libstdc++-v3/doc/html/ext/lwg-active.html

   1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd">
   2 <html><head>
   3 <meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
   4
   5
   6 <title>C++ Standard Library Active Issues List</title>
   7 <style type="text/css">
   8 p {text-align:justify}
   9 li {text-align:justify}
  10 blockquote.note
  11 {
  12     background-color:#E0E0E0;
  13     padding-left: 15px;
  14     padding-right: 15px;
  15     padding-top: 1px;
  16     padding-bottom: 1px;
  17 }
  18 ins {background-color:#A0FFA0}
  19 del {background-color:#FFA0A0}
  20 </style>
  21 </head><body>
  22 <table>
  23 <tbody><tr>
  24 <td align="left">Doc. no.</td>
  25 <td align="left">N3011=09-0201</td>
  26 </tr>
  27 <tr>
  28 <td align="left">Date:</td>
  29 <td align="left">2009-11-08</td>
  30 </tr>
  31 <tr>
  32 <td align="left">Project:</td>
  33 <td align="left">Programming Language C++</td>
  34 </tr>
  35 <tr>
  36 <td align="left">Reply to:</td>
  37 <td align="left">Howard Hinnant &lt;<a href="mailto:howard.hinnant@gmail.com">howard.hinnant@gmail.com</a>&gt;</td>
  38 </tr>
  39 </tbody></table>
  40 <h1>C++ Standard Library Active Issues List (Revision R68)</h1>
  41
  42   <p>Reference ISO/IEC IS 14882:2003(E)</p>
  43   <p>Also see:</p>
  44   <ul>
  45       <li><a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-toc.html">Table of Contents</a> for all library issues.</li>
  46       <li><a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html">Index by Section</a> for all library issues.</li>
  47       <li><a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html">Index by Status</a> for all library issues.</li>
  48       <li><a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html">Library Defect Reports List</a></li>
  49       <li><a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html">Library Closed Issues List</a></li>
  50   </ul>
  51   <p>The purpose of this document is to record the status of issues
  52   which have come before the Library Working Group (LWG) of the ANSI
  53   (J16) and ISO (WG21) C++ Standards Committee. Issues represent
  54   potential defects in the ISO/IEC IS 14882:2003(E) document.
  55   </p>
  56
  57   <p>This document contains only library issues which are actively being
  58   considered by the Library Working Group, i.e., issues which have a
  59   status of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>,
  60   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>, or <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>. See
  61   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html">Library Defect Reports List</a> for issues considered defects and
  62   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html">Library Closed Issues List</a> for issues considered closed.</p>
  63
  64   <p>The issues in these lists are not necessarily formal ISO Defect
  65   Reports (DR's). While some issues will eventually be elevated to
  66   official Defect Report status, other issues will be disposed of in
  67   other ways. See <a href="#Status">Issue Status</a>.</p>
  68
  69   <p>Prior to Revision 14, library issues lists existed in two slightly
  70   different versions; a Committee Version and a Public
  71   Version. Beginning with Revision 14 the two versions were combined
  72   into a single version.</p>
  73
  74   <p>This document includes <i>[bracketed italicized notes]</i> as a
  75   reminder to the LWG of current progress on issues. Such notes are
  76   strictly unofficial and should be read with caution as they may be
  77   incomplete or incorrect. Be aware that LWG support for a particular
  78   resolution can quickly change if new viewpoints or killer examples are
  79   presented in subsequent discussions.</p>
  80
  81   <p>For the most current official version of this document see
  82   <a href="http://www.open-std.org/jtc1/sc22/wg21/">http://www.open-std.org/jtc1/sc22/wg21/</a>.
  83   Requests for further information about this document should include
  84   the document number above, reference ISO/IEC 14882:2003(E), and be
  85   submitted to Information Technology Industry Council (ITI), 1250 Eye
  86   Street NW, Washington, DC 20005.</p>
  87
  88   <p>Public information as to how to obtain a copy of the C++ Standard,
  89   join the standards committee, submit an issue, or comment on an issue
  90   can be found in the comp.std.c++ FAQ.
  91   </p>
  92
  93 <p><a name="submit_issue"></a><b>How to submit an issue</b></p>
  94
  95 <ol type="A">
  96 <a name="submit_issue_A"></a><li>
  97 Mail your issue to the author of this list.
  98 </li>
  99 <a name="submit_issue_B"></a><li>
 100 Specify a short descriptive title.  If you fail to do so, the subject line of your
 101 mail will be used as the issue title.
 102 </li>
 103 <a name="submit_issue_C"></a><li>
 104 If the "From" on your email is not the name you wish to appear as issue submitter,
 105 then specify issue submitter.
 106 </li>
 107 <a name="submit_issue_D"></a><li>
 108 Provide a brief discussion of the problem you wish to correct.  Refer to the latest
 109 working draft or standard using [section.tag] and paragraph numbers where appropriate.
 110 </li>
 111 <a name="submit_issue_E"></a><li>
 112 Provide proposed wording.  This should indicate exactly how you want the standard
 113 to be changed.  General solution statements belong in the discussion area.  This
 114 area contains very clear and specific directions on how to modify the current
 115 draft.  If you are not sure how to word a solution, you may omit this part.
 116 But your chances of a successful issue greatly increase if you attempt wording.
 117 </li>
 118 <a name="submit_issue_F"></a><li>
 119 It is not necessary for you to use html markup.  However, if you want to, you can
 120 &lt;ins&gt;<ins>insert text like this</ins>&lt;/ins&gt; and &lt;del&gt;<del>delete text like
 121 this</del>&lt;/del&gt;.  The only strict requirement is to communicate clearly to
 122 the list maintainer exactly how you want your issue to look.
 123 </li>
 124 <a name="submit_issue_G"></a><li>
 125 It is not necessary for you to specify other html font/formatting
 126 mark-up, but if you do the list maintainer will attempt to respect your
 127 formatting wishes (as described by html markup, or other common idioms).
 128 </li>
 129 <a name="submit_issue_H"></a><li>
 130 It is not necessary for you to specify open date or last modified date (the date
 131 of your mail will be used).
 132 </li>
 133 <a name="submit_issue_I"></a><li>
 134 It is not necessary for you to cross reference other issues, but you can if you
 135 like.  You do not need to form the hyperlinks when you do, the list maintainer will
 136 take care of that.
 137 </li>
 138 <a name="submit_issue_J"></a><li>
 139 One issue per email is best.
 140 </li>
 141 <a name="submit_issue_K"></a><li>
 142 Between the time you submit the issue, and the next mailing deadline
 143 (date at the top of the Revision History), you <em>own</em> this issue.
 144 You control the content, the stuff that is right, the stuff that is
 145 wrong, the format, the misspellings, etc.  You can even make the issue
 146 disappear if you want.  Just let the list maintainer know how you want
 147 it to look, and he will try his best to accommodate you.  After the
 148 issue appears in an official mailing, you no longer enjoy exclusive
 149 ownership of it.
 150 </li>
 151 </ol>
 152
 153
 154 <h2>Revision History</h2>
 155 <ul>
 156 <li>R68:
 157 2009-11-06 post-Santa Cruz mailing.
 158 <ul>
 159 <li><b>Summary:</b><ul>
 160 <li>205 open issues, down by 77.</li>
 161 <li>1055 closed issues, up by 120.</li>
 162 <li>1260 issues total, up by 43.</li>
 163 </ul></li>
 164 <li><b>Details:</b><ul>
 165 <li>Added the following Dup issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1230">1230</a>.</li>
 166 <li>Added the following NAD issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1229">1229</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1236">1236</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1243">1243</a>.</li>
 167 <li>Added the following NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1232">1232</a>.</li>
 168 <li>Added the following NAD Future issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1235">1235</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1242">1242</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1248">1248</a>.</li>
 169 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1218">1218</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1219">1219</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1221">1221</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1222">1222</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1223">1223</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1224">1224</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1225">1225</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1234">1234</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1240">1240</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1244">1244</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1245">1245</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1246">1246</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1249">1249</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1250">1250</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1251">1251</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1252">1252</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1253">1253</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1254">1254</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1255">1255</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1256">1256</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1257">1257</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1258">1258</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1259">1259</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1260">1260</a>.</li>
 170 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1228">1228</a>.</li>
 171 <li>Added the following Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1227">1227</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1237">1237</a>.</li>
 172 <li>Added the following Review issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1247">1247</a>.</li>
 173 <li>Added the following Tentatively NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1233">1233</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1239">1239</a>.</li>
 174 <li>Added the following Tentatively NAD Future issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1238">1238</a>.</li>
 175 <li>Added the following Tentatively Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1220">1220</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1226">1226</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1231">1231</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1241">1241</a>.</li>
 176 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1132">1132</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1148">1148</a>.</li>
 177 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#96">96</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#458">458</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#916">916</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#917">917</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#919">919</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#955">955</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#977">977</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1009">1009</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1020">1020</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1035">1035</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1042">1042</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1051">1051</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1064">1064</a>.</li>
 178 <li>Changed the following issues from Review to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#668">668</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#930">930</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1091">1091</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1102">1102</a>.</li>
 179 <li>Changed the following issues from Tentatively NAD to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#588">588</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#617">617</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#971">971</a>.</li>
 180 <li>Changed the following issues from Tentatively NAD Future to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1062">1062</a>.</li>
 181 <li>Changed the following issues from NAD Concepts to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1143">1143</a>.</li>
 182 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1116">1116</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1117">1117</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1122">1122</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1129">1129</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1145">1145</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1146">1146</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1147">1147</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1155">1155</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1166">1166</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1172">1172</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1174">1174</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1179">1179</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1195">1195</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1196">1196</a>.</li>
 183 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#431">431</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#580">580</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#635">635</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#719">719</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#827">827</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#879">879</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#880">880</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#908">908</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#923">923</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#924">924</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#926">926</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#944">944</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#947">947</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#958">958</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1046">1046</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1048">1048</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1054">1054</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1055">1055</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1075">1075</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1088">1088</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1160">1160</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1161">1161</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1162">1162</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1163">1163</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1165">1165</a>.</li>
 184 <li>Changed the following issues from Review to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#828">828</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#897">897</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1043">1043</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1047">1047</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1049">1049</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1050">1050</a>.</li>
 185 <li>Changed the following issues from New to NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1120">1120</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1150">1150</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1184">1184</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1203">1203</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1217">1217</a>.</li>
 186 <li>Changed the following issues from Open to NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#484">484</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#851">851</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#933">933</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#935">935</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#936">936</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#961">961</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1041">1041</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1053">1053</a>.</li>
 187 <li>Changed the following issues from Tentatively NAD Future to NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1031">1031</a>.</li>
 188 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1118">1118</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1119">1119</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1151">1151</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1153">1153</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1156">1156</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1171">1171</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1173">1173</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1183">1183</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1191">1191</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1211">1211</a>.</li>
 189 <li>Changed the following issues from Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#430">430</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#834">834</a>.</li>
 190 <li>Changed the following issues from Review to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#397">397</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#408">408</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#835">835</a>.</li>
 191 <li>Changed the following issues from Tentatively NAD to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#625">625</a>.</li>
 192 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1123">1123</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1134">1134</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1135">1135</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1136">1136</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1144">1144</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1177">1177</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1194">1194</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1216">1216</a>.</li>
 193 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#296">296</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#471">471</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#485">485</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#539">539</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#816">816</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#860">860</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#865">865</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#872">872</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#920">920</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#932">932</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#939">939</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#940">940</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#960">960</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#963">963</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#974">974</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#978">978</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1011">1011</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1030">1030</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1079">1079</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1098">1098</a>.</li>
 194 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#473">473</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#671">671</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#836">836</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#854">854</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#868">868</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#871">871</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#889">889</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#893">893</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#950">950</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#954">954</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#957">957</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#967">967</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#968">968</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#983">983</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1052">1052</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1094">1094</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1095">1095</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1097">1097</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1100">1100</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1104">1104</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1157">1157</a>.</li>
 195 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1130">1130</a>.</li>
 196 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#556">556</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#631">631</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a>.</li>
 197 <li>Changed the following issues from Open to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1056">1056</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1099">1099</a>.</li>
 198 <li>Changed the following issues from New to Tentatively NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1186">1186</a>.</li>
 199 <li>Changed the following issues from New to Tentatively NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1115">1115</a>.</li>
 200 <li>Changed the following issues from New to Tentatively NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1121">1121</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1201">1201</a>.</li>
 201 <li>Changed the following issues from Open to Tentatively NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1112">1112</a>.</li>
 202 <li>Changed the following issues from New to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1126">1126</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1138">1138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1152">1152</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1158">1158</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1189">1189</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1192">1192</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1208">1208</a>.</li>
 203 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#987">987</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#999">999</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1071">1071</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1090">1090</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1108">1108</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1113">1113</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1114">1114</a>.</li>
 204 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#149">149</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#419">419</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#498">498</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#564">564</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#565">565</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#630">630</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#659">659</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#696">696</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#711">711</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#716">716</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#723">723</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#788">788</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#838">838</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#847">847</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#857">857</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#876">876</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#881">881</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#883">883</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#886">886</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#934">934</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1004">1004</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1178">1178</a>.</li>
 205 <li>Changed the following issues from Tentatively Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1019">1019</a>.</li>
 206 </ul></li>
 207 </ul>
 208 </li>
 209 <li>R67:
 210 2009-09-25 pre-Santa Cruz mailing.
 211 <ul>
 212 <li><b>Summary:</b><ul>
 213 <li>282 open issues, up by 32.</li>
 214 <li>935 closed issues, down by 1.</li>
 215 <li>1217 issues total, up by 31.</li>
 216 </ul></li>
 217 <li><b>Details:</b><ul>
 218 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1187">1187</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1188">1188</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1189">1189</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1190">1190</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1191">1191</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1192">1192</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1193">1193</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1194">1194</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1195">1195</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1196">1196</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1197">1197</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1198">1198</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1199">1199</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1200">1200</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1201">1201</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1202">1202</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1203">1203</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1205">1205</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1206">1206</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1207">1207</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1208">1208</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1209">1209</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1210">1210</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1211">1211</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1212">1212</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1213">1213</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1214">1214</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1215">1215</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1216">1216</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1217">1217</a>.</li>
 219 <li>Changed the following issues from NAD to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#296">296</a>.</li>
 220 <li>Changed the following issues from WP to Pending WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#970">970</a>.</li>
 221 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1052">1052</a>.</li>
 222 <li>Changed the following issues from Ready to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#780">780</a>.</li>
 223 </ul></li>
 224 </ul>
 225 </li>
 226 <li>R66:
 227 2009-07-31 post-Frankfurt mailing.
 228 <ul>
 229 <li><b>Summary:</b><ul>
 230 <li>250 open issues, down by 128.</li>
 231 <li>936 closed issues, up by 171.</li>
 232 <li>1186 issues total, up by 43.</li>
 233 </ul></li>
 234 <li><b>Details:</b><ul>
 235 <li>Added the following NAD issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1164">1164</a>.</li>
 236 <li>Added the following NAD Concepts issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1149">1149</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1167">1167</a>.</li>
 237 <li>Added the following NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1168">1168</a>.</li>
 238 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1144">1144</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1145">1145</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1146">1146</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1147">1147</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1148">1148</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1150">1150</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1151">1151</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1152">1152</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1153">1153</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1154">1154</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1155">1155</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1156">1156</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1158">1158</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1159">1159</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1166">1166</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1169">1169</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1170">1170</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1171">1171</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1172">1172</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1173">1173</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1174">1174</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1175">1175</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1176">1176</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1177">1177</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1179">1179</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1180">1180</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1181">1181</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1182">1182</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1183">1183</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1184">1184</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1185">1185</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1186">1186</a>.</li>
 239 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1160">1160</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1161">1161</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1162">1162</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1163">1163</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1165">1165</a>.</li>
 240 <li>Added the following Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1178">1178</a>.</li>
 241 <li>Added the following Review issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1157">1157</a>.</li>
 242 <li>Changed the following issues from Open to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#750">750</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#895">895</a>.</li>
 243 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#111">111</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#128">128</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#138">138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#190">190</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#219">219</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#290">290</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#309">309</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#342">342</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#343">343</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#382">382</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#394">394</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#398">398</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#417">417</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#418">418</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#421">421</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#459">459</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#466">466</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#492">492</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#502">502</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#503">503</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#546">546</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#573">573</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#582">582</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#585">585</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#597">597</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#606">606</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#614">614</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#632">632</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#721">721</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#747">747</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#751">751</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#833">833</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#941">941</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#992">992</a>.</li>
 244 <li>Changed the following issues from Review to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1003">1003</a>.</li>
 245 <li>Changed the following issues from Tentatively NAD to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#568">568</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#667">667</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#669">669</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#701">701</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#702">702</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#785">785</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#863">863</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#901">901</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#903">903</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#946">946</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#988">988</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#995">995</a>.</li>
 246 <li>Changed the following issues from Tentatively Ready to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1002">1002</a>.</li>
 247 <li>Changed the following issues from New to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1124">1124</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1127">1127</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1128">1128</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1139">1139</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1140">1140</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1141">1141</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1142">1142</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1143">1143</a>.</li>
 248 <li>Changed the following issues from Open to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#902">902</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#989">989</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1000">1000</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1007">1007</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1010">1010</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1015">1015</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1016">1016</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1017">1017</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1018">1018</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1026">1026</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1027">1027</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1028">1028</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1029">1029</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1032">1032</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1036">1036</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1057">1057</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1059">1059</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1072">1072</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1078">1078</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1081">1081</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1082">1082</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1083">1083</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1084">1084</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1085">1085</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1086">1086</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1092">1092</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1096">1096</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1105">1105</a>.</li>
 249 <li>Changed the following issues from Review to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1001">1001</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1005">1005</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1080">1080</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1087">1087</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1111">1111</a>.</li>
 250 <li>Changed the following issues from Tentatively NAD to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#912">912</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#918">918</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1074">1074</a>.</li>
 251 <li>Changed the following issues from Tentatively NAD Editorial to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#927">927</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1109">1109</a>.</li>
 252 <li>Changed the following issues from Tentatively Ready to NAD Concepts: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#906">906</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#913">913</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#914">914</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#928">928</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1024">1024</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1063">1063</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1067">1067</a>.</li>
 253 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#718">718</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#873">873</a>.</li>
 254 <li>Changed the following issues from Tentatively NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#424">424</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#825">825</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#830">830</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#837">837</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#862">862</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#867">867</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#884">884</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#945">945</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#952">952</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#969">969</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#972">972</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#973">973</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#979">979</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1023">1023</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1058">1058</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1060">1060</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1061">1061</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1077">1077</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1101">1101</a>.</li>
 255 <li>Changed the following issues from Tentatively Ready to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1013">1013</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1107">1107</a>.</li>
 256 <li>Changed the following issues from Open to NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#255">255</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#423">423</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#523">523</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#708">708</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#760">760</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#839">839</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#877">877</a>.</li>
 257 <li>Changed the following issues from CD1 to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a>.</li>
 258 <li>Changed the following issues from NAD Editorial to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#299">299</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#484">484</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#556">556</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#631">631</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#704">704</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#742">742</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#811">811</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#870">870</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#872">872</a>.</li>
 259 <li>Changed the following issues from Review to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#879">879</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#919">919</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#939">939</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#987">987</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1009">1009</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1093">1093</a>.</li>
 260 <li>Changed the following issues from Tentatively NAD to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#458">458</a>.</li>
 261 <li>Changed the following issues from Tentatively NAD Future to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#96">96</a>.</li>
 262 <li>Changed the following issues from Tentatively Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#910">910</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#915">915</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#932">932</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#940">940</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#974">974</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#999">999</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1011">1011</a>.</li>
 263 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#149">149</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#419">419</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#430">430</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#498">498</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#564">564</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#565">565</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#630">630</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#659">659</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#696">696</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#711">711</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#716">716</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#723">723</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#788">788</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#834">834</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#838">838</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#847">847</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#857">857</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#876">876</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#881">881</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#883">883</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#886">886</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1004">1004</a>.</li>
 264 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#780">780</a>.</li>
 265 <li>Changed the following issues from Tentatively NAD to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>.</li>
 266 <li>Changed the following issues from Tentatively Ready to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#934">934</a>.</li>
 267 <li>Changed the following issues from NAD to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#871">871</a>.</li>
 268 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#397">397</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#408">408</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#473">473</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#671">671</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#836">836</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#868">868</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#889">889</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#893">893</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#930">930</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#954">954</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#967">967</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#968">968</a>.</li>
 269 <li>Changed the following issues from Tentatively NAD to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#668">668</a>.</li>
 270 <li>Changed the following issues from Tentatively Ready to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#950">950</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1100">1100</a>.</li>
 271 <li>Changed the following issues from Open to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#588">588</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#617">617</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#625">625</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#971">971</a>.</li>
 272 <li>Changed the following issues from Open to Tentatively NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1031">1031</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1062">1062</a>.</li>
 273 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1019">1019</a>.</li>
 274 <li>Changed the following issues from Tentatively Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#688">688</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#765">765</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#810">810</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#814">814</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#869">869</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#878">878</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#888">888</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#890">890</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#898">898</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#899">899</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#904">904</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#907">907</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#909">909</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#922">922</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#925">925</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#931">931</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#938">938</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#943">943</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#948">948</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#949">949</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#965">965</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#970">970</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#975">975</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#981">981</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#982">982</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#984">984</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#986">986</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#990">990</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#991">991</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#993">993</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#994">994</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#997">997</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#998">998</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1006">1006</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1014">1014</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1021">1021</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1037">1037</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1038">1038</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1039">1039</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1040">1040</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1044">1044</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1045">1045</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1065">1065</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1066">1066</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1070">1070</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1073">1073</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1103">1103</a>.</li>
 275 </ul></li>
 276 </ul>
 277 </li>
 278 <li>R65:
 279 2009-06-19 pre-Frankfurt mailing.
 280 <ul>
 281 <li><b>Summary:</b><ul>
 282 <li>378 open issues, up by 32.</li>
 283 <li>765 closed issues, up by 0.</li>
 284 <li>1143 issues total, up by 32.</li>
 285 </ul></li>
 286 <li><b>Details:</b><ul>
 287 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1115">1115</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1116">1116</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1117">1117</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1118">1118</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1119">1119</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1120">1120</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1121">1121</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1122">1122</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1123">1123</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1124">1124</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1125">1125</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1126">1126</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1127">1127</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1128">1128</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1129">1129</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1130">1130</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1131">1131</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1132">1132</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1133">1133</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1134">1134</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1135">1135</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1136">1136</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1137">1137</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1138">1138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1139">1139</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1140">1140</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1141">1141</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1142">1142</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1143">1143</a>.</li>
 288 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1112">1112</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1113">1113</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1114">1114</a>.</li>
 289 <li>Changed the following issues from Pending NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#937">937</a>.</li>
 290 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#696">696</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#716">716</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#865">865</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#900">900</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#911">911</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#916">916</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#917">917</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#920">920</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#933">933</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#935">935</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#941">941</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#947">947</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#951">951</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#953">953</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#954">954</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#955">955</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#956">956</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#977">977</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#978">978</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#985">985</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#989">989</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#996">996</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1033">1033</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1054">1054</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1056">1056</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1057">1057</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1059">1059</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1062">1062</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1068">1068</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1069">1069</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1071">1071</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1072">1072</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1076">1076</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1090">1090</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1092">1092</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1096">1096</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1098">1098</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1099">1099</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1105">1105</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1106">1106</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1108">1108</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1110">1110</a>.</li>
 291 <li>Changed the following issues from Review to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#971">971</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#992">992</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1004">1004</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1010">1010</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1015">1015</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1019">1019</a>.</li>
 292 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#780">780</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#835">835</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#897">897</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#919">919</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#939">939</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#957">957</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#983">983</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1001">1001</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1080">1080</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1091">1091</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1093">1093</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1094">1094</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1095">1095</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1097">1097</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1102">1102</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1104">1104</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1111">1111</a>.</li>
 293 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#987">987</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1087">1087</a>.</li>
 294 <li>Changed the following issues from New to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#568">568</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#701">701</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#702">702</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#785">785</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#863">863</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#903">903</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#912">912</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#918">918</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#946">946</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#995">995</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1074">1074</a>.</li>
 295 <li>Changed the following issues from Open to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#458">458</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#667">667</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#668">668</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#669">669</a>.</li>
 296 <li>Changed the following issues from Review to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#901">901</a>.</li>
 297 <li>Changed the following issues from Tentatively Ready to Tentatively NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#988">988</a>.</li>
 298 <li>Changed the following issues from New to Tentatively NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#837">837</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#862">862</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#867">867</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#927">927</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#945">945</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#952">952</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#969">969</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#972">972</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#973">973</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#979">979</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1058">1058</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1060">1060</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1061">1061</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1077">1077</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1101">1101</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1109">1109</a>.</li>
 299 <li>Changed the following issues from Open to Tentatively NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#424">424</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#825">825</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#830">830</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#884">884</a>.</li>
 300 <li>Changed the following issues from Review to Tentatively NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1023">1023</a>.</li>
 301 <li>Changed the following issues from Open to Tentatively NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#96">96</a>.</li>
 302 <li>Changed the following issues from New to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#810">810</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#898">898</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#906">906</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#910">910</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#913">913</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#914">914</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#915">915</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#925">925</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#974">974</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#981">981</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#982">982</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#984">984</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#990">990</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#998">998</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#999">999</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1063">1063</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1067">1067</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1070">1070</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1073">1073</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1100">1100</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1103">1103</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1107">1107</a>.</li>
 303 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#688">688</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#814">814</a>.</li>
 304 <li>Changed the following issues from Review to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#899">899</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#907">907</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#909">909</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#934">934</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#938">938</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#940">940</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#943">943</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#950">950</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#965">965</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#970">970</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#975">975</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#986">986</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#991">991</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#993">993</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#994">994</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#997">997</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1002">1002</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1006">1006</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1011">1011</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1013">1013</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1014">1014</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1021">1021</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1024">1024</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1037">1037</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1038">1038</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1039">1039</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1040">1040</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1044">1044</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1045">1045</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1065">1065</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1066">1066</a>.</li>
 305 </ul></li>
 306 </ul>
 307 </li>
 308 <li>R64:
 309 2009-05-01 mid-term mailing.
 310 <ul>
 311 <li><b>Summary:</b><ul>
 312 <li>346 open issues, up by 19.</li>
 313 <li>765 closed issues, up by 0.</li>
 314 <li>1111 issues total, up by 19.</li>
 315 </ul></li>
 316 <li><b>Details:</b><ul>
 317 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1093">1093</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1094">1094</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1095">1095</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1096">1096</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1097">1097</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1098">1098</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1099">1099</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1100">1100</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1101">1101</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1102">1102</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1103">1103</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1104">1104</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1105">1105</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1106">1106</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1107">1107</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1108">1108</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1109">1109</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1110">1110</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1111">1111</a>.</li>
 318 <li>Changed the following issues from DR to CD1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#130">130</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#386">386</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#406">406</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#409">409</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#413">413</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#434">434</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#438">438</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#444">444</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#445">445</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#455">455</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#457">457</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#460">460</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#469">469</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#533">533</a>.</li>
 319 <li>Changed the following issues from Review to New: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1070">1070</a>.</li>
 320 </ul></li>
 321 </ul>
 322 </li>
 323 <li>R63:
 324 2009-03-20 post-Summit mailing.
 325 <ul>
 326 <li><b>Summary:</b><ul>
 327 <li>327 open issues, up by 96.</li>
 328 <li>765 closed issues, up by 14.</li>
 329 <li>1092 issues total, up by 110.</li>
 330 </ul></li>
 331 <li><b>Details:</b><ul>
 332 <li>Added the following NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1022">1022</a>.</li>
 333 <li>Added the following NAD Future issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1025">1025</a>.</li>
 334 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#983">983</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#984">984</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#985">985</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#989">989</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#990">990</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#995">995</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#996">996</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#998">998</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#999">999</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1001">1001</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1033">1033</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1054">1054</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1056">1056</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1057">1057</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1058">1058</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1059">1059</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1060">1060</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1061">1061</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1062">1062</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1063">1063</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1067">1067</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1068">1068</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1069">1069</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1071">1071</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1072">1072</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1073">1073</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1074">1074</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1076">1076</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1077">1077</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1080">1080</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1090">1090</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1091">1091</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1092">1092</a>.</li>
 335 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#987">987</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1000">1000</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1007">1007</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1008">1008</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1016">1016</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1017">1017</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1018">1018</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1020">1020</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1026">1026</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1027">1027</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1028">1028</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1029">1029</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1030">1030</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1031">1031</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1032">1032</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1034">1034</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1035">1035</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1036">1036</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1041">1041</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1042">1042</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1046">1046</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1048">1048</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1051">1051</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1052">1052</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1053">1053</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1055">1055</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1064">1064</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1075">1075</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1078">1078</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1079">1079</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1081">1081</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1082">1082</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1083">1083</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1084">1084</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1085">1085</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1086">1086</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1087">1087</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1088">1088</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1089">1089</a>.</li>
 336 <li>Added the following Review issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#986">986</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#991">991</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#992">992</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#993">993</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#994">994</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#997">997</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1002">1002</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1003">1003</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1004">1004</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1005">1005</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1006">1006</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1009">1009</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1010">1010</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1011">1011</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1013">1013</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1014">1014</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1015">1015</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1019">1019</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1021">1021</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1023">1023</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1024">1024</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1037">1037</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1038">1038</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1039">1039</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1040">1040</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1043">1043</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1044">1044</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1045">1045</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1047">1047</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1049">1049</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1050">1050</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1065">1065</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1066">1066</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1070">1070</a>.</li>
 337 <li>Added the following Tentatively Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#988">988</a>.</li>
 338 <li>Changed the following issues from New to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#905">905</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#942">942</a>.</li>
 339 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#980">980</a>.</li>
 340 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#874">874</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#875">875</a>.</li>
 341 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#732">732</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#793">793</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#794">794</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#800">800</a>.</li>
 342 <li>Changed the following issues from Pending NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#683">683</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#892">892</a>.</li>
 343 <li>Changed the following issues from Ready to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#803">803</a>.</li>
 344 <li>Changed the following issues from NAD to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#466">466</a>.</li>
 345 <li>Changed the following issues from NAD Future to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#111">111</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#138">138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#149">149</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#219">219</a>.</li>
 346 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#880">880</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#891">891</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#893">893</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#902">902</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#908">908</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#923">923</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#924">924</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#926">926</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#930">930</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#936">936</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#944">944</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#958">958</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#959">959</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#960">960</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#961">961</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#963">963</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#964">964</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#966">966</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#967">967</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#968">968</a>.</li>
 347 <li>Changed the following issues from Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#788">788</a>.</li>
 348 <li>Changed the following issues from New to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#937">937</a>.</li>
 349 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#879">879</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#899">899</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#901">901</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#907">907</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#909">909</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#934">934</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#938">938</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#940">940</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#943">943</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#950">950</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#965">965</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#970">970</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#971">971</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#975">975</a>.</li>
 350 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>.</li>
 351 <li>Changed the following issues from New to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#904">904</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#922">922</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#928">928</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#931">931</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#932">932</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#948">948</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#949">949</a>.</li>
 352 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#890">890</a>.</li>
 353 <li>Changed the following issues from Review to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#765">765</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#869">869</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#878">878</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#888">888</a>.</li>
 354 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#752">752</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#753">753</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#758">758</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#821">821</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#866">866</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#894">894</a>.</li>
 355 </ul></li>
 356 </ul>
 357 </li>
 358 <li>R62:
 359 2009-02-06 pre-Summit mailing.
 360 <ul>
 361 <li><b>Summary:</b><ul>
 362 <li>231 open issues, up by 44.</li>
 363 <li>751 closed issues, up by 0.</li>
 364 <li>982 issues total, up by 44.</li>
 365 </ul></li>
 366 <li><b>Details:</b><ul>
 367 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#939">939</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#940">940</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#941">941</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#942">942</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#943">943</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#944">944</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#945">945</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#946">946</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#947">947</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#948">948</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#949">949</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#950">950</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#951">951</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#952">952</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#953">953</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#954">954</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#955">955</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#956">956</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#957">957</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#958">958</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#959">959</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#960">960</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#961">961</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#963">963</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#964">964</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#965">965</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#966">966</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#967">967</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#968">968</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#969">969</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#970">970</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#971">971</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#972">972</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#973">973</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#974">974</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#975">975</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#977">977</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#978">978</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#979">979</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#980">980</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#981">981</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#982">982</a>.</li>
 368 </ul></li>
 369 </ul>
 370 </li>
 371 <li>R61:
 372 2008-12-05 mid-term mailing.
 373 <ul>
 374 <li><b>Summary:</b><ul>
 375 <li>187 open issues, up by 20.</li>
 376 <li>751 closed issues, up by 0.</li>
 377 <li>938 issues total, up by 20.</li>
 378 </ul></li>
 379 <li><b>Details:</b><ul>
 380 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#919">919</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#920">920</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#922">922</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#923">923</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#924">924</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#925">925</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#926">926</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#927">927</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#928">928</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#930">930</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#931">931</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#932">932</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#933">933</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#934">934</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#935">935</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#936">936</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#937">937</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#938">938</a>.</li>
 381 </ul></li>
 382 </ul>
 383 </li>
 384 <li>R60:
 385 2008-10-03 post-San Francisco mailing.
 386 <ul>
 387 <li><b>Summary:</b><ul>
 388 <li>167 open issues, down by 25.</li>
 389 <li>751 closed issues, up by 65.</li>
 390 <li>918 issues total, up by 40.</li>
 391 </ul></li>
 392 <li><b>Details:</b><ul>
 393 <li>Added the following CD1 issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#882">882</a>.</li>
 394 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#879">879</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#880">880</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#891">891</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#893">893</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#897">897</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#898">898</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#899">899</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#900">900</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#901">901</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#902">902</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#903">903</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#904">904</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#905">905</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#906">906</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#907">907</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#908">908</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#909">909</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#910">910</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#911">911</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#912">912</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#913">913</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#914">914</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#915">915</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#916">916</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#917">917</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#918">918</a>.</li>
 395 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#881">881</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#883">883</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#884">884</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#885">885</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#886">886</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#887">887</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#889">889</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#890">890</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#895">895</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#896">896</a>.</li>
 396 <li>Added the following Pending NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#892">892</a>.</li>
 397 <li>Added the following Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#894">894</a>.</li>
 398 <li>Added the following Review issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#888">888</a>.</li>
 399 <li>Changed the following issues from New to CD1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#818">818</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#820">820</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#843">843</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#845">845</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#846">846</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#856">856</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#858">858</a>.</li>
 400 <li>Changed the following issues from Ready to CD1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#180">180</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#387">387</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#396">396</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#522">522</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#691">691</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#713">713</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#714">714</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#720">720</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#728">728</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#762">762</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#769">769</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#771">771</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#772">772</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#776">776</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#779">779</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#787">787</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#805">805</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#806">806</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#807">807</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#808">808</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#809">809</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#813">813</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#824">824</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#829">829</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#842">842</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#844">844</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#848">848</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#850">850</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#852">852</a>.</li>
 401 <li>Changed the following issues from Review to CD1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#23">23</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#692">692</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#698">698</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#709">709</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#734">734</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#804">804</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a>.</li>
 402 <li>Changed the following issues from WP to CD1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#44">44</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#49">49</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#76">76</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#91">91</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#92">92</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#98">98</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#109">109</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#117">117</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#118">118</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#120">120</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#123">123</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#136">136</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#153">153</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#165">165</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#167">167</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#171">171</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#179">179</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#182">182</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#183">183</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#184">184</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#185">185</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#186">186</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#187">187</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#198">198</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#200">200</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#201">201</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#202">202</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#206">206</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#214">214</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#221">221</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#225">225</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#226">226</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#228">228</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#229">229</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#230">230</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#231">231</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#232">232</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#233">233</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#234">234</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#235">235</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#237">237</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#238">238</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#239">239</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#240">240</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#241">241</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#242">242</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#243">243</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#247">247</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#248">248</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#250">250</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#251">251</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#252">252</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#253">253</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#254">254</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#256">256</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#258">258</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#259">259</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#260">260</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#261">261</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#262">262</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#263">263</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#264">264</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#265">265</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#266">266</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#268">268</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#270">270</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#271">271</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#272">272</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#273">273</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#274">274</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#275">275</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#276">276</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#278">278</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#280">280</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#281">281</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#282">282</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#283">283</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#284">284</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#285">285</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#286">286</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#288">288</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#291">291</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#292">292</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#294">294</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#295">295</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#297">297</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#298">298</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#300">300</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#301">301</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#303">303</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#305">305</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#306">306</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#307">307</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#308">308</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#310">310</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#311">311</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#312">312</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#315">315</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#316">316</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#317">317</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#318">318</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#319">319</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#320">320</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#321">321</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#322">322</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#324">324</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#325">325</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#327">327</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#328">328</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#329">329</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#331">331</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#333">333</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#334">334</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#335">335</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#336">336</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#337">337</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#338">338</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#339">339</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#340">340</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#341">341</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#345">345</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#346">346</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#347">347</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#349">349</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#352">352</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#354">354</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#355">355</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#358">358</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#359">359</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#360">360</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#362">362</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#363">363</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#364">364</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#365">365</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#369">369</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#370">370</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#371">371</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#373">373</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#375">375</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#376">376</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#379">379</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#380">380</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#381">381</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#383">383</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#384">384</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#389">389</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#391">391</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#395">395</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#400">400</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#401">401</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#402">402</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#403">403</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#404">404</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#405">405</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#407">407</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#410">410</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#411">411</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#412">412</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#414">414</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#415">415</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#416">416</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#420">420</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#422">422</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#425">425</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#426">426</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#428">428</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#432">432</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#435">435</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#436">436</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#441">441</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#442">442</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#443">443</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#448">448</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#449">449</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#453">453</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#456">456</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#461">461</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#464">464</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#465">465</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#467">467</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#468">468</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#474">474</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#475">475</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#478">478</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#488">488</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#495">495</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#496">496</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#497">497</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#505">505</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#507">507</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#508">508</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#519">519</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#520">520</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#521">521</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#524">524</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#530">530</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#534">534</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#535">535</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#537">537</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#538">538</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#540">540</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#541">541</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#542">542</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#543">543</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#545">545</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#550">550</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#551">551</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#552">552</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#559">559</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#561">561</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#562">562</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#563">563</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#566">566</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#567">567</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#574">574</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#575">575</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#576">576</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#577">577</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#578">578</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#581">581</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#586">586</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#589">589</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#593">593</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#595">595</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#596">596</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#607">607</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#608">608</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#609">609</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#610">610</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#611">611</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#612">612</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#613">613</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#616">616</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#618">618</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#619">619</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#620">620</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#621">621</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#622">622</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#623">623</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#624">624</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#628">628</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#634">634</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#638">638</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#640">640</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#643">643</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#646">646</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#650">650</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#651">651</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#652">652</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#654">654</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#655">655</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#660">660</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#661">661</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#664">664</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#665">665</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#666">666</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#674">674</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#677">677</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#678">678</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#679">679</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#680">680</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#681">681</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#682">682</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#685">685</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#687">687</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#689">689</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#693">693</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#694">694</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#695">695</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#699">699</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#700">700</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#703">703</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#705">705</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#706">706</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#710">710</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#712">712</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#715">715</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#722">722</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#740">740</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#743">743</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#744">744</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#746">746</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#749">749</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#755">755</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#759">759</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#761">761</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#766">766</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#768">768</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#770">770</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#775">775</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#777">777</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#778">778</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#781">781</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#782">782</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#783">783</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#789">789</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#792">792</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#798">798</a>.</li>
 403 <li>Changed the following issues from Open to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#670">670</a>.</li>
 404 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#849">849</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#855">855</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#871">871</a>.</li>
 405 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#454">454</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#832">832</a>.</li>
 406 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#811">811</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#812">812</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#841">841</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#864">864</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#870">870</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#872">872</a>.</li>
 407 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#299">299</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#484">484</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#556">556</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#631">631</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#704">704</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#742">742</a>.</li>
 408 <li>Changed the following issues from Pending NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#717">717</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#725">725</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#738">738</a>.</li>
 409 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#721">721</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#751">751</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#814">814</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#816">816</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#819">819</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#827">827</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#836">836</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#838">838</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#847">847</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#857">857</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#860">860</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#861">861</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#868">868</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#873">873</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#876">876</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#877">877</a>.</li>
 410 <li>Changed the following issues from Pending NAD Editorial to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#424">424</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#625">625</a>.</li>
 411 <li>Changed the following issues from Review to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#851">851</a>.</li>
 412 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#788">788</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#821">821</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#866">866</a>.</li>
 413 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#753">753</a>.</li>
 414 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#752">752</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#758">758</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#803">803</a>.</li>
 415 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#765">765</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#854">854</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#869">869</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#878">878</a>.</li>
 416 <li>Changed the following issues from TC to TC1: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1">1</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#3">3</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#5">5</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#7">7</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#8">8</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#9">9</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#11">11</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#13">13</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#14">14</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#15">15</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#16">16</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#17">17</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#18">18</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#19">19</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#20">20</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#21">21</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#22">22</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#24">24</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#25">25</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#26">26</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#27">27</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#28">28</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#29">29</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#30">30</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#31">31</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#32">32</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#33">33</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#34">34</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#35">35</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#36">36</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#37">37</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#38">38</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#39">39</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#40">40</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#41">41</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#42">42</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#46">46</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#47">47</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#48">48</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#50">50</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#51">51</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#52">52</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#53">53</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#54">54</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#55">55</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#56">56</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#57">57</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#59">59</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#60">60</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#61">61</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#62">62</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#63">63</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#64">64</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#66">66</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#68">68</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#69">69</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#70">70</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#71">71</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#74">74</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#75">75</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#78">78</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#79">79</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#80">80</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#83">83</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#86">86</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#90">90</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#106">106</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#108">108</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#110">110</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#112">112</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#114">114</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#115">115</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#119">119</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#122">122</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#124">124</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#125">125</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#126">126</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#127">127</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#129">129</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#132">132</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#133">133</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#134">134</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#137">137</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#139">139</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#141">141</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#142">142</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#144">144</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#146">146</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#147">147</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#148">148</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#150">150</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#151">151</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#152">152</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#154">154</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#155">155</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#156">156</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#158">158</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#159">159</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#160">160</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#161">161</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#164">164</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#168">168</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#169">169</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#170">170</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#172">172</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#173">173</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#174">174</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#175">175</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#176">176</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#181">181</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#189">189</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#193">193</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#195">195</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#199">199</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#208">208</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#209">209</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#210">210</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#211">211</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#212">212</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#217">217</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#220">220</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#222">222</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#223">223</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#224">224</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#227">227</a>.</li>
 417 </ul></li>
 418 </ul>
 419 </li>
 420 <li>R59:
 421 2008-08-22 pre-San Francisco mailing.
 422 <ul>
 423 <li><b>Summary:</b><ul>
 424 <li>192 open issues, up by 9.</li>
 425 <li>686 closed issues, up by 0.</li>
 426 <li>878 issues total, up by 9.</li>
 427 </ul></li>
 428 <li><b>Details:</b><ul>
 429 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#870">870</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#871">871</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#872">872</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#873">873</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#874">874</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#875">875</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#876">876</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#877">877</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#878">878</a>.</li>
 430 </ul></li>
 431 </ul>
 432 </li>
 433 <li>R58:
 434 2008-07-28 mid-term mailing.
 435 <ul>
 436 <li><b>Summary:</b><ul>
 437 <li>183 open issues, up by 12.</li>
 438 <li>686 closed issues, down by 4.</li>
 439 <li>869 issues total, up by 8.</li>
 440 </ul></li>
 441 <li><b>Details:</b><ul>
 442 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#862">862</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#863">863</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#864">864</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#865">865</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#866">866</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#867">867</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#868">868</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#869">869</a>.</li>
 443 <li>Changed the following issues from Pending NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#393">393</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#557">557</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#592">592</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#754">754</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#757">757</a>.</li>
 444 <li>Changed the following issues from Pending WP to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>.</li>
 445 <li>Changed the following issues from WP to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#387">387</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>.</li>
 446 <li>Changed the following issues from Pending NAD Editorial to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#709">709</a>.</li>
 447 </ul></li>
 448 </ul>
 449 </li>
 450 <li>R57:
 451 2008-06-27 post-Sophia Antipolis mailing.
 452 <ul>
 453 <li><b>Summary:</b><ul>
 454 <li>171 open issues, down by 20.</li>
 455 <li>690 closed issues, up by 43.</li>
 456 <li>861 issues total, up by 23.</li>
 457 </ul></li>
 458 <li><b>Details:</b><ul>
 459 <li>Added the following NAD issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#840">840</a>.</li>
 460 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#841">841</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#843">843</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#845">845</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#846">846</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#847">847</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#849">849</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#854">854</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#855">855</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#856">856</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#857">857</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#858">858</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#860">860</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#861">861</a>.</li>
 461 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#839">839</a>.</li>
 462 <li>Added the following Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#842">842</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#844">844</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#848">848</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#850">850</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#852">852</a>.</li>
 463 <li>Added the following Review issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#851">851</a>.</li>
 464 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#826">826</a>.</li>
 465 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#570">570</a>.</li>
 466 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#786">786</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#831">831</a>.</li>
 467 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#756">756</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#767">767</a>.</li>
 468 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#723">723</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#726">726</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#794">794</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#815">815</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#825">825</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#830">830</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#833">833</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#834">834</a>.</li>
 469 <li>Changed the following issues from Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#471">471</a>.</li>
 470 <li>Changed the following issues from Review to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#539">539</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#711">711</a>.</li>
 471 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#713">713</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#714">714</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#769">769</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#772">772</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#779">779</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#787">787</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#805">805</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#806">806</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#807">807</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#808">808</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#809">809</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#813">813</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#824">824</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#829">829</a>.</li>
 472 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#180">180</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#396">396</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#522">522</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#720">720</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#762">762</a>.</li>
 473 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#691">691</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#728">728</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#771">771</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#776">776</a>.</li>
 474 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#692">692</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#698">698</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#752">752</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#804">804</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#828">828</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#832">832</a>.</li>
 475 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#23">23</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#734">734</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#803">803</a>.</li>
 476 <li>Changed the following issues from Ready to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#758">758</a>.</li>
 477 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#387">387</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#550">550</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#574">574</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#595">595</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#596">596</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#612">612</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#618">618</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#638">638</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#685">685</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#710">710</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#715">715</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#722">722</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#740">740</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#743">743</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#744">744</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#746">746</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#749">749</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#755">755</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#759">759</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#761">761</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#766">766</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#768">768</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#770">770</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#775">775</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#777">777</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#778">778</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#781">781</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#782">782</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#783">783</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#789">789</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#792">792</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#798">798</a>.</li>
 478 </ul></li>
 479 </ul>
 480 </li>
 481 <li>R56:
 482 2008-05-16 pre-Sophia Antipolis mailing.
 483 <ul>
 484 <li><b>Summary:</b><ul>
 485 <li>191 open issues, up by 24.</li>
 486 <li>647 closed issues, up by 1.</li>
 487 <li>838 issues total, up by 25.</li>
 488 </ul></li>
 489 <li><b>Details:</b><ul>
 490 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#814">814</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#815">815</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#816">816</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#818">818</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#819">819</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#820">820</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#821">821</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#822">822</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#824">824</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#825">825</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#826">826</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#827">827</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#828">828</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#829">829</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#830">830</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#831">831</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#832">832</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#833">833</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#834">834</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#835">835</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#836">836</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#837">837</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#838">838</a>.</li>
 491 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#802">802</a>.</li>
 492 </ul></li>
 493 </ul>
 494 </li>
 495 <li>R55:
 496 2008-03-14 post-Bellevue mailing.
 497 <ul>
 498 <li><b>Summary:</b><ul>
 499 <li>167 open issues, down by 39.</li>
 500 <li>646 closed issues, up by 65.</li>
 501 <li>813 issues total, up by 26.</li>
 502 </ul></li>
 503 <li><b>Details:</b><ul>
 504 <li>Added the following Dup issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#795">795</a>.</li>
 505 <li>Added the following NAD issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#790">790</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#791">791</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#796">796</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#797">797</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#799">799</a>.</li>
 506 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#788">788</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#794">794</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#802">802</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#804">804</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#805">805</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#806">806</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#807">807</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#808">808</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#809">809</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#810">810</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#811">811</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#812">812</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#813">813</a>.</li>
 507 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#793">793</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#800">800</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#801">801</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#803">803</a>.</li>
 508 <li>Added the following Ready issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#789">789</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#792">792</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#798">798</a>.</li>
 509 <li>Changed the following issues from NAD Future to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#116">116</a>.</li>
 510 <li>Changed the following issues from NAD Future to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#188">188</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#323">323</a>.</li>
 511 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#729">729</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#730">730</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#731">731</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#733">733</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#735">735</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#736">736</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#737">737</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#739">739</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#741">741</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#745">745</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#748">748</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#763">763</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#764">764</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#773">773</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#784">784</a>.</li>
 512 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#388">388</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#462">462</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#579">579</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#627">627</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#653">653</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#686">686</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#707">707</a>.</li>
 513 <li>Changed the following issues from NAD Future to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#140">140</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#390">390</a>.</li>
 514 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#529">529</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#626">626</a>.</li>
 515 <li>Changed the following issues from Review to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#645">645</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#684">684</a>.</li>
 516 <li>Changed the following issues from NAD Future to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#128">128</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#180">180</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#190">190</a>.</li>
 517 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#617">617</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#718">718</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#719">719</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#720">720</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#732">732</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#734">734</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#742">742</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#747">747</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#750">750</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#753">753</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#756">756</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#760">760</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#762">762</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#767">767</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#774">774</a>.</li>
 518 <li>Changed the following issues from Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#688">688</a>.</li>
 519 <li>Changed the following issues from New to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#709">709</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#717">717</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#725">725</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#738">738</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#754">754</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#757">757</a>.</li>
 520 <li>Changed the following issues from Open to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#424">424</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#557">557</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#625">625</a>.</li>
 521 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#710">710</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#715">715</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#722">722</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#740">740</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#743">743</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#744">744</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#746">746</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#749">749</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#755">755</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#758">758</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#759">759</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#761">761</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#766">766</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#768">768</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#770">770</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#775">775</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#777">777</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#778">778</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#781">781</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#782">782</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#783">783</a>.</li>
 522 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#387">387</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#471">471</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#550">550</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#612">612</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a>.</li>
 523 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#574">574</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#596">596</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#618">618</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#638">638</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#685">685</a>.</li>
 524 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#711">711</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#728">728</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#771">771</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#776">776</a>.</li>
 525 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#539">539</a>.</li>
 526 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#561">561</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#562">562</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#563">563</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#567">567</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#581">581</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#620">620</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#621">621</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#622">622</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#623">623</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#624">624</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#661">661</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#664">664</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#665">665</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#666">666</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#674">674</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#679">679</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#680">680</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#687">687</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#689">689</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#693">693</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#694">694</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#695">695</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#700">700</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#703">703</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#705">705</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#706">706</a>.</li>
 527 <li>Changed the following issues from Tentatively Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a>.</li>
 528 </ul></li>
 529 </ul>
 530 </li>
 531 <li>R54:
 532 2008-02-01 pre-Bellevue mailing.
 533 <ul>
 534 <li><b>Summary:</b><ul>
 535 <li>206 open issues, up by 23.</li>
 536 <li>581 closed issues, up by 0.</li>
 537 <li>787 issues total, up by 23.</li>
 538 </ul></li>
 539 <li><b>Details:</b><ul>
 540 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#765">765</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#766">766</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#767">767</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#768">768</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#769">769</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#770">770</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#771">771</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#772">772</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#773">773</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#774">774</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#775">775</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#776">776</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#777">777</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#778">778</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#779">779</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#780">780</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#781">781</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#782">782</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#783">783</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#784">784</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#785">785</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#786">786</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#787">787</a>.</li>
 541 <li>Changed the following issues from NAD Future to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#105">105</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#348">348</a>.</li>
 542 <li>Changed the following issues from NAD Future to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#353">353</a>.</li>
 543 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#697">697</a>.</li>
 544 <li>Changed the following issues from NAD Future to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#388">388</a>.</li>
 545 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a>.</li>
 546 </ul></li>
 547 </ul>
 548 </li>
 549 <li>R53:
 550 2007-12-09 mid-term mailing.
 551 <ul>
 552 <li><b>Summary:</b><ul>
 553 <li>183 open issues, up by 11.</li>
 554 <li>581 closed issues, down by 1.</li>
 555 <li>764 issues total, up by 10.</li>
 556 </ul></li>
 557 <li><b>Details:</b><ul>
 558 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#755">755</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#756">756</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#757">757</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#758">758</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#759">759</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#760">760</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#761">761</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#762">762</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#763">763</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#764">764</a>.</li>
 559 <li>Changed the following issues from NAD to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>.</li>
 560 <li>Changed the following issues from Pending WP to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#607">607</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#608">608</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#654">654</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#655">655</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#677">677</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#682">682</a>.</li>
 561 </ul></li>
 562 </ul>
 563 </li>
 564 <li>R52:
 565 2007-10-19 post-Kona mailing.
 566 <ul>
 567 <li><b>Summary:</b><ul>
 568 <li>172 open issues, up by 4.</li>
 569 <li>582 closed issues, up by 27.</li>
 570 <li>754 issues total, up by 31.</li>
 571 </ul></li>
 572 <li><b>Details:</b><ul>
 573 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#725">725</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#726">726</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#728">728</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#729">729</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#730">730</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#731">731</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#732">732</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#733">733</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#734">734</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#735">735</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#736">736</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#737">737</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#738">738</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#739">739</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#740">740</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#741">741</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#742">742</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#743">743</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#744">744</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#745">745</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#746">746</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#747">747</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#748">748</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#749">749</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#750">750</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#751">751</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#752">752</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#753">753</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#754">754</a>.</li>
 574 <li>Changed the following issues from NAD Future to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#77">77</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#350">350</a>.</li>
 575 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#639">639</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#657">657</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#663">663</a>.</li>
 576 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#548">548</a>.</li>
 577 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#546">546</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#550">550</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#564">564</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#565">565</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#573">573</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#585">585</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#588">588</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#627">627</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#630">630</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#632">632</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#635">635</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#653">653</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#659">659</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#667">667</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#668">668</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#669">669</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#670">670</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#671">671</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#686">686</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#704">704</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#707">707</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#708">708</a>.</li>
 578 <li>Changed the following issues from New to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#393">393</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#592">592</a>.</li>
 579 <li>Changed the following issues from New to Pending WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#607">607</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#608">608</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#654">654</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#655">655</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#677">677</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#682">682</a>.</li>
 580 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#561">561</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#562">562</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#563">563</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#567">567</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#581">581</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#595">595</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#620">620</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#621">621</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#622">622</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#623">623</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#624">624</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#661">661</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#664">664</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#665">665</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#666">666</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#674">674</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#679">679</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#687">687</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#688">688</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#689">689</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#693">693</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#694">694</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#695">695</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#700">700</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#703">703</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#705">705</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#706">706</a>.</li>
 581 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#680">680</a>.</li>
 582 <li>Changed the following issues from New to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#574">574</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#596">596</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#618">618</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#638">638</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#645">645</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#684">684</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#685">685</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#691">691</a>.</li>
 583 <li>Changed the following issues from New to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#552">552</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#634">634</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#650">650</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#651">651</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#652">652</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#678">678</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#681">681</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#699">699</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#712">712</a>.</li>
 584 <li>Changed the following issues from Open to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#258">258</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#401">401</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#524">524</a>.</li>
 585 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#488">488</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#577">577</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#660">660</a>.</li>
 586 </ul></li>
 587 </ul>
 588 </li>
 589 <li>R51:
 590 2007-09-09 pre-Kona mailing.
 591 <ul>
 592 <li><b>Summary:</b><ul>
 593 <li>168 open issues, up by 15.</li>
 594 <li>555 closed issues, up by 0.</li>
 595 <li>723 issues total, up by 15.</li>
 596 </ul></li>
 597 <li><b>Details:</b><ul>
 598 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#709">709</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#710">710</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#711">711</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#712">712</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#713">713</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#714">714</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#715">715</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#716">716</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#717">717</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#718">718</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#719">719</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#720">720</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#721">721</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#722">722</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#723">723</a>.</li>
 599 </ul></li>
 600 </ul>
 601 </li>
 602 <li>R50:
 603 2007-08-05 post-Toronto mailing.
 604 <ul>
 605 <li><b>Summary:</b><ul>
 606 <li>153 open issues, down by 5.</li>
 607 <li>555 closed issues, up by 17.</li>
 608 <li>708 issues total, up by 12.</li>
 609 </ul></li>
 610 <li><b>Details:</b><ul>
 611 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#697">697</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#698">698</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#699">699</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#700">700</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#701">701</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#702">702</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#703">703</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#704">704</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#705">705</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#706">706</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#707">707</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#708">708</a>.</li>
 612 <li>Changed the following issues from New to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#583">583</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#584">584</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#662">662</a>.</li>
 613 <li>Changed the following issues from Open to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#528">528</a>.</li>
 614 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#637">637</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#647">647</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#658">658</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#690">690</a>.</li>
 615 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#525">525</a>.</li>
 616 <li>Changed the following issues from Pending NAD Editorial to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#553">553</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#571">571</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#591">591</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#633">633</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#636">636</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#641">641</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#642">642</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#648">648</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#649">649</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#656">656</a>.</li>
 617 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#579">579</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#631">631</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#680">680</a>.</li>
 618 <li>Changed the following issues from Pending WP to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#258">258</a>.</li>
 619 <li>Changed the following issues from Ready to Pending WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>.</li>
 620 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#577">577</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#660">660</a>.</li>
 621 <li>Changed the following issues from Open to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#488">488</a>.</li>
 622 <li>Changed the following issues from Open to Review: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a>.</li>
 623 <li>Changed the following issues from Ready to TRDec: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#604">604</a>.</li>
 624 <li>Changed the following issues from DR to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#453">453</a>.</li>
 625 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#551">551</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#566">566</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#628">628</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#640">640</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#643">643</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#646">646</a>.</li>
 626 </ul></li>
 627 </ul>
 628 </li>
 629 <li>R49:
 630 2007-06-23 pre-Toronto mailing.
 631 <ul>
 632 <li><b>Summary:</b><ul>
 633 <li>158 open issues, up by 13.</li>
 634 <li>538 closed issues, up by 7.</li>
 635 <li>696 issues total, up by 20.</li>
 636 </ul></li>
 637 <li><b>Details:</b><ul>
 638 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#677">677</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#678">678</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#679">679</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#680">680</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#681">681</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#682">682</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#684">684</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#685">685</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#686">686</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#687">687</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#688">688</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#689">689</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#690">690</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#691">691</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#692">692</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#693">693</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#694">694</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#695">695</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#696">696</a>.</li>
 639 <li>Added the following Pending NAD Editorial issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#683">683</a>.</li>
 640 <li>Changed the following issues from New to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#587">587</a>.</li>
 641 <li>Changed the following issues from Open to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#590">590</a>.</li>
 642 <li>Changed the following issues from New to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#636">636</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#642">642</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#648">648</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#649">649</a>.</li>
 643 </ul></li>
 644 </ul>
 645 </li>
 646 <li>R48:
 647 2007-05-06 post-Oxford mailing.
 648 <ul>
 649 <li><b>Summary:</b><ul>
 650 <li>145 open issues, down by 33.</li>
 651 <li>531 closed issues, up by 53.</li>
 652 <li>676 issues total, up by 20.</li>
 653 </ul></li>
 654 <li><b>Details:</b><ul>
 655 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#657">657</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#658">658</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#659">659</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#660">660</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#661">661</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#662">662</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#663">663</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#664">664</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#665">665</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#666">666</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#667">667</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#668">668</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#669">669</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#670">670</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#671">671</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#674">674</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#676">676</a>.</li>
 656 <li>Changed the following issues from Tentatively Ready to Dup: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#479">479</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#536">536</a>.</li>
 657 <li>Changed the following issues from Tentatively Ready to NAD: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#385">385</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#466">466</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#470">470</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#515">515</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#526">526</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#547">547</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#560">560</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#572">572</a>.</li>
 658 <li>Changed the following issues from NAD to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#351">351</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#357">357</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#368">368</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#499">499</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#504">504</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#512">512</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#513">513</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#514">514</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#516">516</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#544">544</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#555">555</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#558">558</a>.</li>
 659 <li>Changed the following issues from Tentatively Ready to NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#482">482</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#615">615</a>.</li>
 660 <li>Changed the following issues from NAD_Future to NAD Future: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#77">77</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#105">105</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#111">111</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#116">116</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#128">128</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#138">138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#140">140</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#149">149</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#180">180</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#188">188</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#190">190</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#219">219</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#323">323</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#348">348</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#350">350</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#353">353</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#388">388</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#390">390</a>.</li>
 661 <li>Changed the following issues from Tentatively Ready to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#471">471</a>.</li>
 662 <li>Changed the following issues from New to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#633">633</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#641">641</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#656">656</a>.</li>
 663 <li>Changed the following issues from Tentatively Ready to Pending NAD Editorial: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#553">553</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#571">571</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#591">591</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>.</li>
 664 <li>Changed the following issues from Tentatively Ready to Pending WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#258">258</a>.</li>
 665 <li>Changed the following issues from New to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#566">566</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#628">628</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#640">640</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#643">643</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#646">646</a>.</li>
 666 <li>Changed the following issues from Review to Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#551">551</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#604">604</a>.</li>
 667 <li>Changed the following issues from Ready to TRDec: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#598">598</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#599">599</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#600">600</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#601">601</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#602">602</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#603">603</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#605">605</a>.</li>
 668 <li>Changed the following issues from Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#543">543</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#545">545</a>.</li>
 669 <li>Changed the following issues from Tentatively Ready to WP: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#201">201</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#206">206</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#233">233</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#254">254</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#416">416</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#422">422</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#456">456</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#534">534</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#542">542</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#559">559</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#575">575</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#576">576</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#578">578</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#586">586</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#589">589</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#593">593</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#609">609</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#610">610</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#611">611</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#613">613</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#616">616</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#619">619</a>.</li>
 670 </ul></li>
 671 </ul>
 672 </li>
 673 <li>R47:
 674 2007-03-09 pre-Oxford mailing.
 675 <ul>
 676 <li><b>Summary:</b><ul>
 677 <li>178 open issues, up by 37.</li>
 678 <li>478 closed issues, up by 0.</li>
 679 <li>656 issues total, up by 37.</li>
 680 </ul></li>
 681 <li><b>Details:</b><ul>
 682 <li>Added the following New issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#620">620</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#621">621</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#622">622</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#623">623</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#624">624</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#627">627</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#628">628</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#629">629</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#630">630</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#631">631</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#632">632</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#633">633</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#634">634</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#635">635</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#636">636</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#637">637</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#638">638</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#639">639</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#640">640</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#641">641</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#642">642</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#643">643</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#644">644</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#645">645</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#646">646</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#647">647</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#648">648</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#649">649</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#650">650</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#651">651</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#652">652</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#653">653</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#654">654</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#655">655</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#656">656</a>.</li>
 683 <li>Added the following Open issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#625">625</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#626">626</a>.</li>
 684 <li>Changed the following issues from New to Open: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#570">570</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#580">580</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#582">582</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#590">590</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#612">612</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#614">614</a>.</li>
 685 <li>Changed the following issues from New to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#547">547</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#553">553</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#560">560</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#571">571</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#572">572</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#575">575</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#576">576</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#578">578</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#586">586</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#589">589</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#591">591</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#593">593</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#609">609</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#610">610</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#611">611</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#613">613</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#615">615</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#616">616</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#619">619</a>.</li>
 686 <li>Changed the following issues from Open to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#201">201</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#206">206</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#233">233</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#254">254</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#258">258</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#385">385</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#416">416</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#422">422</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#456">456</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#466">466</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#470">470</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#471">471</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#479">479</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#482">482</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#515">515</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#526">526</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#536">536</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#542">542</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#559">559</a>.</li>
 687 <li>Changed the following issues from Review to Tentatively Ready: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#534">534</a>.</li>
 688 </ul></li>
 689 </ul>
 690 </li>
 691 <li>R46:
 692 2007-01-12 mid-term mailing.
 693 <ul>
 694 <li><b>Summary:</b><ul>
 695 <li>141 open issues, up by 11.</li>
 696 <li>478 closed issues, down by 1.</li>
 697 <li>619 issues total, up by 10.</li>
 698 </ul></li>
 699 <li><b>Details:</b><ul>
 700 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#610">610</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#611">611</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#612">612</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#613">613</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#614">614</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#615">615</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#616">616</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#617">617</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#618">618</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#619">619</a>.</li>
 701 </ul></li>
 702 </ul>
 703 </li>
 704 <li>R45:
 705 2006-11-03 post-Portland mailing.
 706 <ul>
 707 <li><b>Summary:</b><ul>
 708 <li>130 open issues, up by 0.</li>
 709 <li>479 closed issues, up by 17.</li>
 710 <li>609 issues total, up by 17.</li>
 711 </ul></li>
 712 <li><b>Details:</b><ul>
 713 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#520">520</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#521">521</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#530">530</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#535">535</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#537">537</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#538">538</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#540">540</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#541">541</a> to WP.</li>
 714 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#504">504</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#512">512</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#516">516</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#544">544</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#554">554</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#555">555</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#558">558</a> to NAD.</li>
 715 <li>Moved issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#569">569</a> to Dup.</li>
 716 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#523">523</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#524">524</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#542">542</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#556">556</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#557">557</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#559">559</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#597">597</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#606">606</a> to Open.</li>
 717 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#543">543</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#545">545</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#598">598</a> - <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#603">603</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#605">605</a> to Ready.</li>
 718 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#551">551</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#604">604</a> to Review.</li>
 719 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#593">593</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#595">595</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#596">596</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#597">597</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#598">598</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#599">599</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#600">600</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#601">601</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#602">602</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#603">603</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#604">604</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#605">605</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#606">606</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#607">607</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#608">608</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#609">609</a>.</li>
 720 </ul></li>
 721 </ul>
 722 </li>
 723 <li>R44:
 724 2006-09-08 pre-Portland mailing.
 725 <ul>
 726 <li><b>Summary:</b><ul>
 727 <li>130 open issues, up by 6.</li>
 728 <li>462 closed issues, down by 1.</li>
 729 <li>592 issues total, up by 5.</li>
 730 </ul></li>
 731 <li><b>Details:</b><ul>
 732 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#583">583</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#584">584</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#585">585</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#586">586</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#587">587</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#588">588</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#589">589</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#590">590</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#591">591</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#592">592</a>.</li>
 733 </ul></li>
 734 </ul>
 735 </li>
 736 <li>R43:
 737 2006-06-23 mid-term mailing.
 738 <ul>
 739 <li><b>Summary:</b><ul>
 740 <li>124 open issues, up by 14.</li>
 741 <li>463 closed issues, down by 1.</li>
 742 <li>587 issues total, up by 13.</li>
 743 </ul></li>
 744 <li><b>Details:</b><ul>
 745 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#575">575</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#576">576</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#577">577</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#578">578</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#579">579</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#580">580</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#581">581</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#582">582</a>.</li>
 746 <li>Reopened <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#255">255</a>.</li>
 747 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#520">520</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#541">541</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#544">544</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#569">569</a> to Tentatively Ready.</li>
 748 </ul></li>
 749 </ul>
 750 </li>
 751 <li>R42:
 752 2006-04-21 post-Berlin mailing.
 753 <ul>
 754 <li><b>Summary:</b><ul>
 755 <li>110 open issues, down by 16.</li>
 756 <li>464 closed issues, up by 24.</li>
 757 <li>574 issues total, up by 8.</li>
 758 </ul></li>
 759 <li><b>Details:</b><ul>
 760 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#567">567</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#568">568</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#569">569</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#570">570</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#571">571</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#572">572</a>.</li>
 761 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#499">499</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#501">501</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#506">506</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#509">509</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#510">510</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#511">511</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#513">513</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#514">514</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#517">517</a> to NAD.</li>
 762 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#502">502</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#503">503</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#515">515</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#516">516</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#522">522</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#525">525</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#526">526</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#528">528</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#529">529</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#536">536</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#539">539</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#548">548</a> to Open.</li>
 763 <li>Moved issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#504">504</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#512">512</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#521">521</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#530">530</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#535">535</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#537">537</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#538">538</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#540">540</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a> to Ready.</li>
 764 <li>Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#247">247</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#294">294</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#362">362</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#369">369</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#371">371</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#376">376</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#384">384</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#475">475</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#478">478</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#495">495</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#497">497</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#505">505</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#507">507</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#508">508</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#519">519</a> to WP.</li>
 765 <li>Moved issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#534">534</a> to Review.</li>
 766 </ul></li>
 767 </ul>
 768 </li>
 769 <li>R41:
 770 2006-02-24 pre-Berlin mailing.
 771 <ul>
 772 <li><b>Summary:</b><ul>
 773 <li>126 open issues, up by 31.</li>
 774 <li>440 closed issues, up by 0.</li>
 775 <li>566 issues total, up by 31.</li>
 776 </ul></li>
 777 <li><b>Details:</b><ul>
 778 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#536">536</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#537">537</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#538">538</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#539">539</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#540">540</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#541">541</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#542">542</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#543">543</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#544">544</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#545">545</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#546">546</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#547">547</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#548">548</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#549">549</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#550">550</a> ,<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#551">551</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#552">552</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#553">553</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#554">554</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#555">555</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#556">556</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#557">557</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#558">558</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#559">559</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#560">560</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#561">561</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#562">562</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#563">563</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#564">564</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#565">565</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#566">566</a>.</li>
 779 <li>Moved <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#342">342</a> from Ready to Open.</li>
 780 <li>Reopened <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#309">309</a>.</li>
 781 </ul></li>
 782 </ul>
 783 </li>
 784 <li>R40:
 785 2005-12-16 mid-term mailing.
 786 <ul>
 787 <li><b>Summary:</b><ul>
 788 <li>95 open issues.</li>
 789 <li>440 closed issues.</li>
 790 <li>535 issues total.</li>
 791 </ul></li>
 792 <li><b>Details:</b><ul>
 793 <li>Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#529">529</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#530">530</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#531">531</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#532">532</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#533">533</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#534">534</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#535">535</a>.</li>
 794 </ul></li>
 795 </ul>
 796 </li>
 797 <li>R39:
 798 2005-10-14 post-Mont Tremblant mailing.
 799 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#526">526</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#528">528</a>.
 800 Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#280">280</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#461">461</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#464">464</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#465">465</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#467">467</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#468">468</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#474">474</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#496">496</a> from Ready to WP as per the vote from Mont Tremblant.
 801 Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#247">247</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#294">294</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#342">342</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#362">362</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#369">369</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#371">371</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#376">376</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#384">384</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#475">475</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#478">478</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#495">495</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#497">497</a> from Review to Ready.
 802 Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#498">498</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#504">504</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#506">506</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#509">509</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#510">510</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#511">511</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#512">512</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#513">513</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#514">514</a> from New to Open.
 803 Moved issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#505">505</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#507">507</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#508">508</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#519">519</a> from New to Ready.
 804 Moved issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#500">500</a> from New to NAD.
 805 Moved issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#518">518</a> from New to Review.
 806 </li>
 807 <li>R38:
 808 2005-07-03 pre-Mont Tremblant mailing.
 809 Merged open TR1 issues in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#504">504</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#522">522</a>.
 810 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#523">523</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#523">523</a>
 811 </li>
 812 <li>R37:
 813 2005-06 mid-term mailing.
 814 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#498">498</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#503">503</a>.
 815 </li>
 816 <li>R36:
 817 2005-04 post-Lillehammer mailing. All issues in "ready" status except
 818 for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#454">454</a> were moved to "DR" status, and all issues
 819 previously in "DR" status were moved to "WP".
 820 </li>
 821 <li>R35:
 822 2005-03 pre-Lillehammer mailing.
 823 </li>
 824 <li>R34:
 825 2005-01 mid-term mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#488">488</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#494">494</a>.
 826 </li>
 827 <li>R33:
 828 2004-11 post-Redmond mailing. Reflects actions taken in Redmond.
 829 </li>
 830 <li>R32:
 831 2004-09 pre-Redmond mailing: reflects new proposed resolutions and
 832 new issues received after the 2004-07 mailing.  Added
 833 new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#479">479</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#481">481</a>.
 834 </li>
 835 <li>R31:
 836 2004-07 mid-term mailing: reflects new proposed resolutions and
 837 new issues received after the post-Sydney mailing.  Added
 838 new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#478">478</a>.
 839 </li>
 840 <li>R30:
 841 Post-Sydney mailing: reflects decisions made at the Sydney meeting.
 842 Voted all "Ready" issues from R29 into the working paper.
 843 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#460">460</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#462">462</a>.
 844 </li>
 845 <li>R29:
 846 Pre-Sydney mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#441">441</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#457">457</a>.
 847 </li>
 848 <li>R28:
 849 Post-Kona mailing: reflects decisions made at the Kona meeting.
 850 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#432">432</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#440">440</a>.
 851 </li>
 852 <li>R27:
 853 Pre-Kona mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#404">404</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#431">431</a>.
 854 </li>
 855 <li>R26:
 856 Post-Oxford mailing: reflects decisions made at the Oxford meeting.
 857 All issues in Ready status were voted into DR status.  All issues in
 858 DR status were voted into WP status.
 859 </li>
 860 <li>R25:
 861 Pre-Oxford mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#390">390</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#402">402</a>.
 862 </li>
 863 <li>R24:
 864 Post-Santa Cruz mailing: reflects decisions made at the Santa Cruz
 865 meeting.  All Ready issues from R23 with the exception of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#253">253</a>, which has been given a new proposed resolution, were
 866 moved to DR status.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#383">383</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#389">389</a>.  (Issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#387">387</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#389">389</a> were discussed
 867 at the meeting.)  Made progress on issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#225">225</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#226">226</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#229">229</a>: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#225">225</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#229">229</a> have been moved to Ready status, and the only remaining
 868 concerns with <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#226">226</a> involve wording.
 869 </li>
 870 <li>R23:
 871 Pre-Santa Cruz mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#367">367</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#382">382</a>.
 872 Moved issues in the TC to TC status.
 873 </li>
 874 <li>R22:
 875 Post-Curaçao mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#362">362</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#366">366</a>.
 876 </li>
 877 <li>R21:
 878 Pre-Curaçao mailing.  Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#351">351</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#361">361</a>.
 879 </li>
 880 <li>R20:
 881 Post-Redmond mailing; reflects actions taken in Redmond.  Added
 882 new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#336">336</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#350">350</a>, of which issues
 883 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#347">347</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#350">350</a> were added since Redmond, hence
 884 not discussed at the meeting.
 885
 886 All Ready issues were moved to DR status, with the exception of issues
 887 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#284">284</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#241">241</a>, and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#267">267</a>.
 888
 889 Noteworthy issues discussed at Redmond include
 890 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#120">120</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#202">202</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#226">226</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#233">233</a>,
 891 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#270">270</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#253">253</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#254">254</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#323">323</a>.
 892 </li>
 893 <li>R19:
 894 Pre-Redmond mailing.  Added new issues
 895 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#323">323</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#335">335</a>.
 896 </li>
 897 <li>R18:
 898 Post-Copenhagen mailing; reflects actions taken in Copenhagen.
 899 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#312">312</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#317">317</a>, and discussed
 900 new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#271">271</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#314">314</a>.
 901
 902 Changed status of issues
 903 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#118">118</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#136">136</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#153">153</a>
 904 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#165">165</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#171">171</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#183">183</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#184">184</a>
 905 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#185">185</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#186">186</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#214">214</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#221">221</a>
 906 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#234">234</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#237">237</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#243">243</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#248">248</a>
 907 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#251">251</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#252">252</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#256">256</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#260">260</a>
 908 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#261">261</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#262">262</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#263">263</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#265">265</a>
 909 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#268">268</a>
 910 to DR.
 911
 912 Changed status of issues
 913 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#49">49</a>  <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#109">109</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#117">117</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#182">182</a>
 914 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#228">228</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#230">230</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#232">232</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#235">235</a>
 915 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#238">238</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#241">241</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#242">242</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#250">250</a>
 916 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#259">259</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#264">264</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#266">266</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#267">267</a>
 917 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#271">271</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#272">272</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#273">273</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#275">275</a>
 918 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#281">281</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#284">284</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#285">285</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#286">286</a>
 919 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#288">288</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#292">292</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#295">295</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#297">297</a>
 920 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#298">298</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#301">301</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#303">303</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#306">306</a>
 921 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#307">307</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#308">308</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#312">312</a>
 922 to Ready.
 923
 924 Closed issues
 925 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#111">111</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#277">277</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#279">279</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#287">287</a>
 926 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#289">289</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#293">293</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#302">302</a> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#313">313</a>
 927 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#314">314</a>
 928 as NAD.
 929
 930 </li>
 931 <li>R17:
 932 Pre-Copenhagen mailing.  Converted issues list to XML.  Added proposed
 933 resolutions for issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#49">49</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#76">76</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#91">91</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#235">235</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#250">250</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#267">267</a>.
 934 Added new issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#278">278</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#311">311</a>.
 935 </li>
 936 <li>R16:
 937 post-Toronto mailing; reflects actions taken in Toronto. Added new
 938 issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#265">265</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#277">277</a>.  Changed status of issues
 939 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#3">3</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#8">8</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#9">9</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#19">19</a>,
 940 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#26">26</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#31">31</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#61">61</a>,
 941 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#63">63</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#86">86</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#108">108</a>,
 942 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#112">112</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#114">114</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#115">115</a>,
 943 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#122">122</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#127">127</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#129">129</a>,
 944 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#134">134</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#137">137</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#142">142</a>,
 945 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#144">144</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#146">146</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#147">147</a>,
 946 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#159">159</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#164">164</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#170">170</a>,
 947 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#181">181</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#199">199</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#208">208</a>,
 948 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#209">209</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#210">210</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#211">211</a>,
 949 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#212">212</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#217">217</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#220">220</a>,
 950 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#222">222</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#223">223</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#224">224</a>,
 951 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#227">227</a> to "DR".  Reopened issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#23">23</a>. Reopened
 952 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#187">187</a>. Changed issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#2">2</a> and
 953 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#4">4</a> to NAD. Fixed a typo in issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#17">17</a>. Fixed
 954 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#70">70</a>: signature should be changed both places it
 955 appears. Fixed issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#160">160</a>: previous version didn't fix
 956 the bug in enough places.
 957 </li>
 958 <li>R15:
 959 pre-Toronto mailing. Added issues
 960 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#233">233</a>-<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#264">264</a>. Some small HTML formatting
 961 changes so that we pass Weblint tests.
 962 </li>
 963 <li>R14:
 964 post-Tokyo II mailing; reflects committee actions taken in
 965 Tokyo. Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#228">228</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#232">232</a>. (00-0019R1/N1242)
 966 </li>
 967 <li>R13:
 968 pre-Tokyo II updated: Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#212">212</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#227">227</a>.
 969 </li>
 970 <li>R12:
 971 pre-Tokyo II mailing: Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#199">199</a> to
 972 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#211">211</a>. Added "and paragraph 5" to the proposed resolution
 973 of issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#29">29</a>.  Add further rationale to issue
 974 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#178">178</a>.
 975 </li>
 976 <li>R11:
 977 post-Kona mailing: Updated to reflect LWG and full committee actions
 978 in Kona (99-0048/N1224). Note changed resolution of issues
 979 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#4">4</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#38">38</a>. Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#196">196</a>
 980 to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#198">198</a>. Closed issues list split into "defects" and
 981 "closed" documents.  Changed the proposed resolution of issue
 982 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#4">4</a> to NAD, and changed the wording of proposed resolution
 983 of issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#38">38</a>.
 984 </li>
 985 <li>R10:
 986 pre-Kona updated.  Added proposed resolutions <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#83">83</a>,
 987 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#86">86</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#91">91</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#92">92</a>,
 988 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#109">109</a>. Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#190">190</a> to
 989 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#195">195</a>. (99-0033/D1209, 14 Oct 99)
 990 </li>
 991 <li>R9:
 992 pre-Kona mailing.  Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#140">140</a> to
 993 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#189">189</a>. Issues list split into separate "active" and
 994 "closed" documents. (99-0030/N1206, 25 Aug 99)
 995 </li>
 996 <li>R8:
 997 post-Dublin mailing. Updated to reflect LWG and full committee actions
 998 in Dublin. (99-0016/N1193, 21 Apr 99)
 999 </li>
1000 <li>R7:
1001 pre-Dublin updated: Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#130">130</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#131">131</a>,
1002 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#132">132</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#133">133</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#134">134</a>,
1003 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#135">135</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#136">136</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#137">137</a>,
1004 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#138">138</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#139">139</a> (31 Mar 99)
1005 </li>
1006 <li>R6:
1007 pre-Dublin mailing. Added issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#127">127</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#128">128</a>,
1008 and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#129">129</a>.  (99-0007/N1194, 22 Feb 99)
1009 </li>
1010 <li>R5:
1011 update issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#112">112</a>; added issues
1012 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#114">114</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#126">126</a>. Format revisions to prepare
1013 for making list public. (30 Dec 98)
1014 </li>
1015 <li>R4:
1016 post-Santa Cruz II updated: Issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#110">110</a>,
1017 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#111">111</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#112">112</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#113">113</a> added, several
1018 issues corrected. (22 Oct 98)
1019 </li>
1020 <li>R3:
1021 post-Santa Cruz II: Issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#94">94</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#109">109</a>
1022 added, many issues updated to reflect LWG consensus (12 Oct 98)
1023 </li>
1024 <li>R2:
1025 pre-Santa Cruz II: Issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#73">73</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#93">93</a> added,
1026 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#17">17</a> updated. (29 Sep 98)
1027 </li>
1028 <li>R1:
1029 Correction to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#55">55</a> resolution, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#60">60</a> code
1030 format, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#64">64</a> title. (17 Sep 98)
1031 </li>
1032 </ul>
1033
1034 <h2><a name="Status"></a>Issue Status</h2>
1035
1036   <p><b><a name="New">New</a></b> - The issue has not yet been
1037   reviewed by the LWG. Any <b>Proposed Resolution</b> is purely a
1038   suggestion from the issue submitter, and should not be construed as
1039   the view of LWG.</p>
1040
1041   <p><b><a name="Open">Open</a></b> - The LWG has discussed the issue
1042   but is not yet ready to move the issue forward. There are several
1043   possible reasons for open status:</p>
1044      <ul>
1045         <li>Consensus may have not yet have been reached as to how to deal
1046             with the issue.</li>
1047         <li>Informal consensus may have been reached, but the LWG awaits
1048             exact <b>Proposed Resolution</b> wording for review.</li>
1049         <li>The LWG wishes to consult additional technical experts before
1050             proceeding.</li>
1051         <li>The issue may require further study.</li>
1052      </ul>
1053
1054   <p>A <b>Proposed Resolution</b> for an open issue is still not be
1055   construed as the view of LWG. Comments on the current state of
1056   discussions are often given at the end of open issues in an italic
1057   font. Such comments are for information only and should not be given
1058   undue importance.</p>
1059
1060   <p><b><a name="Dup">Dup</a></b> - The LWG has reached consensus that
1061   the issue is a duplicate of another issue, and will not be further
1062   dealt with. A <b>Rationale</b> identifies the duplicated issue's
1063   issue number.  </p>
1064
1065   <p><b><a name="NAD">NAD</a></b> - The LWG has reached consensus that
1066   the issue is not a defect in the Standard.</p>
1067
1068   <p><b><a name="NAD Editorial">NAD Editorial</a></b> - The LWG has reached consensus that
1069   the issue can either be handled editorially, or is handled by a paper (usually
1070   linked to in the rationale).</p>
1071
1072   <p><b><a name="NAD Concepts">NAD Concepts</a></b> - The LWG has reached consensus that
1073   the issue is NAD for now, but represents a real issue when the library is
1074   done with language-supported concepts.</p>
1075
1076   <p><b><a name="NAD Future">NAD Future</a></b> - In addition to the regular
1077   status, the LWG believes that this issue should be revisited at the
1078   next revision of the standard.</p>
1079
1080   <p><b><a name="Review">Review</a></b> - Exact wording of a
1081   <b>Proposed Resolution</b> is now available for review on an issue
1082   for which the LWG previously reached informal consensus.</p>
1083
1084   <p><b><a name="Ready">Ready</a></b> - The LWG has reached consensus
1085   that the issue is a defect in the Standard, the <b>Proposed
1086   Resolution</b> is correct, and the issue is ready to forward to the
1087   full committee for further action as a Defect Report (DR).</p>
1088
1089   <p><b><a name="DR">DR</a></b> - (Defect Report) - The full J16
1090   committee has voted to forward the issue to the Project Editor to be
1091   processed as a Potential Defect Report. The Project Editor reviews
1092   the issue, and then forwards it to the WG21 Convenor, who returns it
1093   to the full committee for final disposition. This issues list
1094   accords the status of DR to all these Defect Reports regardless of
1095   where they are in that process.</p>
1096
1097   <p><b><a name="TC1">TC1</a></b> - (Technical Corrigenda 1) - The full
1098   WG21 committee has voted to accept the Defect Report's Proposed
1099   Resolution as a Technical Corrigenda.  Action on this issue is thus
1100   complete and no further action is possible under ISO rules.</p>
1101
1102   <p><b><a name="CD1">CD1</a></b> - (Committee Draft 2008) - The full
1103   WG21 committee has voted to accept the Defect Report's Proposed
1104   Resolution into the Fall 2008 Committee Draft.</p>
1105
1106   <p><b><a name="TRDec">TRDec</a></b> - (Decimal TR defect) - The
1107   LWG has voted to accept the Defect Report's Proposed
1108   Resolution into the Decimal TR.  Action on this issue is thus
1109   complete and no further action is expected.</p>
1110
1111   <p><b><a name="WP">WP</a></b> - (Working Paper) - The proposed
1112   resolution has not been accepted as a Technical Corrigendum, but
1113   the full WG21 committee has voted to apply the Defect Report's Proposed
1114   Resolution to the working paper.</p>
1115
1116   <p><b>Tentatively</b> - This is a <i>status qualifier</i>.  The issue has
1117   been reviewed online, or at an unofficial meeting, but not in an official meeting, and some support has been formed
1118   for the qualified status.  Tentatively qualified issues may be moved to the unqualified status
1119   and forwarded to full committee (if Ready) within the same meeting.  Unlike Ready issues, Tentatively Ready issues
1120   will be reviewed in subcommittee prior to forwarding to full committee.  When a status is
1121   qualified with Tentatively, the issue is still considered active.</p>
1122
1123   <p><b>Pending</b> - This is a <i>status qualifier</i>.  When prepended to
1124   a status this indicates the issue has been
1125   processed by the committee, and a decision has been made to move the issue to
1126   the associated unqualified status.  However for logistical reasons the indicated
1127   outcome of the issue has not yet appeared in the latest working paper.
1128
1129   </p><p>Issues are always given the status of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a> when
1130   they first appear on the issues list. They may progress to
1131   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a> or <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a> while the LWG
1132   is actively working on them. When the LWG has reached consensus on
1133   the disposition of an issue, the status will then change to
1134   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Dup">Dup</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD">NAD</a>, or
1135   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a> as appropriate.  Once the full J16 committee votes to
1136   forward Ready issues to the Project Editor, they are given the
1137   status of Defect Report ( <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#DR">DR</a>). These in turn may
1138   become the basis for Technical Corrigenda (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#TC">TC</a>),
1139   or are closed without action other than a Record of Response
1140   (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#RR">RR</a> ). The intent of this LWG process is that
1141   only issues which are truly defects in the Standard move to the
1142   formal ISO DR status.
1143   </p>
1144
1145
1146 <h2>Active Issues</h2>
1147 <hr>
1148 <h3><a name="296"></a>296. Missing descriptions and requirements of pair operators</h3>
1149 <p><b>Section:</b> 20.3.4 [pairs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
1150  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2001-01-14  <b>Last modified:</b> 2009-10-26</p>
1151 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#pairs">active issues</a> in [pairs].</p>
1152 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#pairs">issues</a> in [pairs].</p>
1153 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
1154 <p><b>Discussion:</b></p>
1155 <p>The synopsis of the header <tt>&lt;utility&gt;</tt> in 20.3 [utility]
1156 lists the complete set of equality and relational operators for <tt>pair</tt>
1157 but the section describing the template and the operators only describes
1158 <tt>operator==()</tt> and <tt>operator&lt;()</tt>, and it fails to mention
1159 any requirements on the template arguments. The remaining operators are
1160 not mentioned at all.
1161 </p>
1162
1163 <p><i>[
1164 2009-09-27 Alisdair reopens.
1165 ]</i></p>
1166
1167
1168 <blockquote>
1169 <p>
1170 The issue is a lack of wording specifying the semantics of <tt>std::pair</tt>
1171 relational operators.  The rationale is that this is covered by
1172 catch-all wording in the relops component, and that as relops directly
1173 precedes <tt>pair</tt> in the document this is an easy connection to make.
1174 </p>
1175
1176 <p>
1177 Reading the current working paper I make two observations:
1178 </p>
1179
1180 <ol type="i">
1181 <li>
1182 relops no longer immediately precedes <tt>pair</tt> in the order of
1183 specification.  However, even if it did, there is a lot of <tt>pair</tt>
1184 specification itself between the (apparently) unrelated relops and the
1185 relational operators for <tt>pair</tt>.  (The catch-all still requires
1186 <tt>operator==</tt> and <tt>operator&lt;</tt> to be specified
1187 explicitly)
1188 </li>
1189
1190 <li>
1191 No other library component relies on the catch-all clause. The following
1192 all explicitly document all six relational operators, usually in a
1193 manner that could have deferred to the relops clause.
1194 </li>
1195 </ol>
1196
1197 <blockquote><pre>tuple
1198 unique_ptr
1199 duration
1200 time_point
1201 basic_string
1202 queue
1203 stack
1204 move_iterator
1205 reverse_iterator
1206 regex submatch
1207 thread::id
1208 </pre></blockquote>
1209
1210 <p>
1211 The container components provide their own (equivalent) definition in
1212 23.2.1 [container.requirements.general] Table 90 -- Container
1213 requirements and do so do not defer to relops.
1214 </p>
1215
1216 <p>
1217 <tt>Shared_ptr</tt> explicitly documents <tt>operator!=</tt> and does
1218 not supply the other 3 missing operators
1219 (<tt>&gt;</tt>,<tt>&gt;=</tt>,<tt>&lt;=</tt>) so does not meet the
1220 reqirements of the relops clause.
1221 </p>
1222
1223 <p>
1224 <tt>Weak_ptr</tt> only supports <tt>operator&lt;</tt> so would not be
1225 covered by relops.
1226 </p>
1227
1228 <p>
1229 At the very least I would request a note pointing to the relops clause
1230 we rely on to provide this definition.  If this route is taken, I would
1231 recommend reducing many of the above listed clauses to a similar note
1232 rather than providing redundant specification.
1233 </p>
1234
1235 <p>
1236 My preference would be to supply the 4 missing specifications consistent
1237 with the rest of the library.
1238 </p>
1239
1240 </blockquote>
1241
1242 <p><i>[
1243 2009-10-11 Daniel opens <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1233">1233</a> which deals with the same issue as
1244 it pertains to <tt>unique_ptr</tt>.
1245 ]</i></p>
1246
1247
1248 <p><i>[
1249 2009-10 Santa Cruz:
1250 ]</i></p>
1251
1252
1253 <blockquote>
1254 Move to Ready
1255 </blockquote>
1256
1257
1258
1259 <p><b>Proposed resolution:</b></p>
1260 <p>
1261 After p20 20.3.4 [pairs] add:
1262 </p>
1263
1264 <blockquote><pre>template &lt;class T1, class T2&gt;
1265 bool operator!=(const pair&lt;T1,T2&gt;&amp; x, const pair&lt;T1,T2&gt;&amp; y);
1266 </pre>
1267
1268 <blockquote>
1269 <i>Returns:</i> <tt>!(x==y)</tt>
1270 </blockquote>
1271
1272 <pre>template &lt;class T1, class T2&gt;
1273 bool operator&gt; (const pair&lt;T1,T2&gt;&amp; x, const pair&lt;T1,T2&gt;&amp; y);
1274 </pre>
1275
1276 <blockquote>
1277 <i>Returns:</i> <tt>y &lt; x</tt>
1278 </blockquote>
1279
1280 <pre>template &lt;class T1, class T2&gt;
1281 bool operator&gt;=(const pair&lt;T1,T2&gt;&amp; x, const pair&lt;T1,T2&gt;&amp; y);
1282 </pre>
1283
1284 <blockquote>
1285 <i>Returns:</i> <tt>!(x &lt; y)</tt>
1286 </blockquote>
1287
1288 <pre>template &lt;class T1, class T2&gt;
1289 bool operator&lt;=(const pair&lt;T1,T2&gt;&amp; x, const pair&lt;T1,T2&gt;&amp; y);
1290 </pre>
1291
1292 <blockquote>
1293 <i>Returns:</i> <tt>!(y &lt; x)</tt>
1294 </blockquote>
1295 </blockquote>
1296
1297
1298 <p><b>Rationale:</b></p>
1299 <p>20.3.1 [operators] paragraph 10 already specifies the semantics.
1300 That paragraph says that, if declarations of operator!=, operator&gt;,
1301 operator&lt;=, and operator&gt;= appear without definitions, they are
1302 defined as specified in 20.3.1 [operators].  There should be no user
1303 confusion, since that paragraph happens to immediately precede the
1304 specification of <tt>pair</tt>.</p>
1305
1306
1307
1308
1309
1310 <hr>
1311 <h3><a name="299"></a>299. Incorrect return types for iterator dereference</h3>
1312 <p><b>Section:</b> 24.2.4 [bidirectional.iterators], 24.2.5 [random.access.iterators] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
1313  <b>Submitter:</b> John Potter <b>Opened:</b> 2001-01-22  <b>Last modified:</b> 2009-10-26</p>
1314 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#bidirectional.iterators">issues</a> in [bidirectional.iterators].</p>
1315 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
1316 <p><b>Discussion:</b></p>
1317 <p>
1318 In section 24.2.4 [bidirectional.iterators],
1319 Table 75 gives the return type of *r-- as convertible to T.  This is
1320 not consistent with Table 74 which gives the return type of *r++ as
1321 T&amp;.  *r++ = t is valid while *r-- = t is invalid.
1322 </p>
1323
1324 <p>
1325 In section 24.2.5 [random.access.iterators],
1326 Table 76 gives the return type of a[n] as convertible to T.  This is
1327 not consistent with the semantics of *(a + n) which returns T&amp; by
1328 Table 74.  *(a + n) = t is valid while a[n] = t is invalid.
1329 </p>
1330
1331 <p>
1332 Discussion from the Copenhagen meeting: the first part is
1333 uncontroversial.  The second part, operator[] for Random Access
1334 Iterators, requires more thought.  There are reasonable arguments on
1335 both sides.  Return by value from operator[] enables some potentially
1336 useful iterators, e.g. a random access "iota iterator" (a.k.a
1337 "counting iterator" or "int iterator").  There isn't any obvious way
1338 to do this with return-by-reference, since the reference would be to a
1339 temporary.  On the other hand, <tt>reverse_iterator</tt> takes an
1340 arbitrary Random Access Iterator as template argument, and its
1341 operator[] returns by reference.  If we decided that the return type
1342 in Table 76 was correct, we would have to change
1343 <tt>reverse_iterator</tt>.  This change would probably affect user
1344 code.
1345 </p>
1346
1347 <p>
1348 History: the contradiction between <tt>reverse_iterator</tt> and the
1349 Random Access Iterator requirements has been present from an early
1350 stage.  In both the STL proposal adopted by the committee
1351 (N0527==94-0140) and the STL technical report (HPL-95-11 (R.1), by
1352 Stepanov and Lee), the Random Access Iterator requirements say that
1353 operator[]'s return value is "convertible to T".  In N0527
1354 reverse_iterator's operator[] returns by value, but in HPL-95-11
1355 (R.1), and in the STL implementation that HP released to the public,
1356 reverse_iterator's operator[] returns by reference.  In 1995, the
1357 standard was amended to reflect the contents of HPL-95-11 (R.1).  The
1358 original intent for operator[] is unclear.
1359 </p>
1360
1361 <p>
1362 In the long term it may be desirable to add more fine-grained
1363 iterator requirements, so that access method and traversal strategy
1364 can be decoupled.  (See "Improved Iterator Categories and
1365 Requirements", N1297 = 01-0011, by Jeremy Siek.)  Any decisions
1366 about issue 299 should keep this possibility in mind.
1367 </p>
1368
1369 <p>Further discussion: I propose a compromise between John Potter's
1370 resolution, which requires <tt>T&amp;</tt> as the return type of
1371 <tt>a[n]</tt>, and the current wording, which requires convertible to
1372 <tt>T</tt>. The compromise is to keep the convertible to <tt>T</tt>
1373 for the return type of the expression <tt>a[n]</tt>, but to also add
1374 <tt>a[n] = t</tt> as a valid expression. This compromise "saves" the
1375 common case uses of random access iterators, while at the same time
1376 allowing iterators such as counting iterator and caching file
1377 iterators to remain random access iterators (iterators where the
1378 lifetime of the object returned by <tt>operator*()</tt> is tied to the
1379 lifetime of the iterator).
1380 </p>
1381
1382 <p>
1383 Note that the compromise resolution necessitates a change to
1384 <tt>reverse_iterator</tt>. It would need to use a proxy to support
1385 <tt>a[n] = t</tt>.
1386 </p>
1387
1388 <p>
1389 Note also there is one kind of mutable random access iterator that
1390 will no longer meet the new requirements. Currently, iterators that
1391 return an r-value from <tt>operator[]</tt> meet the requirements for a
1392 mutable random access iterartor, even though the expression <tt>a[n] =
1393 t</tt> will only modify a temporary that goes away. With this proposed
1394 resolution, <tt>a[n] = t</tt> will be required to have the same
1395 operational semantics as <tt>*(a + n) = t</tt>.
1396 </p>
1397
1398 <p><i>[
1399 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
1400 ]</i></p>
1401
1402
1403 <p><i>[
1404 2009-09-18 Alisdair adds:
1405 ]</i></p>
1406
1407
1408 <blockquote>
1409 <p>
1410 Why can't we write through the reference returned from operator[] on a
1411 random access iterator?
1412 </p>
1413
1414 <p>
1415 Recommended solution:
1416 </p>
1417
1418 <p>
1419 In table Table 104 -- Random access iterator requirements, replace
1420 </p>
1421
1422 <blockquote>
1423 a[n] : convertible to <del><tt>const T &amp;</tt></del>
1424 <ins><tt>T&amp;</tt> if <tt>X</tt> is mutable, otherwise convertible to <tt>const T&amp;</tt></ins>
1425 </blockquote>
1426 </blockquote>
1427
1428 <p><i>[
1429 2009-10 Santa Cruz:
1430 ]</i></p>
1431
1432
1433 <blockquote>
1434 Leave Open. Alisdair to spearhead a paper on revivification.
1435 </blockquote>
1436
1437
1438
1439 <p><b>Proposed resolution:</b></p>
1440
1441 <p>
1442 In section 24.1.4 [lib.bidirectdional.iterators], change the return
1443 type in table 75 from "convertible to <tt>T</tt>" to
1444 <tt>T&amp;</tt>.
1445 </p>
1446
1447 <p>
1448 In section 24.1.5 [lib.random.access.iterators], change the
1449 operational semantics for <tt>a[n]</tt> to " the r-value of
1450 <tt>a[n]</tt> is equivalent to the r-value of <tt>*(a +
1451 n)</tt>". Add a new row in the table for the expression <tt>a[n] = t</tt>
1452 with a return type of convertible to <tt>T</tt> and operational semantics of
1453 <tt>*(a + n) = t</tt>.
1454 </p>
1455
1456 <p><i>[Lillehammer: Real problem, but should be addressed as part of
1457   iterator redesign]</i></p>
1458
1459
1460
1461
1462 <p><b>Rationale:</b></p>
1463 <p><i>[
1464 San Francisco:
1465 ]</i></p>
1466
1467
1468 <blockquote>
1469 Solved by
1470 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2758.pdf">N2758</a>.
1471 </blockquote>
1472
1473
1474
1475
1476
1477
1478
1479 <hr>
1480 <h3><a name="397"></a>397. ostream::sentry dtor throws exceptions</h3>
1481 <p><b>Section:</b> 27.7.2.4 [ostream::sentry] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
1482  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2003-01-05  <b>Last modified:</b> 2009-10-20</p>
1483 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#ostream::sentry">issues</a> in [ostream::sentry].</p>
1484 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
1485 <p><b>Discussion:</b></p>
1486     <p>
1487 17.4.4.8, p3 prohibits library dtors from throwing exceptions.
1488     </p>
1489     <p>
1490 27.6.2.3, p4 says this about the ostream::sentry dtor:
1491     </p>
1492     <pre>    -4- If ((os.flags() &amp; ios_base::unitbuf) &amp;&amp; !uncaught_exception())
1493         is true, calls os.flush().
1494     </pre>
1495     <p>
1496 27.6.2.6, p7 that describes ostream::flush() says:
1497     </p>
1498     <pre>    -7- If rdbuf() is not a null pointer, calls rdbuf()-&gt;pubsync().
1499         If that function returns ?-1 calls setstate(badbit) (which
1500         may throw ios_base::failure (27.4.4.3)).
1501     </pre>
1502     <p>
1503 That seems like a defect, since both pubsync() and setstate() can
1504 throw an exception.
1505     </p>
1506 <p><i>[
1507 The contradiction is real.  Clause 17 says destructors may never
1508 throw exceptions, and clause 27 specifies a destructor that does
1509 throw.  In principle we might change either one.  We're leaning
1510 toward changing clause 17: putting in an "unless otherwise specified"
1511 clause, and then putting in a footnote saying the sentry destructor
1512 is the only one that can throw.  PJP suggests specifying that
1513 sentry::~sentry() should internally catch any exceptions it might cause.
1514 ]</i></p>
1515
1516
1517 <p><i>[
1518 See <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#418">418</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#622">622</a> for related issues.
1519 ]</i></p>
1520
1521
1522 <p><i>[
1523 2009-07 Frankfurt
1524 ]</i></p>
1525
1526
1527 <blockquote>
1528 <p>
1529 Move to Review. Add "Throws: nothing" to the specification of ostream::sentry::~sentry().
1530 </p>
1531 </blockquote>
1532
1533 <p><i>[
1534 2009-10-13 Daniel adds:
1535 ]</i></p>
1536
1537
1538 <blockquote>
1539 The proposed resolution of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#835">835</a> is written to match the outcome
1540 of this issue.
1541 </blockquote>
1542
1543 <p><i>[
1544 2009 Santa Cruz:
1545 ]</i></p>
1546
1547
1548 <blockquote>
1549 Move to Open.  Our intent is to solve this issue with <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#835">835</a>.
1550 </blockquote>
1551
1552
1553
1554 <p><b>Proposed resolution:</b></p>
1555 <p>
1556 Add after 27.7.2.4 [ostream::sentry] p17:
1557 </p>
1558
1559 <blockquote>
1560 <pre>~sentry();
1561 </pre>
1562 <blockquote>
1563 <p>
1564 -17- If <tt>((os.flags() &amp; ios_base::unitbuf) &amp;&amp; !uncaught_exception())</tt>
1565 is <tt>true</tt>, calls <tt>os.flush()</tt>.
1566 </p>
1567
1568 <p><ins>
1569 <i>Throws:</i> Nothing.
1570 </ins></p>
1571 </blockquote>
1572 </blockquote>
1573
1574
1575
1576
1577
1578 <hr>
1579 <h3><a name="408"></a>408. Is vector&lt;reverse_iterator&lt;char*&gt; &gt; forbidden?</h3>
1580 <p><b>Section:</b> 24.2 [iterator.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
1581  <b>Submitter:</b> Nathan Myers <b>Opened:</b> 2003-06-03  <b>Last modified:</b> 2009-11-03</p>
1582 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
1583 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
1584 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
1585 <p><b>Discussion:</b></p>
1586 <p>
1587 I've been discussing iterator semantics with Dave Abrahams, and a
1588 surprise has popped up.  I don't think this has been discussed before.
1589 </p>
1590
1591 <p>
1592 X [iterator.concepts] says that the only operation that can be performed on "singular"
1593 iterator values is to assign a non-singular value to them.  (It
1594 doesn't say they can be destroyed, and that's probably a defect.)
1595 Some implementations have taken this to imply that there is no need
1596 to initialize the data member of a reverse_iterator&lt;&gt; in the default
1597 constructor.  As a result, code like
1598 </p>
1599 <blockquote><pre>  std::vector&lt;std::reverse_iterator&lt;char*&gt; &gt; v(7);
1600   v.reserve(1000);
1601 </pre></blockquote>
1602 <p>
1603 invokes undefined behavior, because it must default-initialize the
1604 vector elements, and then copy them to other storage.  Of course many
1605 other vector operations on these adapters are also left undefined,
1606 and which those are is not reliably deducible from the standard.
1607 </p>
1608
1609 <p>
1610 I don't think that 24.1 was meant to make standard-library iterator
1611 types unsafe.  Rather, it was meant to restrict what operations may
1612 be performed by functions which take general user- and standard
1613 iterators as arguments, so that raw pointers would qualify as
1614 iterators.  However, this is not clear in the text, others have come
1615 to the opposite conclusion.
1616 </p>
1617
1618 <p>
1619 One question is whether the standard iterator adaptors have defined
1620 copy semantics.  Another is whether they have defined destructor
1621 semantics: is
1622 </p>
1623 <blockquote><pre>  { std::vector&lt;std::reverse_iterator&lt;char*&gt; &gt;  v(7); }
1624 </pre></blockquote>
1625 <p>
1626 undefined too?
1627 </p>
1628
1629 <p>
1630 Note this is not a question of whether algorithms are allowed to
1631 rely on copy semantics for arbitrary iterators, just whether the
1632 types we actually supply support those operations.  I believe the
1633 resolution must be expressed in terms of the semantics of the
1634 adapter's argument type.  It should make clear that, e.g., the
1635 reverse_iterator&lt;T&gt; constructor is actually required to execute
1636 T(), and so copying is defined if the result of T() is copyable.
1637 </p>
1638
1639 <p>
1640 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#235">235</a>, which defines reverse_iterator's default
1641 constructor more precisely, has some relevance to this issue.
1642 However, it is not the whole story.
1643 </p>
1644
1645 <p>
1646 The issue was whether
1647 </p>
1648 <blockquote><pre>  reverse_iterator() { }
1649 </pre></blockquote>
1650 <p>
1651 is allowed, vs.
1652 </p>
1653 <blockquote><pre>  reverse_iterator() : current() { }
1654 </pre></blockquote>
1655
1656 <p>
1657 The difference is when T is char*, where the first leaves the member
1658 uninitialized, and possibly equal to an existing pointer value, or
1659 (on some targets) may result in a hardware trap when copied.
1660 </p>
1661
1662 <p>
1663 8.5 paragraph 5 seems to make clear that the second is required to
1664 satisfy DR <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#235">235</a>, at least for non-class Iterator argument
1665 types.
1666 </p>
1667
1668 <p>
1669 But that only takes care of reverse_iterator, and doesn't establish
1670 a policy for all iterators.  (The reverse iterator adapter was just
1671 an example.)  In particular, does my function
1672 </p>
1673 <blockquote><pre>  template &lt;typename Iterator&gt;
1674     void f() { std::vector&lt;Iterator&gt;  v(7); }
1675 </pre></blockquote>
1676 <p>
1677 evoke undefined behavior for some conforming iterator definitions?
1678 I think it does, now, because vector&lt;&gt; will destroy those singular
1679 iterator values, and that's explicitly disallowed.
1680 </p>
1681
1682 <p>
1683 24.1 shouldn't give blanket permission to copy all singular iterators,
1684 because then pointers wouldn't qualify as iterators.  However, it
1685 should allow copying of that subset of singular iterator values that
1686 are default-initialized, and it should explicitly allow destroying any
1687 iterator value, singular or not, default-initialized or not.
1688 </p>
1689
1690 <p>Related issues: <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#407">407</a>, <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a></p>
1691 <p><i>[
1692 We don't want to require all singular iterators to be copyable,
1693 because that is not the case for pointers.  However, default
1694 construction may be a special case.  Issue: is it really default
1695 construction we want to talk about, or is it something like value
1696 initialization?  We need to check with core to see whether default
1697 constructed pointers are required to be copyable; if not, it would be
1698 wrong to impose so strict a requirement for iterators.
1699 ]</i></p>
1700
1701
1702 <p><i>[
1703 2009-05-10 Alisdair provided wording.
1704 ]</i></p>
1705
1706
1707 <blockquote>
1708 The comments regarding destroying singular iterators have already been
1709 resolved.  That just leaves copying (with moving implied).
1710 </blockquote>
1711
1712 <p><i>[
1713 2009-07 Frankfurt
1714 ]</i></p>
1715
1716
1717 <blockquote>
1718 <p>
1719 This is related to LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>.
1720 </p>
1721 <p>
1722 Note that there is a bug in the proposed resolution to LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>. The
1723 change to  [reverse.iter.con] should be modified so that the word
1724 "default" in the second sentence of the Effects clause is replaced by
1725 "value."
1726 </p>
1727 <p>
1728 We believe that the proposed fix to LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a> (now corrected) is
1729 sufficient to solve the problem for reverse_iterator. However, Alisdair
1730 pointed out that LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a> does not solve the general problem for authors
1731 of iterator adaptors.
1732 </p>
1733 <p>
1734 There are some problems with the proposed resolution. The phrase "safely
1735 copyable" is not a term of art. Also, it mentions a
1736 DefaultConstructible? concept.
1737 </p>
1738 <p>
1739 Move to Review after Alisdair updates the wording.
1740 </p>
1741 </blockquote>
1742
1743 <p><i>[
1744 2009-07-31 Alisdair revised wording:
1745 ]</i></p>
1746
1747
1748 <p><i>[
1749 2009-08-17 Alisdair and Daniel collaborate on slightly revised wording.
1750 This issue depends upon <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>
1751 ]</i></p>
1752
1753
1754 <p><i>[
1755 2009-10-14 Daniel adds:
1756 ]</i></p>
1757
1758
1759 <blockquote>
1760 There is a clear dependency on <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1213">1213</a>, because the term "singular",
1761 which is used as part of the resolution, is not properly defined yet.
1762 </blockquote>
1763
1764 <p><i>[
1765 2009-10 Santa Cruz:
1766 ]</i></p>
1767
1768
1769 <blockquote>
1770 Moved to Open. Alisdair will provide improved wording to make
1771 this have "value semantics" and otherwise behave like a valid iterator.
1772 </blockquote>
1773
1774
1775
1776 <p><b>Proposed resolution:</b></p>
1777 <p>
1778 Add a new paragrpah to Iterator concepts 24.2 [iterator.requirements] after para 5 (the one describing
1779 singular iterators)
1780 </p>
1781 <blockquote>
1782 <p>
1783 Just as a regular pointer to an array guarantees that there is a pointer
1784 value pointing past the last element of the array, so for any iterator
1785 type there is an iterator value that points past the last element of a
1786 corresponding container. These values are called <i>past-the-end</i> values.
1787 Values of an iterator <tt>i</tt> for which the expression <tt>*i</tt> is defined are called
1788 <i>dereferenceable</i>. The library never assumes that past-the-end values are
1789 dereferenceable. Iterators can also have singular values that are not
1790 associated with any container. [<i>Example:</i> After the declaration of an
1791 uninitialized pointer <tt>x</tt> (as with <tt>int* x;</tt>), <tt>x</tt> must always be assumed to
1792 have a singular value of a pointer. &#8212; <i>end example</i>] Results of most
1793 expressions are undefined for singular values; the only exceptions are
1794 destroying an iterator that holds a singular value and the assignment of
1795 a non-singular value to an iterator that holds a singular value. In this
1796 case the singular value is overwritten the same way as any other value.
1797 Dereferenceable values are always non-singular.
1798 </p>
1799 <p><ins>
1800 After value-initialization, any iterator that satisfies the
1801 <tt>DefaultConstructible</tt> requirements ([defaultconstructible]) shall not introduce undefined behaviour
1802 when used <ins>as</ins> the
1803 source of a copy or move operation, even if it would
1804 otherwise be singular. [<i>Note:</i> This guarantee is not offered for
1805 default-initialization (8.5 [dcl.init]), although the distinction only
1806 matters for types with trivial default constructors such as pointers. &#8212;
1807 <i>end note</i>]
1808 </ins></p>
1809
1810
1811 </blockquote>
1812
1813
1814
1815
1816
1817
1818 <hr>
1819 <h3><a name="427"></a>427. stage 2 and rationale of DR 221</h3>
1820 <p><b>Section:</b> 22.4.2.1.2 [facet.num.get.virtuals] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
1821  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2003-09-18  <b>Last modified:</b> 2009-10-26</p>
1822 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#facet.num.get.virtuals">active issues</a> in [facet.num.get.virtuals].</p>
1823 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#facet.num.get.virtuals">issues</a> in [facet.num.get.virtuals].</p>
1824 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
1825 <p><b>Discussion:</b></p>
1826 <p>
1827 The requirements specified in Stage 2 and reiterated in the rationale
1828 of DR 221 (and echoed again in DR 303) specify that num_get&lt;charT&gt;::
1829 do_get() compares characters on the stream against the widened elements
1830 of "012...abc...ABCX+-"
1831 </p>
1832
1833 <p>
1834 An implementation is required to allow programs to instantiate the num_get
1835 template on any charT that satisfies the requirements on a user-defined
1836 character type. These requirements do not include the ability of the
1837 character type to be equality comparable (the char_traits template must
1838 be used to perform tests for equality). Hence, the num_get template cannot
1839 be implemented to support any arbitrary character type. The num_get template
1840 must either make the assumption that the character type is equality-comparable
1841 (as some popular implementations do), or it may use char_traits&lt;charT&gt; to do
1842 the comparisons (some other popular implementations do that). This diversity
1843 of approaches makes it difficult to write portable programs that attempt to
1844 instantiate the num_get template on user-defined types.
1845 </p>
1846
1847 <p><i>[Kona: the heart of the problem is that we're theoretically
1848   supposed to use traits classes for all fundamental character
1849   operations like assignment and comparison, but facets don't have
1850   traits parameters.  This is a fundamental design flaw and it
1851   appears all over the place, not just in this one place.  It's not
1852   clear what the correct solution is, but a thorough review of facets
1853   and traits is in order.  The LWG considered and rejected the
1854   possibility of changing numeric facets to use narrowing instead of
1855   widening.  This may be a good idea for other reasons (see issue
1856   <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#459">459</a>), but it doesn't solve the problem raised by this
1857   issue.  Whether we use widen or narrow the <tt>num_get</tt> facet
1858   still has no idea which traits class the user wants to use for
1859   the comparison, because only streams, not facets, are passed traits
1860   classes.   The standard does not require that two different
1861   traits classes with the same <tt>char_type</tt> must necessarily
1862   have the same behavior.]</i></p>
1863
1864
1865 <p>Informally, one possibility: require that some of the basic
1866 character operations, such as <tt>eq</tt>, <tt>lt</tt>,
1867 and <tt>assign</tt>, must behave the same way for all traits classes
1868 with the same <tt>char_type</tt>.  If we accept that limitation on
1869 traits classes, then the facet could reasonably be required to
1870 use <tt>char_traits&lt;charT&gt;</tt>.</p>
1871
1872 <p><i>[
1873 2009-07 Frankfurt
1874 ]</i></p>
1875
1876
1877 <blockquote>
1878 <p>
1879 There was general agreement that the standard only needs to specify the
1880 behavior when the character type is char or wchar_t.
1881 </p>
1882 <p>
1883 Beman: we don't need to worry about C++1x because there is a non-zero
1884 possibility that we would have a replacement facility for iostreams that
1885 would solve these problems.
1886 </p>
1887 <p>
1888 We need to change the following sentence in [locale.category], paragraph
1889 6 to specify that C is char and wchar_t:
1890 </p>
1891 <p>
1892 "A template formal parameter with name C represents the set of all
1893 possible specializations on a parameter that satisfies the requirements
1894 for a character on which any member of the iostream components can be
1895 instantiated."
1896 </p>
1897 <p>
1898 We also need to specify in 27 that the basic character operations, such
1899 as eq, lt, and assign use std::char_traits.
1900 </p>
1901 <p>
1902 Daniel volunteered to provide wording.
1903 </p>
1904 </blockquote>
1905
1906 <p><i>[
1907 2009-09-19 Daniel provided wording.
1908 ]</i></p>
1909
1910
1911 <p><i>[
1912 2009-10 Santa Cruz:
1913 ]</i></p>
1914
1915
1916 <blockquote>
1917 Leave as Open. Alisdair and/or Tom will provide wording based on discussions.
1918 We want to clearly state that streams and locales work just on <tt>char</tt>
1919 and <tt>wchar_t</tt> (except where otherwise specified).
1920 </blockquote>
1921
1922
1923
1924 <p><b>Proposed resolution:</b></p>
1925 <ol>
1926 <li>
1927 <p>
1928 Change 22.3.1.1.1 [locale.category]/6:
1929 </p>
1930
1931 <blockquote>
1932 [..] A template formal parameter with name <tt>C</tt> represents the set of all possible
1933 specializations on a <ins><tt>char</tt> or <tt>wchar_t</tt></ins> parameter<del> that satisfies
1934 the requirements for a character on which any of the iostream components
1935 can be instantiated</del>. [..]
1936 </blockquote>
1937 </li>
1938
1939 <li>
1940 <p>
1941 Add the following sentence to the end of 22.4.2 [category.numeric]/2:
1942 </p>
1943
1944 <blockquote>
1945 [..] These specializations refer to [..], and also for the <tt>ctype&lt;&gt;</tt> facet to
1946 perform character classification. <ins>Implementations are encouraged
1947 but not required to use the <tt>char_traits&lt;charT&gt;</tt> functions for all
1948 comparisons and assignments of characters of type <tt>charT</tt> that do
1949 not belong to the set of required specializations.</ins>
1950 </blockquote>
1951 </li>
1952
1953 <li>
1954 <p>
1955 Change 22.4.2.1.2 [facet.num.get.virtuals]/3:
1956 </p>
1957
1958 <blockquote>
1959 <p>
1960 Stage 2: If <tt>in==end</tt> then stage 2 terminates. Otherwise a <tt>charT</tt> is taken
1961 from <tt>in</tt> and local variables are initialized as if by
1962 </p>
1963
1964 <blockquote><pre>char_type ct = *in;
1965 <ins>using tr = char_traits&lt;char_type&gt;;
1966 const char_type* pos = tr::find(atoms, sizeof(src) - 1, ct);</ins>
1967 char c = src[<del>find(atoms, atoms + sizeof(src) - 1, ct) - atoms</del>
1968              <ins>pos ? pos - atoms : sizeof(src) - 1</ins>];
1969 if (<ins>tr::eq(ct, </ins><del>ct == </del>use_facet&lt;numpunct&lt;charT&gt;(loc).decimal_point()<ins>)</ins>)
1970     c = '.';
1971 bool discard =
1972     <ins>tr::eq(ct, </ins><del>ct == </del>use_facet&lt;numpunct&lt;charT&gt;(loc).thousands_sep()<ins>)</ins>
1973     &amp;&amp; use_facet&lt;numpunct&lt;charT&gt; &gt;(loc).grouping().length() != 0;
1974 </pre></blockquote>
1975
1976 <p>
1977 where the values <tt>src</tt> and <tt>atoms</tt> are defined as if by: [..]
1978 </p>
1979 </blockquote>
1980
1981 <p>
1982 [Remark of the author: I considered to replace the initialization
1983 "<tt>char_type ct = *in;</tt>"
1984 by the sequence "<tt>char_type ct; tr::assign(ct, *in);</tt>", but decided
1985 against it, because
1986 it is a copy-initialization context, not an assignment]
1987 </p>
1988 </li>
1989
1990 <li>
1991 <p>
1992 Add the following sentence to the end of 22.4.5 [category.time]/1:
1993 </p>
1994
1995 <blockquote>
1996 [..] Their members use [..] , to determine formatting details.
1997 <ins>Implementations are encouraged but not required to use the
1998 <tt>char_traits&lt;charT&gt;</tt> functions for all comparisons and assignments
1999 of characters of type <tt>charT</tt> that do
2000 not belong to the set of required specializations.</ins>
2001 </blockquote>
2002 </li>
2003
2004 <li>
2005 <p>
2006 Change 22.4.5.1.1 [locale.time.get.members]/8 bullet 4:
2007 </p>
2008
2009 <ul>
2010 <li>
2011 <del>The next element of <tt>fmt</tt> is equal to <tt>'%'</tt></del> <ins>For the next element <tt>c</tt>
2012 of <tt>fmt char_traits&lt;char_type&gt;::eq(c, use_facet&lt;ctype&lt;char_type&gt;&gt;(f.getloc()).widen('%')) == true</tt></ins>,
2013 [..]
2014 </li>
2015 </ul>
2016 </li>
2017
2018 <li>
2019 <p>
2020 Add the following sentence to the end of 22.4.6 [category.monetary]/2:
2021 </p>
2022
2023 <blockquote>
2024 Their members use [..] to determine formatting details.
2025 <ins>Implementations are encouraged but not required to use the
2026 <tt>char_traits&lt;charT&gt;</tt> functions for all comparisons and assignments
2027 of characters of type <tt>charT</tt> that do
2028 not belong to the set of required specializations.</ins>
2029 </blockquote>
2030 </li>
2031
2032 <li>
2033 <p>
2034 Change 22.4.6.1.2 [locale.money.get.virtuals]/4:
2035 </p>
2036
2037 <blockquote>
2038 <p>
2039 [..] The value <tt>units</tt> is produced as if by:
2040 </p>
2041
2042 <blockquote><pre>for (int i = 0; i &lt; n; ++i)
2043   buf2[i] = src[<ins>char_traits&lt;charT&gt;::</ins>find(atoms, <del>atoms+</del>sizeof(src), buf1[i]) - atoms];
2044 buf2[n] = 0;
2045 sscanf(buf2, "%Lf", &amp;units);
2046 </pre></blockquote>
2047 </blockquote>
2048 </li>
2049
2050 <li>
2051 <p>
2052 Change 22.4.6.2.2 [locale.money.put.virtuals]/1:
2053 </p>
2054
2055 <blockquote>
2056 [..] for character buffers <tt>buf1</tt> and <tt>buf2</tt>. If <ins>for</ins> the first
2057 character <ins><tt>c</tt></ins>
2058 in <tt>digits</tt> or <tt>buf2</tt> <del>is equal to
2059 <tt>ct.widen('-')</tt></del><ins><tt>char_traits&lt;charT&gt;::eq(c,
2060 ct.widen('-')) == true</tt></ins>, [..]
2061 </blockquote>
2062 </li>
2063
2064 <li>
2065 <p>
2066 Add a footnote to the first sentence of 27.7.1.2.2 [istream.formatted.arithmetic]/1:
2067 </p>
2068
2069 <blockquote>
2070 <p>
2071 As in the case of the inserters, these extractors depend on the locale's
2072 <tt>num_get&lt;&gt;</tt> (22.4.2.1) object to perform parsing the input stream
2073 data.<ins><sup>(footnote)</sup></ins> [..]
2074 </p>
2075
2076 <p>
2077 <ins>
2078 <sup>footnote)</sup> If the traits of the input stream has different semantics for <tt>lt()</tt>,
2079 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2080 results.
2081 </ins>
2082 </p>
2083 </blockquote>
2084 </li>
2085
2086 <li>
2087 <p>
2088 Add a footnote to the second sentence of 27.7.2.6.2 [ostream.inserters.arithmetic]/1:
2089 </p>
2090
2091 <blockquote>
2092 <p>
2093 <i>Effects:</i> The classes <tt>num_get&lt;&gt;</tt> and
2094 <tt>num_put&lt;&gt;</tt> handle locale-dependent numeric formatting and
2095 parsing. These inserter functions use the imbued locale value to perform
2096 numeric formatting.<ins><sup>(footnote)</sup></ins> [..]
2097 </p>
2098
2099 <p>
2100 <ins>
2101 <sup>footnote)</sup> If the traits of the output stream has different semantics for <tt>lt()</tt>,
2102 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2103 results.
2104 </ins>
2105 </p>
2106 </blockquote>
2107 </li>
2108
2109 <li>
2110 <p>
2111 Add a footnote after the first sentence of 27.7.4 [ext.manip]/4:
2112 </p>
2113
2114 <blockquote>
2115 <p>
2116 <i>Returns:</i> An object of unspecified type such that if in is an object of type
2117 <tt>basic_istream&lt;charT, traits&gt;</tt> then the expression <tt>in &gt;&gt; get_money(mon, intl)</tt>
2118 behaves as if it called <tt>f(in, mon, intl)</tt>, where the function <tt>f</tt> is defined
2119 as:<ins><sup>(footnote)</sup></ins> [..]
2120 </p>
2121
2122 <p>
2123 <ins>
2124 <sup>footnote)</sup> If the traits of the input stream has different semantics for <tt>lt()</tt>,
2125 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2126 results.
2127 </ins>
2128 </p>
2129 </blockquote>
2130 </li>
2131
2132 <li>
2133 <p>
2134 Add a footnote after the first sentence of 27.7.4 [ext.manip]/5:
2135 </p>
2136
2137 <blockquote>
2138 <p>
2139 <i>Returns:</i> An object of unspecified type such that if <tt>out</tt> is an object of type
2140 <tt>basic_ostream&lt;charT, traits&gt;</tt> then the expression <tt>out &lt;&lt; put_money(mon, intl)</tt>
2141 behaves as a formatted input function that calls <tt>f(out, mon, intl)</tt>, where the
2142 function <tt>f</tt> is defined as:<ins><sup>(footnote)</sup></ins> [..]
2143 </p>
2144
2145 <p>
2146 <ins>
2147 <sup>footnote)</sup> If the traits of the output stream has different semantics for <tt>lt()</tt>,
2148 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2149 results.
2150 </ins>
2151 </p>
2152 </blockquote>
2153 </li>
2154
2155 <li>
2156 <p>
2157 13) Add a footnote after the first sentence of 27.7.4 [ext.manip]/8:
2158 </p>
2159
2160 <blockquote>
2161 <p>
2162 <i>Returns:</i> An object of unspecified type such that if <tt>in</tt> is an
2163 object of type b<tt>asic_istream&lt;charT, traits&gt;</tt> then the expression
2164 <tt>in &gt;&gt;get_time(tmb, fmt)</tt> behaves as if it called <tt>f(in, tmb, fmt)</tt>,
2165 where the function <tt>f</tt> is defined as:<ins><sup>(footnote)</sup></ins> [..]
2166 </p>
2167
2168 <p>
2169 <ins>
2170 <sup>footnote)</sup> If the traits of the input stream has different semantics for <tt>lt()</tt>,
2171 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2172 results.
2173 </ins>
2174 </p>
2175 </blockquote>
2176 </li>
2177
2178 <li>
2179 <p>
2180 Add a footnote after the first sentence of 27.7.4 [ext.manip]/10:
2181 </p>
2182
2183 <blockquote>
2184 <p>
2185 Returns: An object of unspecified type such that if <tt>out</tt> is an object of type
2186 <tt>basic_ostream&lt;charT, traits&gt;</tt> then the expression <tt>out &lt;&lt;put_time(tmb, fmt)</tt>
2187 behaves as if it called <tt>f(out, tmb, fmt)</tt>, where the function <tt>f</tt> is defined
2188 as:<ins><sup>(footnote)</sup></ins> [..]
2189 </p>
2190
2191 <p>
2192 <ins>
2193 <sup>footnote)</sup> If the traits of the output stream has different semantics for <tt>lt()</tt>,
2194 <tt>eq()</tt>, and <tt>assign()</tt> than <tt>char_traits&lt;char_type&gt;</tt>, this may give surprising
2195 results.
2196 </ins>
2197 </p>
2198 </blockquote>
2199 </li>
2200 </ol>
2201
2202
2203
2204
2205
2206 <hr>
2207 <h3><a name="430"></a>430. valarray subset operations</h3>
2208 <p><b>Section:</b> 26.6.2.4 [valarray.sub] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
2209  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2003-09-18  <b>Last modified:</b> 2009-11-04</p>
2210 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
2211 <p><b>Discussion:</b></p>
2212 <p>
2213 The standard fails to specify the behavior of valarray::operator[](slice)
2214 and other valarray subset operations when they are passed an "invalid"
2215 slice object, i.e., either a slice that doesn't make sense at all (e.g.,
2216 slice (0, 1, 0) or one that doesn't specify a valid subset of the valarray
2217 object (e.g., slice (2, 1, 1) for a valarray of size 1).
2218 </p>
2219 <p><i>[Kona: the LWG believes that invalid slices should invoke
2220   undefined behavior.  Valarrays are supposed to be designed for high
2221   performance, so we don't want to require specific checking.  We
2222   need wording to express this decision.]</i></p>
2223
2224
2225 <p><i>[
2226 Bellevue:
2227 ]</i></p>
2228
2229
2230 <blockquote>
2231 Please note that the standard also fails to specify the behavior of
2232 slice_array and gslice_array in the valid case. Bill Plauger will
2233 endeavor to provide revised wording for slice_array and gslice_array.
2234 </blockquote>
2235
2236 <p><i>[
2237 post-Bellevue:  Bill provided wording.
2238 ]</i></p>
2239
2240
2241 <p><i>[
2242 2009-07 Frankfurt
2243 ]</i></p>
2244
2245
2246 <blockquote>
2247 <p>
2248 Move to Ready.
2249 </p>
2250 </blockquote>
2251
2252 <p><i>[
2253 2009-11-04 Pete opens:
2254 ]</i></p>
2255
2256
2257 <blockquote>
2258 The resolution to LWG issue 430 has not been applied --- there have been
2259 changes to the underlying text, and the resolution needs to be reworked.
2260 </blockquote>
2261
2262
2263
2264 <p><b>Proposed resolution:</b></p>
2265 <p>
2266 Insert after 26.6.2.4 [valarray.sub], paragraph 1:
2267 </p>
2268
2269 <blockquote>
2270 <p>
2271 The member operator is overloaded to provide several ways to select
2272 sequences
2273 of elements from among those controlled by <tt>*this</tt>. The first group of five
2274 member operators work in conjunction with various overloads of <tt>operator=</tt>
2275 (and other assigning operators) to allow selective replacement (slicing) of
2276 the controlled sequence. The selected elements must exist.
2277 </p>
2278 <p>
2279 The first member operator selects element off. For example:
2280 </p>
2281
2282 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2283 v0[3] = 'A';
2284 // v0 == valarray&lt;char&gt;("abcAefghijklmnop", 16)
2285 </pre></blockquote>
2286
2287 <p>
2288 The second member operator selects those elements of the controlled sequence
2289 designated by <tt>slicearr</tt>. For example:
2290 </p>
2291
2292 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2293 valarray&lt;char&gt; v1("ABCDE", 5);
2294 v0[slice(2, 5, 3)] = v1;
2295 // v0 == valarray&lt;char&gt;("abAdeBghCjkDmnEp", 16)
2296 </pre></blockquote>
2297
2298 <p>
2299 The third member operator selects those elements of the controlled sequence
2300 designated by <tt>gslicearr</tt>. For example:
2301 </p>
2302
2303 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2304 valarray&lt;char&gt; v1("ABCDEF", 6);
2305 const size_t lv[] = {2, 3};
2306 const size_t dv[] = {7, 2};
2307 const valarray&lt;size_t&gt; len(lv, 2), str(dv, 2);
2308 v0[gslice(3, len, str)] = v1;
2309 // v0 == valarray&lt;char&gt;("abcAeBgCijDlEnFp", 16)
2310 </pre></blockquote>
2311
2312 <p>
2313 The fourth member operator selects those elements of the controlled sequence
2314 designated by <tt>boolarr</tt>. For example:
2315 </p>
2316
2317 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2318 valarray&lt;char&gt; v1("ABC", 3);
2319 const bool vb[] = {false, false, true, true, false, true};
2320 v0[valarray&lt;bool&gt;(vb, 6)] = v1;
2321 // v0 == valarray&lt;char&gt;("abABeCghijklmnop", 16)
2322 </pre></blockquote>
2323
2324 <p>
2325 The fifth member operator selects those elements of the controlled sequence
2326 designated by indarr. For example:
2327 </p>
2328
2329 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2330 valarray&lt;char&gt; v1("ABCDE", 5);
2331 const size_t vi[] = {7, 5, 2, 3, 8};
2332 v0[valarray&lt;size_t&gt;(vi, 5)] = v1;
2333 // v0 == valarray&lt;char&gt;("abCDeBgAEjklmnop", 16)
2334 </pre></blockquote>
2335
2336 <p>
2337 The second group of five member operators each construct an object that
2338 represents the value(s) selected. The selected elements must exist.
2339 </p>
2340
2341 <p>
2342 The sixth member operator returns the value of element off. For example:
2343 </p>
2344
2345 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2346 // v0[3] returns 'd'
2347 </pre></blockquote>
2348
2349 <p>
2350 The seventh member operator returns an object of class <tt>valarray&lt;Ty&gt;</tt>
2351 containing those elements of the controlled sequence designated by <tt>slicearr</tt>.
2352 For example:
2353 </p>
2354
2355 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2356 // v0[slice(2, 5, 3)] returns valarray&lt;char&gt;("cfilo", 5)
2357 </pre></blockquote>
2358
2359 <p>
2360 The eighth member operator selects those elements of the controlled sequence
2361 designated by <tt>gslicearr</tt>. For example:
2362 </p>
2363
2364 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2365 const size_t lv[] = {2, 3};
2366 const size_t dv[] = {7, 2};
2367 const valarray&lt;size_t&gt; len(lv, 2), str(dv, 2);
2368 // v0[gslice(3, len, str)] returns
2369 //    valarray&lt;char&gt;("dfhkmo", 6)
2370 </pre></blockquote>
2371
2372 <p>
2373 The ninth member operator selects those elements of the controlled sequence
2374 designated by <tt>boolarr</tt>. For example:
2375 </p>
2376
2377 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2378 const bool vb[] = {false, false, true, true, false, true};
2379 // v0[valarray&lt;bool&gt;(vb, 6)] returns
2380 //    valarray&lt;char&gt;("cdf", 3)
2381 </pre></blockquote>
2382
2383 <p>
2384 The last member operator selects those elements of the controlled sequence
2385 designated by <tt>indarr</tt>. For example:
2386 </p>
2387
2388 <blockquote><pre>valarray&lt;char&gt; v0("abcdefghijklmnop", 16);
2389 const size_t vi[] = {7, 5, 2, 3, 8};
2390 // v0[valarray&lt;size_t&gt;(vi, 5)] returns
2391 //    valarray&lt;char&gt;("hfcdi", 5)
2392 </pre></blockquote>
2393
2394 </blockquote>
2395
2396
2397
2398
2399
2400 <hr>
2401 <h3><a name="446"></a>446. Iterator equality between different containers</h3>
2402 <p><b>Section:</b> 24.2 [iterator.requirements], 23.2 [container.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
2403  <b>Submitter:</b> Andy Koenig <b>Opened:</b> 2003-12-16  <b>Last modified:</b> 2009-11-03</p>
2404 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
2405 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
2406 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
2407 <p><b>Discussion:</b></p>
2408 <p>
2409 What requirements does the standard place on equality comparisons between
2410 iterators that refer to elements of different containers.  For example, if
2411 v1 and v2 are empty vectors, is v1.end() == v2.end() allowed to yield true?
2412 Is it allowed to throw an exception?
2413 </p>
2414
2415 <p>
2416 The standard appears to be silent on both questions.
2417 </p>
2418 <p><i>[Sydney: The intention is that comparing two iterators from
2419 different containers is undefined, but it's not clear if we say that,
2420 or even whether it's something we should be saying in clause 23 or in
2421 clause 24.  Intuitively we might want to say that equality is defined
2422 only if one iterator is reachable from another, but figuring out how
2423 to say it in any sensible way is a bit tricky: reachability is defined
2424 in terms of equality, so we can't also define equality in terms of
2425 reachability.
2426 ]</i></p>
2427
2428
2429 <p><i>[
2430 2009-07 Frankfurt
2431 ]</i></p>
2432
2433
2434 <blockquote>
2435 Daniel volunteered to work on this.
2436 </blockquote>
2437
2438 <p><i>[
2439 2009-09-20 Daniel provided wording.
2440 ]</i></p>
2441
2442
2443 <p><i>[
2444 2009-10 Santa Cruz:
2445 ]</i></p>
2446
2447
2448 <blockquote>
2449 Leave as Open. Alisdair has volunteered to refine the wording.
2450 </blockquote>
2451
2452
2453
2454 <p><b>Proposed resolution:</b></p>
2455 <p>
2456 Insert a new paragraph between 24.2 [iterator.requirements]/7+8:
2457 </p>
2458
2459 <blockquote>
2460 <p>
2461 [..] The result of the application of functions in the library to invalid
2462 ranges is undefined.
2463 </p>
2464
2465 <p><ins>The result of directly or indirectly evaluating any comparison function
2466 or the binary - operator with two iterator values as arguments that
2467 were obtained
2468 from two different ranges <tt>r1</tt> and <tt>r2</tt> (including their past-the-end values) which
2469 are not subranges of one common range is undefined, unless explicitly
2470 described otherwise.</ins>
2471 </p>
2472
2473 </blockquote>
2474
2475
2476
2477
2478
2479
2480 <hr>
2481 <h3><a name="471"></a>471. result of what() implementation-defined</h3>
2482 <p><b>Section:</b> 18.8.1 [exception] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
2483  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2004-06-28  <b>Last modified:</b> 2009-10-26</p>
2484 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
2485 <p><b>Discussion:</b></p>
2486
2487 <p>[lib.exception] specifies the following:</p>
2488 <pre>    exception (const exception&amp;) throw();
2489     exception&amp; operator= (const exception&amp;) throw();
2490
2491     -4- Effects: Copies an exception object.
2492     -5- Notes: The effects of calling what() after assignment
2493         are implementation-defined.
2494 </pre>
2495
2496 <p>
2497 First, does the Note only apply to the assignment operator? If so,
2498 what are the effects of calling what() on a copy of an object? Is
2499 the returned pointer supposed to point to an identical copy of
2500 the NTBS returned by what() called on the original object or not?
2501 </p>
2502
2503 <p>
2504 Second, is this Note intended to extend to all the derived classes
2505 in section 19? I.e., does the standard provide any guarantee for
2506 the effects of what() called on a copy of any of the derived class
2507 described in section 19?
2508 </p>
2509
2510 <p>
2511 Finally, if the answer to the first question is no, I believe it
2512 constitutes a defect since throwing an exception object typically
2513 implies invoking the copy ctor on the object. If the answer is yes,
2514 then I believe the standard ought to be clarified to spell out
2515 exactly what the effects are on the copy (i.e., after the copy
2516 ctor was called).
2517 </p>
2518
2519 <p><i>[Redmond: Yes, this is fuzzy.  The issue of derived classes is
2520   fuzzy too.]</i></p>
2521
2522
2523 <p><i>[
2524 Batavia: Howard provided wording.
2525 ]</i></p>
2526
2527
2528 <p><i>[
2529 Bellevue:
2530 ]</i></p>
2531
2532
2533 <blockquote>
2534 <p>
2535 Eric concerned this is unimplementable, due to nothrow guarantees.
2536 Suggested implementation would involve reference counting.
2537 </p>
2538 <p>
2539 Is the implied reference counting subtle enough to call out a note on
2540 implementation? Probably not.
2541 </p>
2542 <p>
2543 If reference counting required, could we tighten specification further
2544 to require same pointer value? Probably an overspecification, especially
2545 if exception classes defer evalutation of final string to calls to
2546 what().
2547 </p>
2548 <p>
2549 Remember issue moved open and not resolved at Batavia, but cannot
2550 remember who objected to canvas a disenting opinion - please speak up if
2551 you disagree while reading these minutes!
2552 </p>
2553 <p>
2554 Move to Ready as we are accepting words unmodified.
2555 </p>
2556 </blockquote>
2557
2558 <p><i>[
2559 Sophia Antipolis:
2560 ]</i></p>
2561
2562
2563 <blockquote>
2564 The issue was pulled from Ready.  It needs to make clear that only homogenous copying
2565 is intended to be supported, not coping from a derived to a base.
2566 </blockquote>
2567
2568 <p><i>[
2569 Batavia (2009-05):
2570 ]</i></p>
2571
2572 <blockquote>
2573 <p>
2574 Howard supplied the following replacement wording
2575 for paragraph 7 of the proposed resolution:
2576 </p>
2577 <blockquote>
2578 <ins>-7- <i>Postcondition:</i> <tt>what()</tt> shall return the same NTBS
2579   as would be obtained by using <tt>static_cast</tt>
2580   to cast the rhs to the same types as the lhs
2581   and then calling <tt>what()</tt> on that possibly sliced object.</ins>
2582 </blockquote>
2583 <p>
2584 Pete asks what "the same NTBS" means.
2585 </p>
2586 </blockquote>
2587
2588 <p><i>[
2589 2009-07-30 Niels adds:
2590 ]</i></p>
2591
2592
2593 <blockquote>
2594 Further discussion in the thread starting with c++std-lib-24512.
2595 </blockquote>
2596
2597 <p><i>[
2598 2009-09-24 Niels provided updated wording:
2599 ]</i></p>
2600
2601
2602 <blockquote>
2603 <p>
2604 I think the resolution should at least guarantee
2605 that the result of <tt>what()</tt> is independent of whether the compiler does
2606 copy-elision. And for any class derived from <tt>std::excepion</tt> that has a
2607 constructor that allows specifying a <tt>what_arg</tt>, it should make sure that
2608 the text of a user-provided <tt>what_arg</tt> is preserved, when the object is
2609 copied. Note that all the implementations I've tested already appear to
2610 satisfy the proposed resolution, including MSVC 2008 SP1, Apache
2611 stdcxx-4.2.1, GCC 4.1.2, GCC 4.3.2, and CodeGear C++ 6.13.
2612 </p>
2613 <p>
2614 The proposed resolution was updated with help from Daniel Krügler;
2615 the update aims to clarify that the proposed postcondition only
2616 applies to homogeneous copying.
2617 </p>
2618 </blockquote>
2619
2620 <p><i>[
2621 2009-10 Santa Cruz:
2622 ]</i></p>
2623
2624
2625 <blockquote>
2626 Moved to Ready after inserting "publicly accessible" in two places.
2627 </blockquote>
2628
2629
2630
2631 <p><b>Proposed resolution:</b></p>
2632
2633 <p>
2634 Change 18.8.1 [exception] to:
2635 </p>
2636
2637 <blockquote>
2638 <p>
2639 -1- The class <tt>exception</tt> defines the base class for the types of
2640 objects thrown as exceptions by C++ standard library components, and
2641 certain expressions, to report errors detected during program execution.
2642 </p>
2643 <p><ins>
2644 Each standard library class <tt>T</tt> that derives from class
2645 <tt>exception</tt> shall have a publicly accessible copy constructor and a publicly accessible copy assignment
2646 operator that do not exit with an exception. These member functions
2647 shall preserve the following postcondition: If two objects <i>lhs</i>
2648 and <i>rhs</i> both have dynamic type <tt>T</tt>, and <i>lhs</i> is a
2649 copy of <i>rhs</i>, then <tt>strcmp(<i>lhs</i>.what(),
2650 <i>rhs</i>.what()) == 0</tt>.
2651 </ins></p>
2652 <p>
2653  ...
2654 </p>
2655
2656 <pre>exception(const exception&amp; <ins>rhs</ins>) throw();
2657 exception&amp; operator=(const exception&amp; <ins>rhs</ins>) throw();</pre>
2658
2659 <blockquote>
2660 <p>
2661 -4- <i>Effects:</i> Copies an exception object.
2662 </p>
2663 <p>
2664 <del> -5- <i>Remarks:</i> The effects of calling <tt>what()</tt> after assignment
2665 are implementation-defined.</del>
2666 </p>
2667 <p>
2668 <ins>-5- <i>Postcondition:</i>
2669         If <tt>*this</tt>
2670         and <i>rhs</i> both have dynamic type <tt>exception</tt>
2671         then <tt>strcmp(what(), <i>rhs</i>.what()) == 0</tt>.</ins>
2672 </p>
2673
2674 </blockquote>
2675
2676 </blockquote>
2677
2678
2679
2680
2681
2682 <hr>
2683 <h3><a name="473"></a>473. underspecified ctype calls</h3>
2684 <p><b>Section:</b> 22.4.1.1 [locale.ctype] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
2685  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2004-07-01  <b>Last modified:</b> 2009-10-21</p>
2686 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
2687 <p><b>Discussion:</b></p>
2688 <p>
2689 Most ctype member functions come in two forms: one that operates
2690 on a single character at a time and another form that operates
2691 on a range of characters. Both forms are typically described by
2692 a single Effects and/or Returns clause.
2693 </p>
2694 <p>
2695 The Returns clause of each of the single-character non-virtual forms
2696 suggests that the function calls the corresponding single character
2697 virtual function, and that the array form calls the corresponding
2698 virtual array form. Neither of the two forms of each virtual member
2699 function is required to be implemented in terms of the other.
2700 </p>
2701 <p>
2702 There are three problems:
2703 </p>
2704 <p>
2705 1. One is that while the standard does suggest that each non-virtual
2706 member function calls the corresponding form of the virtual function,
2707 it doesn't actually explicitly require it.
2708 </p>
2709 <p>
2710 Implementations that cache results from some of the virtual member
2711 functions for some or all values of their arguments might want to
2712 call the array form from the non-array form the first time to fill
2713 the cache and avoid any or most subsequent virtual calls. Programs
2714 that rely on each form of the virtual function being called from
2715 the corresponding non-virtual function will see unexpected behavior
2716 when using such implementations.
2717 </p>
2718 <p>
2719 2. The second problem is that either form of each of the virtual
2720 functions can be overridden by a user-defined function in a derived
2721 class to return a value that is different from the one produced by
2722 the virtual function of the alternate form that has not been
2723 overriden.
2724 </p>
2725 <p>
2726 Thus, it might be possible for, say, ctype::widen(c) to return one
2727 value, while for ctype::widen(&amp;c, &amp;c + 1, &amp;wc) to set
2728 wc to another value. This is almost certainly not intended. Both
2729 forms of every function should be required to return the same result
2730 for the same character, otherwise the same program using an
2731 implementation that calls one form of the functions will behave
2732 differently than when using another implementation that calls the
2733 other form of the function "under the hood."
2734 </p>
2735 <p>
2736 3. The last problem is that the standard text fails to specify whether
2737 one form of any of the virtual functions is permitted to be implemented
2738 in terms of the other form or not, and if so, whether it is required
2739 or permitted to call the overridden virtual function or not.
2740 </p>
2741 <p>
2742 Thus, a program that overrides one of the virtual functions so that
2743 it calls the other form which then calls the base member might end
2744 up in an infinite loop if the called form of the base implementation
2745 of the function in turn calls the other form.
2746 </p>
2747 <p>
2748 Lillehammer: Part of this isn't a real problem. We already talk about
2749 caching. 22.1.1/6 But part is a real problem. ctype virtuals may call
2750 each other, so users don't know which ones to override to avoid avoid
2751 infinite loops.</p>
2752
2753 <p>This is a problem for all facet virtuals, not just ctype virtuals,
2754 so we probably want a blanket statement in clause 22 for all
2755 facets. The LWG is leaning toward a blanket prohibition, that a
2756 facet's virtuals may never call each other. We might want to do that
2757 in clause 27 too, for that matter. A review is necessary.  Bill will
2758 provide wording.</p>
2759
2760 <p><i>[
2761 2009-07 Frankfurt, Howard provided wording directed by consensus.
2762 ]</i></p>
2763
2764
2765 <p><i>[
2766 2009-10 Santa Cruz:
2767 ]</i></p>
2768
2769
2770 <blockquote>
2771 Move to Ready.
2772 </blockquote>
2773
2774
2775
2776 <p><b>Proposed resolution:</b></p>
2777 <p>
2778 Add paragraph 3 to 22.4 [locale.categories]:
2779 </p>
2780
2781 <blockquote><ins>
2782 -3- Within this clause it is unspecified if one virtual function calls another
2783 virtual function.
2784 </ins></blockquote>
2785
2786
2787
2788 <p><b>Rationale:</b></p>
2789 <p>
2790 We are explicitly not addressing bullet
2791 item #2, thus giving implementors more latitude. Users will have to
2792 override both virtual functions, not just one.
2793 </p>
2794
2795
2796
2797
2798 <hr>
2799 <h3><a name="485"></a>485. output iterator insufficiently constrained</h3>
2800 <p><b>Section:</b> 24.2.2 [output.iterators] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
2801  <b>Submitter:</b> Chris Jefferson <b>Opened:</b> 2004-10-13  <b>Last modified:</b> 2009-10-26</p>
2802 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#output.iterators">issues</a> in [output.iterators].</p>
2803 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
2804 <p><b>Discussion:</b></p>
2805 <p>
2806 The note on 24.1.2 Output iterators insufficiently limits what can be
2807 performed on output iterators. While it requires that each iterator is
2808 progressed through only once and that each iterator is written to only
2809 once, it does not require the following things:</p>
2810
2811 <p>Note: Here it is assumed that x is an output iterator of type X which
2812 has not yet been assigned to.</p>
2813
2814 <p>a) That each value of the output iterator is written to:
2815 The standard allows:
2816 ++x; ++x; ++x;
2817 </p>
2818
2819 <p>
2820 b) That assignments to the output iterator are made in order
2821 X a(x); ++a; *a=1; *x=2; is allowed
2822 </p>
2823
2824 <p>
2825 c) Chains of output iterators cannot be constructed:
2826 X a(x); ++a; X b(a); ++b; X c(b); ++c; is allowed, and under the current
2827 wording (I believe) x,a,b,c could be written to in any order.
2828 </p>
2829
2830 <p>I do not believe this was the intension of the standard?</p>
2831 <p><i>[Lillehammer: Real issue.  There are lots of constraints we
2832   intended but didn't specify.  Should be solved as part of iterator
2833   redesign.]</i></p>
2834
2835
2836 <p><i>[
2837 2009-07 Frankfurt
2838 ]</i></p>
2839
2840
2841 <blockquote>
2842 Bill provided wording according to consensus.
2843 </blockquote>
2844
2845 <p><i>[
2846 2009-07-21 Alisdair requests change from Review to Open.  See thread starting
2847 with c++std-lib-24459 for discussion.
2848 ]</i></p>
2849
2850
2851 <p><i>[
2852 2009-10 Santa Cruz:
2853 ]</i></p>
2854
2855
2856 <blockquote>
2857 Modified wording.  Set to Review.
2858 </blockquote>
2859
2860 <p><i>[
2861 2009-10 Santa Cruz:
2862 ]</i></p>
2863
2864
2865 <blockquote>
2866 Move to Ready after looking at again in a larger group in Santa Cruz.
2867 </blockquote>
2868
2869
2870
2871 <p><b>Proposed resolution:</b></p>
2872 <p>
2873 Change Table 101 &#8212; Output iterator requirements in 24.2.2 [output.iterators]:
2874 </p>
2875 <blockquote>
2876 <table border="1">
2877 <caption>Table 101 &#8212; Output iterator requirements</caption>
2878 <tbody><tr>
2879 <th>Expression</th>
2880 <th>Return type</th>
2881 <th>Operational semantics</th>
2882 <th>Assertion/note pre-/post-condition</th>
2883 </tr>
2884
2885 <tr>
2886 <td>
2887 <tt>X(a)</tt>
2888 </td>
2889 <td>
2890 &nbsp;
2891 </td>
2892 <td>
2893 &nbsp;
2894 </td>
2895 <td>
2896 <tt>a = t</tt> is equivalent to <tt>X(a) = t</tt>. note: a destructor is assumed.
2897 </td>
2898 </tr>
2899
2900 <tr>
2901 <td>
2902 <tt>X u(a);</tt><br>
2903 <tt>X u = a;</tt>
2904 </td>
2905 <td>
2906 &nbsp;
2907 </td>
2908 <td>
2909 &nbsp;
2910 </td>
2911 <td>
2912 &nbsp;
2913 </td>
2914 </tr>
2915
2916 <tr>
2917 <td>
2918 <tt>*r = o</tt>
2919 </td>
2920 <td>
2921 result is not used
2922 </td>
2923 <td>
2924 &nbsp;
2925 </td>
2926 <td>
2927 <ins>
2928 Post: <tt>r</tt> is not required to be dereferenceable.  <tt>r</tt> is incrementable.
2929 </ins>
2930 </td>
2931 </tr>
2932
2933 <tr>
2934 <td>
2935 <tt>++r</tt>
2936 </td>
2937 <td>
2938 <tt>X&amp;</tt>
2939 </td>
2940 <td>
2941 &nbsp;
2942 </td>
2943 <td>
2944 <tt>&amp;r == &amp;++r</tt>
2945 <ins>
2946 Post: <tt>r</tt> is dereferenceable, unless otherwise specified.  <tt>r</tt> is not required to be incrementable.
2947 </ins>
2948 </td>
2949 </tr>
2950
2951 <tr>
2952 <td>
2953 <tt>r++</tt>
2954 </td>
2955 <td>
2956 convertible to <tt>const X&amp;</tt>
2957 </td>
2958 <td>
2959 <tt>{X tmp = r;<br>++r;<br>return tmp;}</tt>
2960 </td>
2961 <td>
2962 <ins>
2963 Post: <tt>r</tt> is dereferenceable, unless otherwise specified. <tt>r</tt> is not required to be incrementable.
2964 </ins>
2965 </td>
2966 </tr>
2967
2968 <tr>
2969 <td>
2970 <tt>*r++ = o;</tt>
2971 </td>
2972 <td>
2973 result is not used
2974 </td>
2975 <td>
2976 &nbsp;
2977 </td>
2978 <td>
2979
2980 </td>
2981 </tr>
2982
2983 </tbody></table>
2984 </blockquote>
2985
2986
2987
2988
2989
2990 <hr>
2991 <h3><a name="539"></a>539. partial_sum and adjacent_difference should mention requirements</h3>
2992 <p><b>Section:</b> 26.7.3 [partial.sum] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
2993  <b>Submitter:</b> Marc Schoolderman <b>Opened:</b> 2006-02-06  <b>Last modified:</b> 2009-10-24</p>
2994 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
2995 <p><b>Discussion:</b></p>
2996 <p>
2997 There are some problems in the definition of partial_sum and
2998 adjacent_difference in 26.4 [lib.numeric.ops]
2999 </p>
3000
3001 <p>
3002 Unlike <tt>accumulate</tt> and <tt>inner_product</tt>, these functions are not
3003 parametrized on a "type T", instead, 26.4.3 [lib.partial.sum] simply
3004 specifies the effects clause as;
3005 </p>
3006
3007 <blockquote><p>
3008 Assigns to every element referred to by iterator <tt>i</tt> in the range
3009 <tt>[result,result + (last - first))</tt> a value correspondingly equal to
3010 </p>
3011 <blockquote><pre>((...(* first + *( first + 1)) + ...) + *( first + ( i - result )))
3012 </pre></blockquote>
3013 </blockquote>
3014
3015 <p>
3016 And similarly for BinaryOperation. Using just this definition, it seems
3017 logical to expect that:
3018 </p>
3019
3020
3021 <blockquote><pre>char i_array[4] = { 100, 100, 100, 100 };
3022 int  o_array[4];
3023
3024 std::partial_sum(i_array, i_array+4, o_array);
3025 </pre></blockquote>
3026
3027 <p>
3028 Is equivalent to
3029 </p>
3030
3031 <blockquote><pre>int o_array[4] = { 100, 100+100, 100+100+100, 100+100+100+100 };
3032 </pre></blockquote>
3033
3034 <p>
3035 i.e. 100, 200, 300, 400, with addition happening in the <tt>result type</tt>,
3036 <tt>int</tt>.
3037 </p>
3038
3039 <p>
3040 Yet all implementations I have tested produce 100, -56, 44, -112,
3041 because they are using an accumulator of the <tt>InputIterator</tt>'s
3042 <tt>value_type</tt>, which in this case is <tt>char</tt>, not <tt>int</tt>.
3043 </p>
3044
3045 <p>
3046 The issue becomes more noticeable when the result of the expression <tt>*i +
3047 *(i+1)</tt> or <tt>binary_op(*i, *i-1)</tt> can't be converted to the
3048 <tt>value_type</tt>. In a contrived example:
3049 </p>
3050
3051 <blockquote><pre>enum not_int { x = 1, y = 2 };
3052 ...
3053 not_int e_array[4] = { x, x, y, y };
3054 std::partial_sum(e_array, e_array+4, o_array);
3055 </pre></blockquote>
3056
3057 <p>
3058 Is it the intent that the operations happen in the <tt>input type</tt>, or in
3059 the <tt>result type</tt>?
3060 </p>
3061
3062 <p>
3063 If the intent is that operations happen in the <tt>result type</tt>, something
3064 like this should be added to the "Requires" clause of 26.4.3/4
3065 [lib.partial.sum]:
3066 </p>
3067
3068 <blockquote><p>
3069 The type of <tt>*i + *(i+1)</tt> or <tt>binary_op(*i, *(i+1))</tt> shall meet the
3070 requirements of <tt>CopyConstructible</tt> (20.1.3) and <tt>Assignable</tt>
3071 (23.1) types.
3072 </p></blockquote>
3073
3074 <p>
3075 (As also required for <tt>T</tt> in 26.4.1 [lib.accumulate] and 26.4.2
3076 [lib.inner.product].)
3077 </p>
3078
3079 <p>
3080 The "auto initializer" feature proposed in
3081 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2005/n1894.pdf">N1894</a>
3082 is not required to
3083 implement <tt>partial_sum</tt> this way. The 'narrowing' behaviour can still be
3084 obtained by using the <tt>std::plus&lt;&gt;</tt> function object.
3085 </p>
3086
3087 <p>
3088 If the intent is that operations happen in the <tt>input type</tt>, then
3089 something like this should be added instead;
3090 </p>
3091
3092 <blockquote><p>
3093 The type of *first shall meet the requirements of
3094 <tt>CopyConstructible</tt> (20.1.3) and <tt>Assignable</tt> (23.1) types.
3095 The result of <tt>*i + *(i+1)</tt> or <tt>binary_op(*i, *(i+1))</tt> shall be
3096 convertible to this type.
3097 </p></blockquote>
3098
3099 <p>
3100 The 'widening' behaviour can then be obtained by writing a custom proxy
3101 iterator, which is somewhat involved.
3102 </p>
3103
3104 <p>
3105 In both cases, the semantics should probably be clarified.
3106 </p>
3107
3108 <p>
3109 26.4.4 [lib.adjacent.difference] is similarly underspecified, although
3110 all implementations seem to perform operations in the 'result' type:
3111 </p>
3112
3113 <blockquote><pre>unsigned char i_array[4] = { 4, 3, 2, 1 };
3114 int o_array[4];
3115
3116 std::adjacent_difference(i_array, i_array+4, o_array);
3117 </pre></blockquote>
3118
3119 <p>
3120 o_array is 4, -1, -1, -1 as expected, not 4, 255, 255, 255.
3121 </p>
3122
3123 <p>
3124 In any case, <tt>adjacent_difference</tt> doesn't mention the requirements on the
3125 <tt>value_type</tt>; it can be brought in line with the rest of 26.4
3126 [lib.numeric.ops] by adding the following to 26.4.4/2
3127 [lib.adjacent.difference]:
3128 </p>
3129
3130 <blockquote><p>
3131 The type of <tt>*first</tt> shall meet the requirements of
3132 <tt>CopyConstructible</tt> (20.1.3) and <tt>Assignable</tt> (23.1) types."
3133 </p></blockquote>
3134 <p><i>[
3135 Berlin: Giving output iterator's value_types very controversial. Suggestion of
3136 adding signatures to allow user to specify "accumulator".
3137 ]</i></p>
3138
3139
3140 <p><i>[
3141 Bellevue:
3142 ]</i></p>
3143
3144
3145 <blockquote>
3146 The intent of the algorithms is to perform their calculations using the type of the input iterator.
3147 Proposed wording provided.
3148 </blockquote>
3149
3150 <p><i>[
3151 Sophia Antipolis:
3152 ]</i></p>
3153
3154
3155 <blockquote>
3156 We did not agree that the proposed resolution was correct. For example,
3157 when the arguments are types <tt>(float*, float*, double*)</tt>, the
3158 highest-quality solution would use double as the type of the
3159 accumulator. If the intent of the wording is to require that the type of
3160 the accumulator must be the <tt>input_iterator</tt>'s <tt>value_type</tt>, the wording
3161 should specify it.
3162 </blockquote>
3163
3164 <p><i>[
3165 2009-05-09 Alisdair adds:
3166 ]</i></p>
3167
3168
3169 <blockquote>
3170 <p>
3171 Now that we have the facility, the 'best' accumulator type could probably be
3172 deduced as:
3173 </p>
3174 <blockquote><pre>std::common_type&lt;InIter::value_type, OutIter::reference&gt;::type
3175 </pre></blockquote>
3176 <p>
3177 This type would then have additional requirements of constructability and
3178 incrementability/assignability.
3179 </p>
3180 <p>
3181 If this extracting an accumulator type from a pair/set of iterators (with
3182 additional requirements on that type) is a problem for multiple functions,
3183 it might be worth extracting into a SharedAccumulator concept or similar.
3184 </p>
3185 <p>
3186 I'll go no further in writing up wording now, until the group gives a
3187 clearer indication of preferred direction.
3188 </p>
3189 </blockquote>
3190
3191 <p><i>[
3192 2009-07 Frankfurt
3193 ]</i></p>
3194
3195
3196 <blockquote>
3197 The proposed resolution isn't quite right. For example, "the type of
3198 *first" should be changed to "iterator::value_type" or similar. Daniel
3199 volunteered to correct the wording.
3200 </blockquote>
3201
3202 <p><i>[
3203 2009-07-29 Daniel corrected wording.
3204 ]</i></p>
3205
3206
3207 <p><i>[
3208 2009-10 Santa Cruz:
3209 ]</i></p>
3210
3211
3212 <blockquote>
3213 Move to Ready.
3214 </blockquote>
3215
3216
3217
3218 <p><b>Proposed resolution:</b></p>
3219
3220
3221
3222 <ol>
3223 <li>
3224 <p>
3225 Change 26.7.3 [partial.sum]/1 as indicated:
3226 </p>
3227
3228 <blockquote>
3229 <p>
3230 <i>Effects:</i> <ins>Let <tt>VT</tt> be <tt>InputIterator</tt>'s value type. For a nonempty range,
3231 initializes an accumulator <tt>acc</tt> of type <tt>VT</tt> with <tt>*first</tt> and performs
3232 <tt>*result = acc</tt>. For every iterator <tt>i</tt> in <tt>[first + 1, last)</tt> in order, <tt>acc</tt> is then
3233 modified by <tt>acc = acc + *i</tt> or <tt>acc = binary_op(acc, *i)</tt> and is assigned
3234 to <tt>*(result + (i - first))</tt>.</ins> <del>Assigns to every element referred to by
3235 iterator <tt>i</tt> in the range <tt>[result,result + (last - first))</tt> a value
3236 correspondingly
3237 equal to</del>
3238 </p>
3239
3240 <blockquote><pre><del>
3241 ((...(*first + *(first + 1)) + ...) + *(first + (i - result)))
3242 </del></pre></blockquote>
3243
3244 <p><del>
3245 or
3246 </del></p>
3247
3248 <blockquote><pre><del>
3249 binary_op(binary_op(...,
3250    binary_op(*first, *(first + 1)),...), *(first + (i - result)))
3251 </del></pre></blockquote>
3252 </blockquote>
3253 </li>
3254
3255 <li>
3256 <p>
3257 Change 26.7.3 [partial.sum]/3 as indicated:
3258 </p>
3259
3260 <blockquote>
3261 <i>Complexity:</i> Exactly <tt><ins>max(</ins>(last - first) - 1<ins>, 0)</ins></tt>
3262 applications
3263 of <tt><del>binary_op</del></tt><ins>the binary operation</ins>.
3264 </blockquote>
3265 </li>
3266
3267 <li>
3268 <p>
3269 Change 26.7.3 [partial.sum]/4 as indicated:
3270 </p>
3271
3272 <blockquote>
3273 <i>Requires:</i> <ins><tt>VT</tt> shall be constructible from the type of <tt>*first</tt>, the result of
3274 <tt>acc + *i</tt> or <tt>binary_op(acc, *i)</tt> shall be implicitly convertible to <tt>VT</tt>, and
3275 the result of the expression <tt>acc</tt> shall be writable to the <tt>result</tt>
3276 output iterator.</ins> In the ranges <tt>[first,last]</tt> and
3277 <tt>[result,result + (last - first)]</tt> [..]
3278 </blockquote>
3279 </li>
3280
3281 <li>
3282 <p>
3283 Change 26.7.4 [adjacent.difference]/1 as indicated:
3284 </p>
3285
3286 <blockquote>
3287 <p>
3288 <i>Effects:</i> <ins>Let <tt>VT</tt> be <tt>InputIterator</tt>'s value type. For a nonempty range,
3289 initializes an accumulator <tt>acc</tt> of type <tt>VT</tt> with <tt>*first</tt> and performs
3290 <tt>*result = acc</tt>. For every iterator <tt>i</tt> in <tt>[first + 1, last)</tt> in order,
3291 initializes a
3292 value <tt>val</tt> of type <tt>VT</tt> with <tt>*i</tt>, assigns the result of <tt>val - acc</tt> or
3293 <tt>binary_op(val, acc)</tt>
3294 to <tt>*(result + (i - first))</tt> and modifies <tt>acc = std::move(val)</tt>.</ins>
3295 <del>Assigns to every element referred to by iterator <tt>i</tt> in the range
3296 <tt>[result + 1,
3297 result + (last - first))</tt> a value correspondingly equal to</del>
3298 </p>
3299
3300 <blockquote><pre><del>
3301 *(first + (i - result)) - *(first + (i - result) - 1)
3302 </del></pre></blockquote>
3303
3304 <p><del>
3305 or
3306 </del></p>
3307
3308 <blockquote><pre><del>
3309 binary_op(*(first + (i - result)), *(first + (i - result) - 1)).
3310 </del></pre></blockquote>
3311
3312 <p><del>
3313 result gets the value of *first.</del>
3314 </p>
3315 </blockquote>
3316 </li>
3317
3318 <li>
3319 <p>
3320 Change 26.7.4 [adjacent.difference]/2 as indicated:
3321 </p>
3322
3323 <blockquote>
3324 <i>Requires:</i> <ins><tt>VT</tt> shall be <tt>MoveAssignable</tt> ([moveassignable])
3325 and shall be
3326 constructible from the type of <tt>*first</tt>. The result
3327 of the expression <tt>acc</tt> and the result of the expression <tt>val - acc</tt> or
3328 <tt>binary_op(val, acc)</tt>
3329 shall be writable to the <tt>result</tt> output iterator.</ins> In the ranges
3330 <tt>[first,last]</tt> [..]
3331 </blockquote>
3332 </li>
3333
3334 <li>
3335 <p>
3336 Change 26.7.4 [adjacent.difference]/5 as indicated:
3337 </p>
3338
3339 <blockquote>
3340 <i>Complexity:</i> Exactly <tt><ins>max(</ins>(last - first) - 1<ins>, 0)</ins></tt>
3341 applications
3342 of <del><tt>binary_op</tt></del><ins>the binary operation</ins>.
3343 </blockquote>
3344 </li>
3345 </ol>
3346
3347
3348
3349
3350
3351
3352
3353
3354 <hr>
3355 <h3><a name="556"></a>556. is Compare a BinaryPredicate?</h3>
3356 <p><b>Section:</b> 25.4 [alg.sorting] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
3357  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2006-02-05  <b>Last modified:</b> 2009-10-25</p>
3358 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#alg.sorting">issues</a> in [alg.sorting].</p>
3359 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
3360 <p><b>Discussion:</b></p>
3361 <p>
3362 In 25, p8 we allow BinaryPredicates to return a type that's convertible
3363 to bool but need not actually be bool. That allows predicates to return
3364 things like proxies and requires that implementations be careful about
3365 what kinds of expressions they use the result of the predicate in (e.g.,
3366 the expression in if (!pred(a, b)) need not be well-formed since the
3367 negation operator may be inaccessible or return a type that's not
3368 convertible to bool).
3369 </p>
3370 <p>
3371 Here's the text for reference:
3372 </p>
3373 <blockquote><p>
3374   ...if an algorithm takes BinaryPredicate binary_pred as its argument
3375  and first1 and first2 as its iterator arguments, it should work
3376  correctly in the construct if (binary_pred(*first1, first2)){...}.
3377 </p></blockquote>
3378
3379 <p>
3380 In 25.3, p2 we require that the Compare function object return true
3381 of false, which would seem to preclude such proxies. The relevant text
3382 is here:
3383 </p>
3384 <blockquote><p>
3385   Compare is used as a function object which returns true if the first
3386   argument is less than the second, and false otherwise...
3387 </p></blockquote>
3388
3389 <p><i>[
3390 Portland:  Jack to define "convertible to bool" such that short circuiting isn't
3391 destroyed.
3392 ]</i></p>
3393
3394
3395 <p><i>[
3396 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
3397 ]</i></p>
3398
3399
3400 <p><i>[
3401 2009-10 Santa Cruz:
3402 ]</i></p>
3403
3404
3405 <blockquote>
3406 Move to Review once wording received. Stefanus to send proposed wording.
3407 </blockquote>
3408
3409 <p><i>[
3410 2009-10 Santa Cruz:
3411 ]</i></p>
3412
3413
3414 <blockquote>
3415 Move to Review once wording received. Stefanus to send proposed wording.
3416 </blockquote>
3417
3418 <p><i>[
3419 2009-10-24 Stefanus supplied wording.
3420 ]</i></p>
3421
3422
3423 <blockquote>
3424 Move to Review once wording received. Stefanus to send proposed wording.
3425 Current proposed wording proposed here:
3426 <blockquote>
3427 <p>
3428 I think we could fix this by rewording 25.3, p2 to read somthing like:
3429 </p>
3430 <blockquote><p>
3431 -2- <tt>Compare</tt> is <del>used as a function object which returns
3432 <tt>true</tt> if the first argument</del> <ins>a <tt>BinaryPredicate</tt>. The
3433 return value of the function call operator applied to an object of type
3434 <tt>Compare</tt>, when converted to type <tt>bool</tt>, yields <tt>true</tt>
3435 if the first argument of the call</ins> is less than the second, and
3436 <tt>false</tt> otherwise. <tt>Compare <i>comp</i></tt> is used throughout for
3437 algorithms assuming an ordering relation. It is assumed that <tt><i>comp</i></tt>
3438 will not apply any non-constant function through the dereferenced iterator.
3439 </p></blockquote>
3440 </blockquote>
3441 </blockquote>
3442
3443
3444
3445 <p><b>Proposed resolution:</b></p>
3446 <p>
3447 Change 25.4 [alg.sorting] p2:
3448 </p>
3449 <blockquote>
3450 <tt>Compare</tt> is used as a function object<ins>. The return value of
3451 the function call operator applied to an object of type Compare, when
3452 converted to type bool, yields true if the first argument of the
3453 call</ins> <del>which returns <tt>true</tt> if the first argument</del>
3454 is less than the second, and <tt>false</tt> otherwise. <tt>Compare
3455 comp</tt> is used throughout for algorithms assuming an ordering
3456 relation. It is assumed that <tt>comp</tt> will not apply any
3457 non-constant function through the dereferenced iterator.
3458 </blockquote>
3459
3460
3461 <p><b>Rationale:</b></p>
3462 <p><i>[
3463 San Francisco:
3464 ]</i></p>
3465
3466
3467 <blockquote>
3468 Solved by
3469 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2774.pdf">N2774</a>).
3470 </blockquote>
3471
3472
3473
3474
3475
3476
3477 <hr>
3478 <h3><a name="594"></a>594. Disadvantages of defining Swappable in terms of CopyConstructible and Assignable</h3>
3479 <p><b>Section:</b> 20.2.1 [utility.arg.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
3480  <b>Submitter:</b> Niels Dekker <b>Opened:</b> 2006-11-02  <b>Last modified:</b> 2009-11-08</p>
3481 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#utility.arg.requirements">active issues</a> in [utility.arg.requirements].</p>
3482 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#utility.arg.requirements">issues</a> in [utility.arg.requirements].</p>
3483 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
3484 <p><b>Discussion:</b></p>
3485 <p>
3486 It seems undesirable to define the Swappable requirement in terms of
3487 CopyConstructible and Assignable requirements. And likewise, once the
3488 MoveConstructible and MoveAssignable requirements (N1860) have made it
3489 into the Working Draft, it seems undesirable to define the Swappable
3490 requirement in terms of those requirements. Instead, it appears
3491 preferable to have the Swappable requirement defined exclusively in
3492 terms of the existence of an appropriate swap function.
3493 </p>
3494 <p>
3495 Section 20.1.4 [lib.swappable] of the current Working Draft (N2009)
3496 says:
3497 </p>
3498 <blockquote><p>
3499 The Swappable requirement is met by satisfying one or more of the
3500 following conditions:</p>
3501 <ul>
3502 <li>
3503 T is Swappable if T satisfies the CopyConstructible requirements
3504 (20.1.3) and the Assignable requirements (23.1);
3505 </li>
3506 <li>
3507 T is Swappable if a namespace scope function named swap exists in the
3508 same namespace as the definition of T, such that the expression
3509 swap(t,u) is valid and has the semantics described in Table 33.
3510 </li>
3511 </ul>
3512 </blockquote>
3513 <p>
3514 I can think of three disadvantages of this definition:
3515 </p>
3516 <ol>
3517 <li>
3518 <p>
3519 If a client's type T satisfies the first condition (T is both
3520 CopyConstructible and Assignable), the client cannot stop T from
3521 satisfying the Swappable requirement without stopping T from
3522 satisfying the first condition.
3523 </p>
3524 <p>
3525 A client might want to stop T from satisfying the Swappable
3526 requirement, because swapping by means of copy construction and
3527 assignment might throw an exception, and she might find a throwing
3528 swap unacceptable for her type. On the other hand, she might not feel
3529 the need to fully implement her own swap function for this type. In
3530 this case she would want to be able to simply prevent algorithms that
3531 would swap objects of type T from being used, e.g., by declaring a
3532 swap function for T, and leaving this function purposely undefined.
3533 This would trigger a link error, if an attempt would be made to use
3534 such an algorithm for this type. For most standard library
3535 implementations, this practice would indeed have the effect of
3536 stopping T from satisfying the Swappable requirement.
3537 </p>
3538 </li>
3539 <li>
3540 <p>
3541 A client's type T that does not satisfy the first condition can not be
3542 made Swappable by providing a specialization of std::swap for T.
3543 </p>
3544 <p>
3545 While I'm aware about the fact that people have mixed feelings about
3546 providing a specialization of std::swap, it is well-defined to do so.
3547 It sounds rather counter-intuitive to say that T is not Swappable, if
3548 it has a valid and semantically correct specialization of std::swap.
3549 Also in practice, providing such a specialization will have the same
3550 effect as satisfying the Swappable requirement.
3551 </p>
3552 </li>
3553 <li>
3554 <p>
3555 For a client's type T that satisfies both conditions of the Swappable
3556 requirement, it is not specified which of the two conditions prevails.
3557 After reading section 20.1.4 [lib.swappable], one might wonder whether
3558 objects of T will be swapped by doing copy construction and
3559 assignments, or by calling the swap function of T.
3560 </p>
3561 <p>
3562 I'm aware that the intention of the Draft is to prefer calling the
3563 swap function of T over doing copy construction and assignments. Still
3564 in my opinion, it would be better to make this clear in the wording of
3565 the definition of Swappable.
3566 </p>
3567 </li>
3568 </ol>
3569 <p>
3570 I would like to have the Swappable requirement defined in such a way
3571 that the following code fragment will correctly swap two objects of a
3572 type T, if and only if T is Swappable:
3573 </p>
3574 <pre>   using std::swap;
3575    swap(t, u);  // t and u are of type T.
3576 </pre>
3577 <p>
3578 This is also the way Scott Meyers recommends calling a swap function,
3579 in Effective C++, Third Edition, item 25.
3580 </p>
3581 <p>
3582 Most aspects of this issue have been dealt with in a discussion on
3583 comp.std.c++ about the Swappable requirement, from 13 September to 4
3584 October 2006, including valuable input by David Abrahams, Pete Becker,
3585 Greg Herlihy, Howard Hinnant and others.
3586 </p>
3587
3588 <p><i>[
3589 San Francisco:
3590 ]</i></p>
3591
3592
3593 <blockquote>
3594 Recommend NAD.  Solved by
3595 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2774.pdf">N2774</a>.
3596 </blockquote>
3597
3598 <p><i>[
3599 2009-07 Frankfurt
3600 ]</i></p>
3601
3602
3603 <blockquote>
3604 Moved to Open.  Waiting for non-concepts draft.
3605 </blockquote>
3606
3607 <p><i>[
3608 2009-11-08 Howard adds:
3609 ]</i></p>
3610
3611
3612 <blockquote>
3613 This issue is very closely related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#742">742</a>.
3614 </blockquote>
3615
3616
3617 <p><b>Proposed resolution:</b></p>
3618 <p>
3619 Change section 20.1.4 [lib.swappable] as follows:
3620 </p>
3621 <blockquote><p>
3622 The Swappable requirement is met by satisfying
3623 <del>one or more of the following conditions:</del>
3624 <ins>the following condition:</ins></p>
3625 <ul>
3626
3627 <li>
3628 <del>T is Swappable if T satisfies the CopyConstructible requirements
3629 (20.1.3) and the Assignable requirements (23.1);</del>
3630 </li>
3631 <li>
3632 <del>
3633 T is Swappable if a namespace scope function named swap exists in the
3634 same namespace as the definition of T, such that the expression
3635 swap(t,u) is valid and has the semantics described in Table 33.
3636 </del>
3637 T is Swappable if an unqualified function call swap(t,u) is valid
3638 within the namespace std, and has the semantics described in Table 33.
3639 </li>
3640 </ul>
3641 </blockquote>
3642
3643
3644
3645
3646
3647 <hr>
3648 <h3><a name="625"></a>625. mixed up <i>Effects</i> and <i>Returns</i> clauses</h3>
3649 <p><b>Section:</b> 17 [library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
3650  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2007-01-20  <b>Last modified:</b> 2009-10-20</p>
3651 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
3652 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
3653 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
3654 <p><b>Duplicate of:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#895">895</a></p>
3655 <p><b>Discussion:</b></p>
3656  <p>
3657
3658 Many member functions of <code>basic_string</code> are overloaded,
3659 with some of the overloads taking a <code>string</code> argument,
3660 others <code>value_type*</code>, others <code>size_type</code>, and
3661 others still <code>iterators</code>. Often, the requirements on one of
3662 the overloads are expressed in the form of <i>Effects</i>,
3663 <i>Throws</i>, and in the Working Paper
3664 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2006/n2134.pdf">N2134</a>)
3665 also <i>Remark</i> clauses, while those on the rest of the overloads
3666 via a reference to this overload and using a <i>Returns</i> clause.
3667 </p>
3668
3669 <p>
3670 The difference between the two forms of specification is that per
3671 17.5.1.4 [structure.specifications], p3, an <i>Effects</i> clause specifies
3672 <i>"actions performed by the functions,"</i> i.e., its observable
3673 effects, while a <i>Returns</i> clause is <i>"a description of the
3674 return value(s) of a function"</i> that does not impose any
3675 requirements on the function's observable effects.
3676 </p>
3677
3678 <p>
3679 Since only <i>Notes</i> are explicitly defined to be informative and
3680 all other paragraphs are explicitly defined to be normative, like
3681 <i>Effects</i> and <i>Returns</i>, the new <i>Remark</i> clauses also
3682 impose normative requirements.
3683 </p>
3684
3685 <p>
3686 So by this strict reading of the standard there are some member
3687 functions of <code>basic_string</code> that are required to throw an
3688 exception under some conditions or use specific traits members while
3689 many other otherwise equivalent overloads, while obliged to return the
3690 same values, aren't required to follow the exact same requirements
3691 with regards to the observable effects.
3692 </p>
3693
3694 <p>
3695 Here's an example of this problem that was precipitated by the change
3696 from informative Notes to normative <i>Remark</i>s (presumably made to
3697 address <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#424">424</a>):
3698 </p>
3699
3700 <p>
3701 In the Working Paper, <code>find(string, size_type)</code> contains a
3702 <i>Remark</i> clause (which is just a <i>Note</i> in the current
3703 standard) requiring it to use <code>traits::eq()</code>.
3704 </p>
3705
3706 <p>
3707 <code>find(const charT *s, size_type pos)</code> is specified to
3708 return <code>find(string(s), pos)</code> by a <i>Returns</i> clause
3709 and so it is not required to use <code>traits::eq()</code>. However,
3710 the Working Paper has replaced the original informative <i>Note</i>
3711 about the function using <code>traits::length()</code> with a
3712 normative requirement in the form of a <i>Remark</i>. Calling
3713 <code>traits::length()</code> may be suboptimal, for example when the
3714 argument is a very long array whose initial substring doesn't appear
3715 anywhere in <code>*this</code>.
3716 </p>
3717
3718 <p>
3719 Here's another similar example, one that existed even prior to the
3720 introduction of <i>Remark</i>s:
3721 </p>
3722
3723 <p>
3724 <code> insert(size_type pos, string, size_type, size_type)</code> is
3725 required to throw <code>out_of_range</code> if <code>pos &gt;
3726 size()</code>.
3727 </p>
3728
3729 <p>
3730 <code>insert(size_type pos, string str)</code> is specified to return
3731 <code>insert(pos, str, 0, npos)</code> by a <i>Returns</i> clause and
3732 so its effects when <code>pos &gt; size()</code> are strictly speaking
3733 unspecified.
3734 </p><p>
3735
3736 </p>
3737 I believe a careful review of the current <i>Effects</i> and
3738 <i>Returns</i> clauses is needed in order to identify all such
3739 problematic cases. In addition, a review of the Working Paper should
3740 be done to make sure that the newly introduced normative <i>Remark</i>
3741 clauses do not impose any undesirable normative requirements in place
3742 of the original informative <i>Notes</i>.
3743
3744
3745 <p><i>[
3746 Batavia: Alan and Pete to work.
3747 ]</i></p>
3748
3749
3750 <p><i>[
3751 Bellevue: Marked as NAD Editorial.
3752 ]</i></p>
3753
3754
3755 <p><i>[
3756 Post-Sophia Antipolis:
3757 Martin indicates there is still work to be done on this issue.
3758 Reopened.
3759 ]</i></p>
3760
3761
3762 <p><i>[
3763 Batavia (2009-05):
3764 ]</i></p>
3765
3766 <blockquote>
3767 Tom proposes we say that, unless specified otherwise,
3768 it is always the caller's responsibility to verify that supplied arguments
3769 meet the called function's requirements.
3770 If further semantics are specified
3771 (e.g., that the function throws under certain conditions),
3772 then it is up to the implementer to check those conditions.
3773 Alan feels strongly that our current use of Requires in this context
3774 is confusing, especially now that <tt>requires</tt> is a new keyword.
3775 </blockquote>
3776
3777 <p><i>[
3778 2009-07 Frankfurt
3779 ]</i></p>
3780
3781
3782 <blockquote>
3783 Move to Tentatively NAD.
3784 </blockquote>
3785
3786 <p><i>[
3787 2009 Santa Cruz:
3788 ]</i></p>
3789
3790
3791 <blockquote>
3792 Move to Open.  Martin will work on proposed wording.
3793 </blockquote>
3794
3795
3796
3797 <p><b>Proposed resolution:</b></p>
3798 <p>
3799 </p>
3800
3801
3802
3803
3804
3805 <hr>
3806 <h3><a name="631"></a>631. conflicting requirements for <tt>BinaryPredicate</tt></h3>
3807 <p><b>Section:</b> 25 [algorithms] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
3808  <b>Submitter:</b> James Kanze <b>Opened:</b> 2007-01-31  <b>Last modified:</b> 2009-10-24</p>
3809 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#algorithms">active issues</a> in [algorithms].</p>
3810 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#algorithms">issues</a> in [algorithms].</p>
3811 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
3812 <p><b>Discussion:</b></p>
3813 <p>
3814 The general requirements for <tt><tt>BinaryPredicate</tt></tt> (in 25 [algorithms]/8) contradict the implied specific requirements for
3815 some functions. In particular, it says that:
3816 </p>
3817
3818 <blockquote><p>
3819 [...] if an algorithm takes <tt>BinaryPredicate <i>binary_pred</i></tt>
3820 as its argument and <tt><i>first1</i></tt> and <i>first2</i> as its
3821 iterator arguments, it should work correctly in the construct <tt>if
3822 (binary_pred (*<i>first1</i> , *<i>first2</i> )){...}</tt>.
3823 <tt>BinaryPredicate</tt> always takes the first iterator type as its
3824 first argument, that is, in those cases when <tt>T <i>value</i></tt> is
3825 part of the signature, it should work correctly in the context of <tt>if
3826 (binary_pred (*<i>first1</i> , <i>value</i>)){...}</tt>.
3827 </p></blockquote>
3828
3829 <p>
3830 In the description of <tt>upper_bound</tt> (25.4.3.2 [upper.bound]/2), however, the use is described as
3831 "<tt>!comp(<i>value</i>, <i>e</i>)</tt>", where <tt><i>e</i></tt> is an
3832 element of the sequence (a result of dereferencing
3833 <tt>*<i>first</i></tt>).
3834 </p>
3835
3836 <p>
3837 In the description of <tt>lexicographical_compare</tt>, we have both
3838 "<tt>*<i>first1</i> &lt; *<i>first2</i></tt>" and "<tt>*<i>first2</i>
3839 &lt; *<i>first1</i></tt>" (which presumably implies "<tt>comp(
3840 *<i>first1</i>, *<i>first2</i> )</tt>" and "<tt>comp( *<i>first2</i>,
3841 *<i>first1</i> )</tt>".
3842 </p>
3843
3844 <p><i>[
3845 Toronto:  Moved to Open.  ConceptGCC seems to get <tt>lower_bound</tt>
3846 and <tt>upper_bound</tt> to work withoutt these changes.
3847 ]</i></p>
3848
3849
3850 <p><i>[
3851 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
3852 ]</i></p>
3853
3854
3855 <p><i>[
3856 2009-10 Santa Cruz:
3857 ]</i></p>
3858
3859
3860 <blockquote>
3861 Move to Review. The small problem with the "iterator type"
3862 will be fixed. The cited functions (<tt>lower_bound</tt>, <tt>uppwer_bound</tt>,
3863 <tt>equal_range</tt>) don't actually use <tt>BinaryPredicate</tt> , and where it is used,
3864 it is consistent with  [algorithm]/8, so the main complaint of the issue
3865 is moot.
3866 </blockquote>
3867
3868
3869
3870 <p><b>Proposed resolution:</b></p>
3871 <p>
3872 Logically, the <tt>BinaryPredicate</tt> is used as an ordering
3873 relationship, with the semantics of "less than".  Depending on the
3874 function, it may be used to determine equality, or any of the inequality
3875 relationships; doing this requires being able to use it with either
3876 parameter first.  I would thus suggest that the requirement be:
3877 </p>
3878
3879 <blockquote><p>
3880 [...] <tt>BinaryPredicate</tt> always takes the first iterator
3881 <tt>value_type</tt> as one of its arguments, it is unspecified which. If
3882 an algorithm takes <tt>BinaryPredicate <i>binary_pred</i></tt> as its
3883 argument and <tt><i><i>first1</i></i></tt> and <i>first2</i> as its
3884 iterator arguments, it should work correctly both in the construct
3885 <tt>if (binary_pred (*<i>first1</i> , *<i>first2</i> )){...}</tt> and
3886 <tt>if (binary_pred (*<i>first2</i>, *<i>first1</i>)){...}</tt>.  In
3887 those cases when <tt>T <i>value</i></tt> is part of the signature, it
3888 should work correctly in the context of <tt>if (binary_pred
3889 (*<i>first1</i> , <i>value</i>)){...}</tt> and of <tt>if (binary_pred
3890 (<i>value</i>, *<i>first1</i>)){...}</tt>. [<i>Note:</i> if the two
3891 types are not identical, and neither is convertable to the other, this
3892 may require that the <tt>BinaryPredicate</tt> be a functional object
3893 with two overloaded <tt>operator()()</tt> functions. <i>--end note</i>]
3894 </p></blockquote>
3895
3896 <p>
3897 Alternatively, one could specify an order for each function. IMHO, this
3898 would be more work for the committee, more work for the implementors,
3899 and of no real advantage for the user: some functions, such as
3900 <tt>lexicographical_compare</tt> or <tt>equal_range</tt>, will still require both
3901 functions, and it seems like a much easier rule to teach that both
3902 functions are always required, rather than to have a complicated list of
3903 when you only need one, and which one.
3904 </p>
3905
3906
3907 <p><b>Rationale:</b></p>
3908 <p><i>[
3909 post San Francisco:
3910 ]</i></p>
3911
3912
3913 <blockquote>
3914 Solved by
3915 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2759.pdf">N2759</a>.
3916 </blockquote>
3917
3918
3919
3920
3921
3922
3923 <hr>
3924 <h3><a name="671"></a>671. precision of hexfloat</h3>
3925 <p><b>Section:</b> 22.4.2.2.2 [facet.num.put.virtuals] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
3926  <b>Submitter:</b> John Salmon <b>Opened:</b> 2007-04-20  <b>Last modified:</b> 2009-10-21</p>
3927 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#facet.num.put.virtuals">active issues</a> in [facet.num.put.virtuals].</p>
3928 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#facet.num.put.virtuals">issues</a> in [facet.num.put.virtuals].</p>
3929 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
3930 <p><b>Discussion:</b></p>
3931 <p>
3932 I am trying to understand how TR1 supports hex float (%a) output.
3933 </p>
3934 <p>
3935 As far as I can tell, it does so via the following:
3936 </p>
3937 <p>
3938 8.15 Additions to header &lt;locale&gt; [tr.c99.locale]
3939 </p>
3940 <p>
3941 In subclause 22.4.2.2.2 [facet.num.put.virtuals], Table 58 Floating-point conversions, after
3942 the line:
3943 floatfield == ios_base::scientific %E
3944 </p>
3945 <p>
3946 add the two lines:
3947 </p>
3948 <blockquote><pre>floatfield == ios_base::fixed | ios_base::scientific &amp;&amp; !uppercase %a
3949 floatfield == ios_base::fixed | ios_base::scientific %A 2
3950 </pre></blockquote>
3951 <p>
3952 [Note: The additional requirements on print and scan functions, later
3953 in this clause, ensure that the print functions generate hexadecimal
3954 floating-point fields with a %a or %A conversion specifier, and that
3955 the scan functions match hexadecimal floating-point fields with a %g
3956 conversion specifier.  end note]
3957 </p>
3958 <p>
3959 Following the thread, in 22.4.2.2.2 [facet.num.put.virtuals], we find:
3960 </p>
3961 <p>
3962 For conversion from a floating-point type, if (flags &amp; fixed) != 0 or
3963 if str.precision() &gt; 0, then str.precision() is specified in the
3964 conversion specification.
3965 </p>
3966 <p>
3967 This would seem to imply that when floatfield == fixed|scientific, the
3968 precision of the conversion specifier is to be taken from
3969 str.precision().  Is this really what's intended?  I sincerely hope
3970 that I'm either missing something or this is an oversight.  Please
3971 tell me that the committee did not intend to mandate that hex floats
3972 (and doubles) should by default be printed as if by %.6a.
3973 </p>
3974
3975 <p><i>[
3976 Howard: I think the fundamental issue we overlooked was that with %f,
3977 %e, %g, the default precision was always 6.  With %a the default
3978 precision is not 6, it is infinity.  So for the first time, we need to
3979 distinguish between the default value of precision, and the precision
3980 value 6.
3981 ]</i></p>
3982
3983
3984 <p><i>[
3985 2009-07 Frankfurt
3986 ]</i></p>
3987
3988
3989 <blockquote>
3990 <p>
3991 Leave this open for Robert and Daniel to work on.
3992 </p>
3993 <p>
3994 Straw poll: Disposition?
3995 </p>
3996 <ul>
3997 <li>Default is %.6a (i.e. NAD): 2</li>
3998 <li>Always %a (no precision): 6</li>
3999 <li>precision(-1) == %a: 3</li>
4000 </ul>
4001 <p>
4002 Daniel and Robert have direction to write up wording for the "always %a" solution.
4003 </p>
4004
4005 <p><i>[
4006 2009-07-15 Robert provided wording.
4007 ]</i></p>
4008
4009 </blockquote>
4010
4011 <p><i>[
4012 2009-10 Santa Cruz:
4013 ]</i></p>
4014
4015
4016 <blockquote>
4017 Move to Ready.
4018 </blockquote>
4019
4020
4021
4022 <p><b>Proposed resolution:</b></p>
4023 <p>
4024 Change 22.4.2.2.2 [facet.num.put.virtuals], Stage 1, under p5 (near the end
4025 of Stage 1):
4026 </p>
4027
4028 <blockquote>
4029 For conversion from a floating-point type, <tt>str.precision()</tt> is specified
4030 <ins>as precision</ins> in the conversion specification
4031 <ins>if <tt>floatfield != (ios_base::fixed | ios_base::scientific)</tt>, else no
4032 precision is specified</ins>.
4033 </blockquote>
4034
4035
4036
4037 <p><i>[
4038 Kona (2007): Robert volunteers to propose wording.
4039 ]</i></p>
4040
4041
4042
4043
4044
4045 <hr>
4046 <h3><a name="676"></a>676. Moving the unordered containers</h3>
4047 <p><b>Section:</b> 23.5 [unord] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
4048  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2007-05-05  <b>Last modified:</b> 2009-10-29</p>
4049 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord">issues</a> in [unord].</p>
4050 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
4051 <p><b>Discussion:</b></p>
4052 <p>
4053 Move semantics are missing from the <tt>unordered</tt> containers.  The proposed
4054 resolution below adds move-support consistent with
4055 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2005/n1858.html">N1858</a>
4056 and the current working draft.
4057 </p>
4058
4059 <p>
4060 The current proposed resolution simply lists the requirements for each function.
4061 These might better be hoisted into the requirements table for unordered associative containers.
4062 Futhermore a mild reorganization of the container requirements could well be in order.
4063 This defect report is purposefully ignoring these larger issues and just focusing
4064 on getting the unordered containers "moved".
4065 </p>
4066
4067 <p><i>[
4068 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
4069 ]</i></p>
4070
4071
4072 <p><i>[
4073 2009-10-17 Removed rvalue-swaps from wording.
4074 ]</i></p>
4075
4076
4077 <p><i>[
4078 2009-10 Santa Cruz:
4079 ]</i></p>
4080
4081
4082 <blockquote>
4083 Move to Review. Alisdair will review proposed wording.
4084 </blockquote>
4085
4086 <p><i>[
4087 2009-10-29 Daniel updates wording.
4088 ]</i></p>
4089
4090
4091
4092
4093 <p><b>Proposed resolution:</b></p>
4094
4095 <p><b><tt>unordered_map</tt></b></p>
4096
4097 <p>
4098 Change 23.5.1 [unord.map]:
4099 </p>
4100
4101 <blockquote><pre>class unordered_map
4102 {
4103     ...
4104     unordered_map(const unordered_map&amp;);
4105     <ins>unordered_map(unordered_map&amp;&amp;);</ins>
4106     ~unordered_map();
4107     unordered_map&amp; operator=(const unordered_map&amp;);
4108     <ins>unordered_map&amp; operator=(unordered_map&amp;&amp;);</ins>
4109     ...
4110     // modifiers
4111     <del>std::</del>pair&lt;iterator, bool&gt; insert(const value_type&amp; obj);
4112     <ins>template &lt;class P&gt; pair&lt;iterator, bool&gt; insert(P&amp;&amp; obj);</ins>
4113     iterator       insert(const_iterator hint, const value_type&amp; obj);
4114     <ins>template &lt;class P&gt; iterator       insert(const_iterator hint, P&amp;&amp; obj);</ins>
4115     ...
4116     mapped_type&amp; operator[](const key_type&amp; k);
4117     <ins>mapped_type&amp; operator[](key_type&amp;&amp; k);</ins>
4118     ...
4119 };
4120
4121 </pre></blockquote>
4122
4123 <p>
4124 Add to 23.5.1.1 [unord.map.cnstr]:
4125 </p>
4126
4127 <blockquote>
4128 <pre>template &lt;class InputIterator&gt;
4129   unordered_map(InputIterator f, InputIterator l,
4130                 size_type n = <i>implementation-defined</i>,
4131                 const hasher&amp; hf = hasher(),
4132                 const key_equal&amp; eql = key_equal(),
4133                 const allocator_type&amp; a = allocator_type());
4134 </pre>
4135
4136 <blockquote><p>
4137 <ins>
4138 <i>Requires:</i> If the iterator's dereference operator returns an
4139 lvalue or a const rvalue <tt>pair&lt;key_type, mapped_type&gt;</tt>,
4140 then both <tt>key_type</tt> and <tt>mapped_type</tt> shall be
4141 <tt>CopyConstructible</tt>.
4142 </ins>
4143 </p></blockquote>
4144 </blockquote>
4145
4146 <p>
4147 Add to 23.5.1.2 [unord.map.elem]:
4148 </p>
4149
4150 <blockquote>
4151
4152 <pre>mapped_type&amp; operator[](const key_type&amp; k);</pre>
4153
4154 <blockquote>
4155 <p>...</p>
4156 <p><ins>
4157 <i>Requires:</i> <tt>key_type</tt> shall be <tt>CopyConstructible</tt>
4158 and <tt>mapped_type</tt> shall be <tt>DefaultConstructible</tt>.
4159 </ins></p>
4160 </blockquote>
4161
4162 <pre><ins>mapped_type&amp; operator[](key_type&amp;&amp; k);</ins></pre>
4163
4164 <blockquote>
4165 <p><ins>
4166 <i>Effects:</i> If the <tt>unordered_map</tt> does not already contain an
4167 element whose key is equivalent to <tt>k</tt> , inserts the value
4168 <tt>std::pair&lt;const key_type, mapped_type&gt;(std::move(k), mapped_type())</tt>.
4169 </ins></p>
4170
4171 <p><ins>
4172 <i>Requires:</i> <tt>mapped_type</tt> shall be <tt>DefaultConstructible</tt>.
4173 </ins></p>
4174
4175 <p><ins>
4176 <i>Returns:</i> A reference to <tt>x.second</tt>, where <tt>x</tt> is the
4177 (unique) element whose key is equivalent to <tt>k</tt>.
4178 </ins></p>
4179
4180 </blockquote>
4181
4182 </blockquote>
4183
4184 <p>
4185 Add new section [unord.map.modifiers]:
4186 </p>
4187
4188 <blockquote>
4189 <pre><ins>pair&lt;iterator, bool&gt; insert(const value_type&amp; x);</ins>
4190 <ins>template &lt;class P&gt; pair&lt;iterator, bool&gt; insert(P&amp;&amp; x);</ins>
4191 <ins>iterator       insert(const_iterator hint, const value_type&amp; x);</ins>
4192 <ins>template &lt;class P&gt; iterator       insert(const_iterator hint, P&amp;&amp; x);</ins>
4193 <ins>template &lt;class InputIterator&gt;
4194   void insert(InputIterator first, InputIterator last);</ins>
4195 </pre>
4196
4197 <blockquote>
4198 <p><ins>
4199 <i>Requires:</i> Those signatures taking a <tt>const value_type&amp;</tt> parameter
4200 requires both the <tt>key_type</tt> and the <tt>mapped_type</tt> to be
4201 <tt>CopyConstructible</tt>.
4202  If <tt>P</tt> is instantiated as a reference
4203 type, then the argument <tt>x</tt> is copied from. Otherwise <tt>x</tt>
4204 is considered to be an rvalue as it is converted to <tt>value_type</tt>
4205 and inserted into the <tt>unordered_map</tt>. Specifically, in such
4206 cases <tt>CopyConstructible</tt> is not required for <tt>key_type</tt> or
4207 <tt>mapped_type</tt> unless the conversion from <tt>P</tt> specifically
4208 requires it (e.g. if <tt>P</tt> is a <tt>tuple&lt;const key_type,
4209 mapped_type&gt;</tt>, then <tt>key_type</tt> must be
4210 <tt>CopyConstructible</tt>).
4211 </ins></p>
4212
4213 <p><ins>
4214 The signature taking <tt>InputIterator</tt>
4215 parameters requires <tt>CopyConstructible</tt> of both
4216 <tt>key_type</tt> and <tt>mapped_type</tt> if the dereferenced
4217 <tt>InputIterator</tt> returns an lvalue or <tt>const</tt> rvalue
4218 <tt>value_type</tt>.
4219 </ins></p>
4220
4221 </blockquote>
4222
4223 </blockquote>
4224
4225 <p><b><tt>unordered_multimap</tt></b></p>
4226
4227 <p>
4228 Change 23.5.2 [unord.multimap]:
4229 </p>
4230
4231 <blockquote><pre>class unordered_multimap
4232 {
4233     ...
4234     unordered_multimap(const unordered_multimap&amp;);
4235     <ins>unordered_multimap(unordered_multimap&amp;&amp;);</ins>
4236     ~unordered_multimap();
4237     unordered_multimap&amp; operator=(const unordered_multimap&amp;);
4238     <ins>unordered_multimap&amp; operator=(unordered_multimap&amp;&amp;);</ins>
4239     ...
4240     // modifiers
4241     iterator insert(const value_type&amp; obj);
4242     <ins>template &lt;class P&gt; iterator insert(P&amp;&amp; obj);</ins>
4243     iterator       insert(const_iterator hint, const value_type&amp; obj);
4244     <ins>template &lt;class P&gt; iterator       insert(const_iterator hint, P&amp;&amp; obj);</ins>
4245     ...
4246 };
4247
4248 </pre></blockquote>
4249
4250 <p>
4251 Add to 23.5.2.1 [unord.multimap.cnstr]:
4252 </p>
4253
4254 <blockquote>
4255 <pre>template &lt;class InputIterator&gt;
4256   unordered_multimap(InputIterator f, InputIterator l,
4257                 size_type n = <i>implementation-defined</i>,
4258                 const hasher&amp; hf = hasher(),
4259                 const key_equal&amp; eql = key_equal(),
4260                 const allocator_type&amp; a = allocator_type());
4261 </pre>
4262
4263 <blockquote><p>
4264 <ins>
4265 <i>Requires:</i> If the iterator's dereference operator returns an
4266 lvalue or a const rvalue <tt>pair&lt;key_type, mapped_type&gt;</tt>,
4267 then both <tt>key_type</tt> and <tt>mapped_type</tt> shall be
4268 <tt>CopyConstructible</tt>.
4269 </ins>
4270 </p></blockquote>
4271 </blockquote>
4272
4273 <p>
4274 Add new section [unord.multimap.modifiers]:
4275 </p>
4276
4277 <blockquote>
4278 <pre><ins>iterator insert(const value_type&amp; x);</ins>
4279 <ins>template &lt;class P&gt; iterator       insert(P&amp;&amp; x);</ins>
4280 <ins>iterator       insert(const_iterator hint, const value_type&amp; x);</ins>
4281 <ins>template &lt;class P&gt; iterator       insert(const_iterator hint, P&amp;&amp; x);</ins>
4282 <ins>template &lt;class InputIterator&gt;
4283   void insert(InputIterator first, InputIterator last);</ins>
4284 </pre>
4285
4286 <blockquote>
4287 <p><ins>
4288 <i>Requires:</i> Those signatures taking a <tt>const value_type&amp;</tt> parameter
4289 requires both the <tt>key_type</tt> and the <tt>mapped_type</tt> to be
4290 <tt>CopyConstructible</tt>.
4291 If <tt>P</tt> is instantiated as a reference
4292 type, then the argument <tt>x</tt> is copied from. Otherwise <tt>x</tt>
4293 is considered to be an rvalue as it is converted to <tt>value_type</tt>
4294 and inserted into the <tt>unordered_multimap</tt>. Specifically, in such
4295 cases <tt>CopyConstructible</tt> is not required for <tt>key_type</tt> or
4296 <tt>mapped_type</tt> unless the conversion from <tt>P</tt> specifically
4297 requires it (e.g. if <tt>P</tt> is a <tt>tuple&lt;const key_type,
4298 mapped_type&gt;</tt>, then <tt>key_type</tt> must be
4299 <tt>CopyConstructible</tt>).
4300 </ins></p>
4301
4302 <p><ins>
4303 The signature taking <tt>InputIterator</tt>
4304 parameters requires <tt>CopyConstructible</tt> of both
4305 <tt>key_type</tt> and <tt>mapped_type</tt> if the dereferenced
4306 <tt>InputIterator</tt> returns an lvalue or <tt>const</tt> rvalue
4307 <tt>value_type</tt>.
4308 </ins></p>
4309 </blockquote>
4310
4311 </blockquote>
4312
4313 <p><b><tt>unordered_set</tt></b></p>
4314
4315 <p>
4316 Change 23.5.3 [unord.set]:
4317 </p>
4318
4319 <blockquote><pre>class unordered_set
4320 {
4321     ...
4322     unordered_set(const unordered_set&amp;);
4323     <ins>unordered_set(unordered_set&amp;&amp;);</ins>
4324     ~unordered_set();
4325     unordered_set&amp; operator=(const unordered_set&amp;);
4326     <ins>unordered_set&amp; operator=(unordered_set&amp;&amp;);</ins>
4327     ...
4328     // modifiers
4329     <del>std::</del>pair&lt;iterator, bool&gt; insert(const value_type&amp; obj);
4330     <ins>pair&lt;iterator, bool&gt; insert(value_type&amp;&amp; obj);</ins>
4331     iterator       insert(const_iterator hint, const value_type&amp; obj);
4332     <ins>iterator       insert(const_iterator hint, value_type&amp;&amp; obj);</ins>
4333     ...
4334 };
4335 </pre></blockquote>
4336
4337 <p>
4338 Add to 23.5.3.1 [unord.set.cnstr]:
4339 </p>
4340
4341 <blockquote>
4342 <pre>template &lt;class InputIterator&gt;
4343   unordered_set(InputIterator f, InputIterator l,
4344                 size_type n = <i>implementation-defined</i>,
4345                 const hasher&amp; hf = hasher(),
4346                 const key_equal&amp; eql = key_equal(),
4347                 const allocator_type&amp; a = allocator_type());
4348 </pre>
4349
4350 <blockquote><p>
4351 <ins>
4352 <i>Requires:</i> If the iterator's dereference operator returns an
4353 lvalue or a const rvalue <tt>value_type</tt>, then the
4354 <tt>value_type</tt> shall be <tt>CopyConstructible</tt>.
4355 </ins>
4356 </p></blockquote>
4357 </blockquote>
4358
4359 <p>
4360 Add new section [unord.set.modifiers]:
4361 </p>
4362
4363 <blockquote>
4364 <pre><ins>pair&lt;iterator, bool&gt; insert(const value_type&amp; x);</ins>
4365 <ins>pair&lt;iterator, bool&gt; insert(value_type&amp;&amp; x);</ins>
4366 <ins>iterator       insert(const_iterator hint, const value_type&amp; x);</ins>
4367 <ins>iterator       insert(const_iterator hint, value_type&amp;&amp; x);</ins>
4368 <ins>template &lt;class InputIterator&gt;
4369   void insert(InputIterator first, InputIterator last);</ins>
4370 </pre>
4371
4372 <blockquote>
4373
4374 <p><ins>
4375 <i>Requires:</i> Those signatures taking a <tt>const
4376 value_type&amp;</tt> parameter requires the <tt>value_type</tt> to
4377 be <tt>CopyConstructible</tt>.
4378 </ins></p>
4379
4380 <p><ins>
4381 The signature taking <tt>InputIterator</tt> parameters requires
4382 <tt>CopyConstructible</tt> of <tt>value_type</tt> if the dereferenced
4383 <tt>InputIterator</tt> returns an lvalue or <tt>const</tt> rvalue
4384 <tt>value_type</tt>.
4385 </ins></p>
4386
4387 </blockquote>
4388
4389 </blockquote>
4390
4391 <p><b><tt>unordered_multiset</tt></b></p>
4392
4393 <p>
4394 Change 23.5.4 [unord.multiset]:
4395 </p>
4396
4397 <blockquote><pre>class unordered_multiset
4398 {
4399     ...
4400     unordered_multiset(const unordered_multiset&amp;);
4401     <ins>unordered_multiset(unordered_multiset&amp;&amp;);</ins>
4402     ~unordered_multiset();
4403     unordered_multiset&amp; operator=(const unordered_multiset&amp;);
4404     <ins>unordered_multiset&amp; operator=(unordered_multiset&amp;&amp;);</ins>
4405     ...
4406     // modifiers
4407     iterator insert(const value_type&amp; obj);
4408     <ins>iterator insert(value_type&amp;&amp; obj);</ins>
4409     iterator       insert(const_iterator hint, const value_type&amp; obj);
4410     <ins>iterator       insert(const_iterator hint, value_type&amp;&amp; obj);</ins>
4411     ...
4412 };
4413
4414 </pre></blockquote>
4415
4416 <p>
4417 Add to 23.5.4.1 [unord.multiset.cnstr]:
4418 </p>
4419
4420 <blockquote>
4421 <pre>template &lt;class InputIterator&gt;
4422   unordered_multiset(InputIterator f, InputIterator l,
4423                 size_type n = <i>implementation-defined</i>,
4424                 const hasher&amp; hf = hasher(),
4425                 const key_equal&amp; eql = key_equal(),
4426                 const allocator_type&amp; a = allocator_type());
4427 </pre>
4428
4429 <blockquote><p>
4430 <ins>
4431 <i>Requires:</i> If the iterator's dereference operator returns an
4432 lvalue or a const rvalue <tt>value_type</tt>, then the
4433 <tt>value_type</tt> shall be <tt>CopyConstructible</tt>.
4434 </ins>
4435 </p></blockquote>
4436 </blockquote>
4437
4438 <p>
4439 Add new section [unord.multiset.modifiers]:
4440 </p>
4441
4442 <blockquote>
4443 <pre><ins>iterator insert(const value_type&amp; x);</ins>
4444 <ins>iterator insert(value_type&amp;&amp; x);</ins>
4445 <ins>iterator       insert(const_iterator hint, const value_type&amp; x);</ins>
4446 <ins>iterator       insert(const_iterator hint, value_type&amp;&amp; x);</ins>
4447 <ins>template &lt;class InputIterator&gt;
4448   void insert(InputIterator first, InputIterator last);</ins>
4449 </pre>
4450
4451 <blockquote>
4452
4453 <p><ins>
4454 <i>Requires:</i> Those signatures taking a <tt>const
4455 value_type&amp;</tt> parameter requires the <tt>value_type</tt> to
4456 be <tt>CopyConstructible</tt>.
4457 </ins></p>
4458
4459 <p><ins>
4460 The signature taking <tt>InputIterator</tt> parameters requires
4461 <tt>CopyConstructible</tt> of <tt>value_type</tt> if the dereferenced
4462 <tt>InputIterator</tt> returns an lvalue or <tt>const</tt> rvalue
4463 <tt>value_type</tt>.
4464 </ins></p>
4465
4466 </blockquote>
4467
4468 </blockquote>
4469
4470
4471
4472 <p><i>[
4473 Voted to WP in Bellevue.
4474 ]</i></p>
4475
4476
4477 <p><i>[
4478 post Bellevue, Pete notes:
4479 ]</i></p>
4480
4481
4482 <blockquote>
4483 <p>
4484 Please remind people who are reviewing issues to check that the text
4485 modifications match the current draft. Issue 676, for example, adds two
4486 overloads for unordered_map::insert taking a hint. One takes a
4487 const_iterator and returns a const_iterator, and the other takes an
4488 iterator and returns an iterator. This was correct at the time the issue
4489 was written, but was changed in Toronto so there is only one hint
4490 overload, taking a const_iterator and returning an iterator.
4491 </p>
4492 <p>
4493 This issue is not ready. In addition to the relatively minor signature
4494 problem I mentioned earlier, it puts requirements in the wrong places.
4495 Instead of duplicating requirements throughout the template
4496 specifications, it should put them in the front matter that talks about
4497 requirements for unordered containers in general. This presentation
4498 problem is editorial, but I'm not willing to do the extensive rewrite
4499 that it requires. Please put it back into Open status.
4500 </p>
4501 </blockquote>
4502
4503 <p><b>Rationale:</b></p>
4504 <p><i>[
4505 San Francisco:
4506 ]</i></p>
4507
4508
4509 <blockquote>
4510 Solved by
4511 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2776.pdf">N2776</a>.
4512 </blockquote>
4513
4514
4515
4516
4517
4518
4519 <hr>
4520 <h3><a name="704"></a>704. MoveAssignable requirement for container value type overly strict</h3>
4521 <p><b>Section:</b> 23.2 [container.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
4522  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2007-05-20  <b>Last modified:</b> 2009-10-20</p>
4523 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.requirements">active issues</a> in [container.requirements].</p>
4524 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.requirements">issues</a> in [container.requirements].</p>
4525 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
4526 <p><b>Discussion:</b></p>
4527 <p>
4528 The move-related changes inadvertently overwrote the intent of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#276">276</a>.
4529 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#276">276</a> removed the requirement of <tt>CopyAssignable</tt> from
4530 most of the member functions of node-based containers.  But the move-related changes
4531 unnecessarily introduced the <tt>MoveAssignable</tt> requirement for those members which used to
4532 require <tt>CopyAssignable</tt>.
4533 </p>
4534
4535 <p>
4536 We also discussed (c++std-lib-18722) the possibility of dropping <tt>MoveAssignable</tt>
4537 from some of the sequence requirements.  Additionally the <i>in-place</i> construction
4538 work may further reduce requirements.  For purposes of an easy reference, here are the
4539 minimum sequence requirements as I currently understand them.  Those items in requirements
4540 table in the working draft which do not appear below have been purposefully omitted for
4541 brevity as they do not have any requirements of this nature.  Some items which do not
4542 have any requirements of this nature are included below just to confirm that they were
4543 not omitted by mistake.
4544 </p>
4545
4546 <table border="1">
4547 <caption>Container Requirements</caption>
4548 <tbody><tr><td><tt>X u(a)</tt></td><td><tt>value_type</tt> must be <tt>CopyConstructible</tt></td></tr>
4549 <tr><td><tt>X u(rv)</tt></td><td><tt>array</tt> and containers with a <tt>propagate_never</tt> allocator require <tt>value_type</tt> to be <tt>MoveConstructible</tt></td></tr>
4550 <tr><td><tt>a = u</tt></td><td>Sequences require <tt>value_type</tt> to be <tt>CopyConstructible</tt> and <tt>CopyAssignable</tt>.
4551                                Associative containers require <tt>value_type</tt> to be <tt>CopyConstructible</tt>.</td></tr>
4552 <tr><td><tt>a = rv</tt></td><td><tt>array</tt> requires <tt>value_type</tt> to be <tt>MoveAssignable</tt>.
4553                                 Sequences and Associative containers with <tt>propagate_never</tt> and <tt>propagate_on_copy_construction</tt> allocators require <tt>value_type</tt> to be <tt>MoveConstructible</tt>.</td></tr>
4554 <tr><td><tt>swap(a,u)</tt></td><td><tt>array</tt> and containers with <tt>propagate_never</tt> and
4555                   <tt>propagate_on_copy_construction</tt> allocators require <tt>value_type</tt> to be <tt>Swappable</tt>.</td></tr>
4556 </tbody></table>
4557
4558 <p>
4559 </p>
4560
4561 <table border="1">
4562 <caption>Sequence Requirements</caption>
4563 <tbody><tr><td><tt>X(n)</tt></td><td><tt>value_type</tt> must be <tt>DefaultConstructible</tt></td></tr>
4564 <tr><td><tt>X(n, t)</tt></td><td><tt>value_type</tt> must be <tt>CopyConstructible</tt></td></tr>
4565 <tr><td><tt>X(i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4566                                  If the iterators return an rvalue the <tt>value_type</tt> must be <tt>MoveConstructible</tt>.</td></tr>
4567 <tr><td><tt>a.insert(p, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4568                                         The sequences <tt>vector</tt> and <tt>deque</tt> also require the <tt>value_type</tt> to be <tt>CopyAssignable</tt>.</td></tr>
4569 <tr><td><tt>a.insert(p, rv)</tt></td><td>The <tt>value_type</tt> must be <tt>MoveConstructible</tt>.
4570                                         The sequences <tt>vector</tt> and <tt>deque</tt> also require the <tt>value_type</tt> to be <tt>MoveAssignable</tt>.</td></tr>
4571 <tr><td><tt>a.insert(p, n, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4572                                         The sequences <tt>vector</tt> and <tt>deque</tt> also require the <tt>value_type</tt> to be <tt>CopyAssignable</tt>.</td></tr>
4573 <tr><td><tt>a.insert(p, i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4574                                         The sequences <tt>vector</tt> and <tt>deque</tt> also require the <tt>value_type</tt> to be <tt>CopyAssignable</tt> when the iterators return an lvalue.
4575                                         If the iterators return an rvalue the <tt>value_type</tt> must be <tt>MoveConstructible</tt>.
4576                                         The sequences <tt>vector</tt> and <tt>deque</tt> also require the <tt>value_type</tt> to be <tt>MoveAssignable</tt> when the iterators return an rvalue.</td></tr>
4577 <tr><td><tt>a.erase(p)</tt></td><td>The sequences <tt>vector</tt> and <tt>deque</tt> require the <tt>value_type</tt> to be <tt>MoveAssignable</tt>.</td></tr>
4578 <tr><td><tt>a.erase(q1, q2)</tt></td><td>The sequences <tt>vector</tt> and <tt>deque</tt> require the <tt>value_type</tt> to be <tt>MoveAssignable</tt>.</td></tr>
4579 <tr><td><tt>a.clear()</tt></td><td></td></tr>
4580 <tr><td><tt>a.assign(i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt> and <tt>CopyAssignable</tt>.
4581                                         If the iterators return an rvalue the <tt>value_type</tt> must be <tt>MoveConstructible</tt> and <tt>MoveAssignable</tt>.</td></tr>
4582 <tr><td><tt>a.assign(n, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt> and <tt>CopyAssignable</tt>.</td></tr>
4583 <tr><td><tt>a.resize(n)</tt></td><td>The <tt>value_type</tt> must be <tt>DefaultConstructible</tt>.
4584                                      The sequence <tt>vector</tt> also requires the <tt>value_type</tt> to be <tt>MoveConstructible</tt>.</td></tr>
4585 <tr><td><tt>a.resize(n, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4586 </tbody></table>
4587
4588 <p>
4589 </p>
4590
4591 <table border="1">
4592 <caption>Optional Sequence Requirements</caption>
4593 <tbody><tr><td><tt>a.front()</tt></td><td></td></tr>
4594 <tr><td><tt>a.back()</tt></td><td></td></tr>
4595 <tr><td><tt>a.push_front(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4596 <tr><td><tt>a.push_front(rv)</tt></td><td>The <tt>value_type</tt> must be <tt>MoveConstructible</tt>.</td></tr>
4597 <tr><td><tt>a.push_back(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4598 <tr><td><tt>a.push_back(rv)</tt></td><td>The <tt>value_type</tt> must be <tt>MoveConstructible</tt>.</td></tr>
4599 <tr><td><tt>a.pop_front()</tt></td><td></td></tr>
4600 <tr><td><tt>a.pop_back()</tt></td><td></td></tr>
4601 <tr><td><tt>a[n]</tt></td><td></td></tr>
4602 <tr><td><tt>a.at[n]</tt></td><td></td></tr>
4603 </tbody></table>
4604
4605 <p>
4606 </p>
4607
4608 <table border="1">
4609 <caption>Associative Container Requirements</caption>
4610 <tbody><tr><td><tt>X(i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4611                                  If the iterators return an rvalue the <tt>value_type</tt> must be <tt>MoveConstructible</tt>.</td></tr>
4612 <tr><td><tt>a_uniq.insert(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4613 <tr><td><tt>a_uniq.insert(rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4614 <tr><td><tt>a_eq.insert(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4615 <tr><td><tt>a_eq.insert(rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4616 <tr><td><tt>a.insert(p, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4617 <tr><td><tt>a.insert(p, rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4618 <tr><td><tt>a.insert(i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4619                                         If the iterators return an rvalue the <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>..</td></tr>
4620 </tbody></table>
4621
4622 <p>
4623 </p>
4624
4625 <table border="1">
4626 <caption>Unordered Associative Container Requirements</caption>
4627 <tbody><tr><td><tt>X(i, j, n, hf, eq)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4628                                  If the iterators return an rvalue the <tt>value_type</tt> must be <tt>MoveConstructible</tt>.</td></tr>
4629 <tr><td><tt>a_uniq.insert(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4630 <tr><td><tt>a_uniq.insert(rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4631 <tr><td><tt>a_eq.insert(t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4632 <tr><td><tt>a_eq.insert(rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4633 <tr><td><tt>a.insert(p, t)</tt></td><td>The <tt>value_type</tt> must be <tt>CopyConstructible</tt>.</td></tr>
4634 <tr><td><tt>a.insert(p, rv)</tt></td><td>The <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>.</td></tr>
4635 <tr><td><tt>a.insert(i, j)</tt></td><td>If the iterators return an lvalue the <tt>value_type</tt> must be <tt>CopyConstructible</tt>.
4636                                         If the iterators return an rvalue the <tt>key_type</tt> and the <tt>mapped_type</tt> (if it exists) must be <tt>MoveConstructible</tt>..</td></tr>
4637 </tbody></table>
4638
4639 <p>
4640 </p>
4641
4642 <table border="1">
4643 <caption>Miscellaneous Requirements</caption>
4644 <tbody><tr><td><tt>map[lvalue-key]</tt></td><td>The <tt>key_type</tt> must be <tt>CopyConstructible</tt>.
4645                                          The <tt>mapped_type</tt> must be <tt>DefaultConstructible</tt> and <tt>MoveConstructible</tt>.</td></tr>
4646 <tr><td><tt>map[rvalue-key]</tt></td><td>The <tt>key_type</tt> must be <tt>MoveConstructible</tt>.
4647                                          The <tt>mapped_type</tt> must be <tt>DefaultConstructible</tt> and <tt>MoveConstructible</tt>.</td></tr>
4648 </tbody></table>
4649
4650 <p><i>[
4651 Kona (2007): Howard and Alan to update requirements table in issue with emplace signatures.
4652 ]</i></p>
4653
4654
4655 <p><i>[
4656 Bellevue: This should be handled as part of the concepts work.
4657 ]</i></p>
4658
4659
4660 <p><i>[
4661 2009-07-20 Reopened by Howard:
4662 ]</i></p>
4663
4664
4665 <blockquote>
4666 <p>
4667 This is one of the issues that was "solved by concepts" and is now no longer solved.
4668 </p>
4669
4670 <p>
4671 In a nutshell, concepts adopted the "minimum requirements" philosophy outlined
4672 in the discussion of this issue, and enforced it.  My strong suggestion is that
4673 we translate the concepts specification into documentation for the containers.
4674 </p>
4675
4676 <p>
4677 What this means for vendors is that they will have to implement container members
4678 being careful to only use those characteristics of a type that the concepts specification
4679 formally allowed.  Note that I <em>am not</em> talking about <tt>enable_if</tt>'ing
4680 everything.  I am simply suggesting that (for example) we tell the vendor he can't call <tt>T's</tt>
4681 copy constructor or move constructor within the <tt>emplace</tt> member function, etc.
4682 </p>
4683
4684 <p>
4685 What this means for customers is that they will be able to use types within C++03
4686 containers which are sometimes not CopyConstructible, and sometimes not even
4687 MoveConstructible, etc.
4688 </p>
4689 </blockquote>
4690
4691 <p><i>[
4692 2009-10 Santa Cruz:
4693 ]</i></p>
4694
4695
4696 <blockquote>
4697 Leave open. Howard to provide wording.
4698 </blockquote>
4699
4700
4701
4702 <p><b>Proposed resolution:</b></p>
4703
4704
4705
4706 <p><b>Rationale:</b></p>
4707 <p><i>[
4708 post San Francisco:
4709 ]</i></p>
4710
4711
4712 <blockquote>
4713 Solved by
4714 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2776.pdf">N2776</a>.
4715 </blockquote>
4716
4717
4718
4719
4720
4721
4722 <hr>
4723 <h3><a name="724"></a>724. <tt>DefaultConstructible</tt> is not defined</h3>
4724 <p><b>Section:</b> 20.2.1 [utility.arg.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
4725  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2007-09-12  <b>Last modified:</b> 2009-10-23</p>
4726 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#utility.arg.requirements">active issues</a> in [utility.arg.requirements].</p>
4727 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#utility.arg.requirements">issues</a> in [utility.arg.requirements].</p>
4728 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
4729 <p><b>Discussion:</b></p>
4730 <p>
4731 The <tt>DefaultConstructible</tt> requirement is referenced in
4732 several places in the August 2007 working draft
4733 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2369.pdf">N2369</a>,
4734 but is not defined anywhere.
4735 </p>
4736
4737 <p><i>[
4738 Bellevue:
4739 ]</i></p>
4740
4741
4742 <blockquote>
4743 <p>
4744 Walking into the default/value-initialization mess...
4745 </p>
4746 <p>
4747 Why two lines? Because we need both expressions to be valid.
4748 </p>
4749 <p>
4750 AJM not sure what the phrase "default constructed" means. This is
4751 unfortunate, as the phrase is already used 24 times in the library!
4752 </p>
4753 <p>
4754 Example: const int would not accept first line, but will accept the second.
4755 </p>
4756 <p>
4757 This is an issue that must be solved by concepts, but we might need to solve it independantly first.
4758 </p>
4759 <p>
4760 It seems that the requirements are the syntax in the proposed first
4761 column is valid, but not clear what semantics we need.
4762 </p>
4763 <p>
4764 A table where there is no post-condition seems odd, but appears to sum up our position best.
4765 </p>
4766 <p>
4767 At a minimum an object is declared and is destuctible.
4768 </p>
4769 <p>
4770 Move to open, as no-one happy to produce wording on the fly.
4771 </p>
4772 </blockquote>
4773
4774 <p><i>[
4775 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
4776 ]</i></p>
4777
4778
4779 <p><i>[
4780 2009-08-17 Daniel adds "[defaultconstructible]" to table title.  <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#408">408</a>
4781 depends upon this issue.
4782 ]</i></p>
4783
4784
4785 <p><i>[
4786 2009-08-18 Alisdair adds:
4787 ]</i></p>
4788
4789
4790 <blockquote>
4791 <p>
4792 Looking at the proposed table in this issue, it really needs two rows:
4793 </p>
4794
4795 <blockquote>
4796 <table border="1">
4797 <caption>Table 33: <tt>DefaultConstructible</tt> requirements [defaultconstructible]</caption>
4798 <tbody><tr>
4799 <th>expression</th><th>post-condition</th>
4800 </tr>
4801
4802 <tr>
4803 <td><tt>T t;</tt></td><td><tt>t</tt> is default-initialized.</td>
4804 </tr>
4805
4806 <tr>
4807 <td><tt>T{}</tt></td><td>Object of type <tt>T</tt> is value-initialized.</td>
4808 </tr>
4809 </tbody></table>
4810 </blockquote>
4811
4812 <p>
4813 Note I am using the new brace-initialization syntax that is unambiguous
4814 in all use cases (no most vexing parse.)
4815 </p>
4816 </blockquote>
4817
4818 <p><i>[
4819 2009-10-03 Daniel adds:
4820 ]</i></p>
4821
4822
4823 <blockquote>
4824 <p>
4825 The suggested definition <tt>T{}</tt> describing it as
4826 value-initialization is wrong, because it belongs to list-initialization
4827 which would - as the current rules are - always prefer a
4828 initializer-list constructor over a default-constructor. I don't
4829 consider this as an appropriate definition of
4830 <tt>DefaultConstructible</tt>. My primary suggestion is to ask core,
4831 whether the special case <tt>T{}</tt> (which also easily leads to
4832 ambiguity situations for more than one initializer-list in a class)
4833 would always prefer a default-constructor - if any - before considering
4834 an initializer-list constructor or to provide another syntax form to
4835 prefer value-initialization over list-initialization. If that fails I
4836 would fall back to suggest to use the expression <tt>T()</tt> instead of
4837 <tt>T{}</tt> with all it's disadvantages for the meaning of the
4838 expression
4839 </p>
4840
4841 <blockquote><pre>T t();
4842 </pre></blockquote>
4843 </blockquote>
4844
4845 <p><i>[
4846 2009-10 Santa Cruz:
4847 ]</i></p>
4848
4849
4850 <blockquote>
4851 Leave Open. Core is looking to make Alisdair's proposed
4852 resolution correct.
4853 </blockquote>
4854
4855
4856
4857 <p><b>Proposed resolution:</b></p>
4858 <p>
4859 In section 20.2.1 [utility.arg.requirements], before table 33, add the
4860 following table:
4861 </p>
4862
4863 <p style="text-align: center;" align="center">Table 33: <tt>DefaultConstructible</tt> requirements [defaultconstructible]</p>
4864
4865 <div align="center">
4866
4867 <table style="border-collapse: collapse;" border="0" cellpadding="0" cellspacing="0">
4868  <tbody><tr>
4869   <td style="border-style: solid none double solid; border-color: navy -moz-use-text-color navy navy; border-width: 1pt medium 1.5pt 1pt; padding: 0in 5.4pt; width: 85.5pt;" valign="top" width="114">
4870   <p style="margin: 0in 0in 0.0001pt; text-align: center;" align="center">expression</p>
4871   </td>
4872   <td style="border-style: solid solid double none; border-color: navy navy navy -moz-use-text-color; border-width: 1pt 1pt 1.5pt medium; padding: 0in 5.4pt; width: 243pt;" valign="top" width="324">
4873   <p style="margin: 0in 0in 0.0001pt; text-align: center;" align="center">post-condition</p>
4874   </td>
4875  </tr>
4876  <tr>
4877   <td style="border-style: none none solid solid; border-color: -moz-use-text-color -moz-use-text-color navy navy; border-width: medium medium 1pt 1pt; padding: 0in 5.4pt; width: 85.5pt;" valign="top" width="114">
4878   <p style="margin: 0in 0in 0.0001pt;"><tt>T
4879   t;</tt><br>
4880   <tt>T()</tt></p>
4881   </td>
4882   <td style="border-style: none solid solid none; border-color: -moz-use-text-color navy navy -moz-use-text-color; border-width: medium 1pt 1pt medium; padding: 0in 5.4pt; width: 243pt;" valign="top" width="324">
4883   <p style="margin: 0in 0in 0.0001pt;"><tt>T</tt>
4884   is <i>default constructed.</i></p>
4885   </td>
4886  </tr>
4887 </tbody></table>
4888
4889 </div>
4890
4891
4892
4893 <p><b>Rationale:</b></p>
4894 <p><i>[
4895 San Francisco:
4896 ]</i></p>
4897
4898 <blockquote>
4899 We believe concepts will solve this problem
4900 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2774.pdf">N2774</a>).
4901 </blockquote>
4902
4903
4904
4905
4906
4907 <hr>
4908 <h3><a name="726"></a>726. Missing <tt>regex_replace()</tt> overloads</h3>
4909 <p><b>Section:</b> 28.11.4 [re.alg.replace] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
4910  <b>Submitter:</b> Stephan T. Lavavej <b>Opened:</b> 2007-09-22  <b>Last modified:</b> 2009-10-23</p>
4911 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#re.alg.replace">active issues</a> in [re.alg.replace].</p>
4912 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#re.alg.replace">issues</a> in [re.alg.replace].</p>
4913 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
4914 <p><b>Discussion:</b></p>
4915 <p>
4916 Two overloads of <tt>regex_replace()</tt> are currently provided:
4917 </p>
4918
4919 <blockquote><pre>template &lt;class OutputIterator, class BidirectionalIterator,
4920     class traits, class charT&gt;
4921   OutputIterator
4922   regex_replace(OutputIterator out,
4923                 BidirectionalIterator first, BidirectionalIterator last,
4924                 const basic_regex&lt;charT, traits&gt;&amp; e,
4925                 const basic_string&lt;charT&gt;&amp; fmt,
4926                 regex_constants::match_flag_type flags =
4927                   regex_constants::match_default);
4928
4929 template &lt;class traits, class charT&gt;
4930   basic_string&lt;charT&gt;
4931   regex_replace(const basic_string&lt;charT&gt;&amp; s,
4932                 const basic_regex&lt;charT, traits&gt;&amp; e,
4933                 const basic_string&lt;charT&gt;&amp; fmt,
4934                 regex_constants::match_flag_type flags =
4935                   regex_constants::match_default);
4936 </pre></blockquote>
4937
4938 <ol>
4939 <li>Overloads taking <tt>const charT *</tt> are provided for <tt>regex_match()</tt> and
4940 <tt>regex_search()</tt>, but not <tt>regex_replace()</tt>.  This is inconsistent.</li>
4941 <li>
4942 <p>The absence of <tt>const charT *</tt> overloads prevents ordinary-looking code from compiling, such as:</p>
4943
4944 <blockquote><pre>const string s("kitten");
4945 const regex r("en");
4946 cout &lt;&lt; regex_replace(s, r, "y") &lt;&lt; endl;
4947 </pre></blockquote>
4948
4949 <p>
4950 The compiler error message will be something like "could not deduce
4951 template argument for 'const std::basic_string&lt;_Elem&gt; &amp;' from 'const
4952 char[1]'".
4953 </p>
4954
4955 <p>
4956 Users expect that anything taking a <tt>basic_string&lt;charT&gt;</tt> can also take a
4957 <tt>const charT *</tt>.  In their own code, when they write a function taking
4958 <tt>std::string</tt> (or <tt>std::wstring</tt>), they can pass a <tt>const char *</tt> (or <tt>const
4959 wchar_t *</tt>), thanks to <tt>basic_string</tt>'s implicit constructor.  Because the
4960 regex algorithms are templated on <tt>charT</tt>, they can't rely on
4961 <tt>basic_string</tt>'s implicit constructor (as the compiler error message
4962 indicates, template argument deduction fails first).
4963 </p>
4964
4965 <p>
4966 If a user figures out what the compiler error message means, workarounds
4967 are available - but they are all verbose.  Explicit template arguments
4968 could be given to <tt>regex_replace()</tt>, allowing <tt>basic_string</tt>'s implicit
4969 constructor to be invoked - but <tt>charT</tt> is the last template argument, not
4970 the first, so this would be extremely verbose.  Therefore, constructing
4971 a <tt>basic_string</tt> from each C string is the simplest workaround.
4972 </p>
4973 </li>
4974
4975 <li>
4976 There is an efficiency consideration: constructing <tt>basic_string</tt>s can
4977 impose performance costs that could be avoided by a library
4978 implementation taking C strings and dealing with them directly.
4979 (Currently, for replacement sources, C strings can be converted into
4980 iterator pairs at the cost of verbosity, but for format strings, there
4981 is no way to avoid constructing a <tt>basic_string</tt>.)
4982 </li>
4983 </ol>
4984
4985 <p><i>[
4986 Sophia Antipolis:
4987 ]</i></p>
4988
4989
4990 <blockquote>
4991 We note that Boost already has these overloads. However, the proposed
4992 wording is provided only for 28.11.4 [re.alg.replace]; wording is needed for the synopsis
4993 as well. We also note that this has impact on <tt>match_results::format</tt>,
4994 which may require further overloads.
4995 </blockquote>
4996
4997 <p><i>[
4998 2009-07 Frankfurt:
4999 ]</i></p>
5000
5001
5002 <blockquote>
5003 Daniel to tweak for us.
5004 </blockquote>
5005
5006 <p><i>[
5007 2009-07-25 Daniel tweaks both this issue and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>.
5008 ]</i></p>
5009
5010
5011 <blockquote>
5012 <p>
5013 This is solved by the proposed resolution of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>.
5014 </p>
5015 </blockquote>
5016
5017 <p><i>[
5018 2009-10 Santa Cruz:
5019 ]</i></p>
5020
5021
5022 <blockquote>
5023 Leave Open. Though we believe this is solved by the proposed resolution
5024 to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#727">727</a>.
5025 </blockquote>
5026
5027
5028
5029 <p><b>Proposed resolution:</b></p>
5030
5031 <p>
5032 Provide additional overloads for <tt>regex_replace()</tt>: one additional
5033 overload of the iterator-based form (taking <tt>const charT* fmt</tt>), and three
5034 additional overloads of the convenience form (one taking <tt>const charT*
5035 str</tt>, another taking <tt>const charT* fmt</tt>, and the third taking both <tt>const
5036 charT* str</tt> and <tt>const charT* fmt</tt>).  28.11.4 [re.alg.replace]:
5037 </p>
5038
5039 <blockquote>
5040 <pre>template &lt;class OutputIterator, class BidirectionalIterator,
5041     class traits, class charT&gt;
5042   OutputIterator
5043   regex_replace(OutputIterator out,
5044                 BidirectionalIterator first, BidirectionalIterator last,
5045                 const basic_regex&lt;charT, traits&gt;&amp; e,
5046                 const basic_string&lt;charT&gt;&amp; fmt,
5047                 regex_constants::match_flag_type flags =
5048                   regex_constants::match_default);
5049
5050 <ins>template &lt;class OutputIterator, class BidirectionalIterator,
5051     class traits, class charT&gt;
5052   OutputIterator
5053   regex_replace(OutputIterator out,
5054                 BidirectionalIterator first, BidirectionalIterator last,
5055                 const basic_regex&lt;charT, traits&gt;&amp; e,
5056                 const charT* fmt,
5057                 regex_constants::match_flag_type flags =
5058                   regex_constants::match_default);</ins>
5059 </pre>
5060 <p>...</p>
5061 <pre>template &lt;class traits, class charT&gt;
5062   basic_string&lt;charT&gt;
5063   regex_replace(const basic_string&lt;charT&gt;&amp; s,
5064                 const basic_regex&lt;charT, traits&gt;&amp; e,
5065                 const basic_string&lt;charT&gt;&amp; fmt,
5066                 regex_constants::match_flag_type flags =
5067                   regex_constants::match_default);
5068
5069 <ins>template &lt;class traits, class charT&gt;
5070   basic_string&lt;charT&gt;
5071   regex_replace(const basic_string&lt;charT&gt;&amp; s,
5072                 const basic_regex&lt;charT, traits&gt;&amp; e,
5073                 const charT* fmt,
5074                 regex_constants::match_flag_type flags =
5075                   regex_constants::match_default);</ins>
5076
5077 <ins>template &lt;class traits, class charT&gt;
5078   basic_string&lt;charT&gt;
5079   regex_replace(const charT* s,
5080                 const basic_regex&lt;charT, traits&gt;&amp; e,
5081                 const basic_string&lt;charT&gt;&amp; fmt,
5082                 regex_constants::match_flag_type flags =
5083                   regex_constants::match_default);</ins>
5084
5085 <ins>template &lt;class traits, class charT&gt;
5086   basic_string&lt;charT&gt;
5087   regex_replace(const charT* s,
5088                 const basic_regex&lt;charT, traits&gt;&amp; e,
5089                 const charT* fmt,
5090                 regex_constants::match_flag_type flags =
5091                   regex_constants::match_default);</ins>
5092 </pre>
5093 </blockquote>
5094
5095
5096
5097
5098
5099
5100 <hr>
5101 <h3><a name="727"></a>727. <tt>regex_replace()</tt> doesn't accept <tt>basic_string</tt>s with custom traits and allocators</h3>
5102 <p><b>Section:</b> 28.11.4 [re.alg.replace] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
5103  <b>Submitter:</b> Stephan T. Lavavej <b>Opened:</b> 2007-09-22  <b>Last modified:</b> 2009-10-23</p>
5104 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#re.alg.replace">active issues</a> in [re.alg.replace].</p>
5105 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#re.alg.replace">issues</a> in [re.alg.replace].</p>
5106 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
5107 <p><b>Discussion:</b></p>
5108 <p>
5109 <tt>regex_match()</tt> and <tt>regex_search()</tt> take <tt>const basic_string&lt;charT, ST,
5110 SA&gt;&amp;</tt>.  <tt>regex_replace()</tt> takes <tt>const basic_string&lt;charT&gt;&amp;</tt>.  This prevents
5111 <tt>regex_replace()</tt> from accepting <tt>basic_string</tt>s with custom traits and
5112 allocators.
5113 </p>
5114
5115 <p>
5116 Overloads of <tt>regex_replace()</tt> taking <tt>basic_string</tt> should be additionally
5117 templated on <tt>class ST, class SA</tt> and take <tt>const basic_string&lt;charT, ST,
5118 SA&gt;&amp;</tt>.  Consistency with <tt>regex_match()</tt> and <tt>regex_search()</tt> would place
5119 <tt>class ST, class SA</tt> as the first template arguments; compatibility with
5120 existing code using TR1 and giving explicit template arguments to
5121 <tt>regex_replace()</tt> would place <tt>class ST, class SA</tt> as the last template
5122 arguments.
5123 </p>
5124
5125 <p><i>[
5126 Batavia (2009-05):
5127 ]</i></p>
5128
5129 <blockquote>
5130 <p>
5131 Bill comments, "We need to look at the depth of this change."
5132 </p>
5133 <p>
5134 Pete remarks that we are here dealing with a convenience function
5135 that saves a user from calling the iterato-based overload.
5136 </p>
5137 <p>
5138 Move to Open.
5139 </p>
5140 </blockquote>
5141
5142 <p><i>[
5143 2009-07 Frankfurt:
5144 ]</i></p>
5145
5146
5147 <blockquote>
5148 Howard to ask Stephan Lavavej to provide wording.
5149 </blockquote>
5150
5151 <p><i>[
5152 2009-07-17 Stephan provided wording.
5153 ]</i></p>
5154
5155
5156 <p><i>[
5157 2009-07-25 Daniel tweaks both this issue and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#726">726</a>.
5158 ]</i></p>
5159
5160
5161 <blockquote>
5162 <p>
5163 One relevant part of the proposed resolution below suggests
5164 to add a new overload of the format member function in the
5165 <tt>match_results</tt> class template that accepts two character pointers
5166 defining the <tt>begin</tt> and <tt>end</tt> of a format range. A more general
5167 approach could have proposed a pair of iterators instead, but
5168 the used pair of char pointers reflects existing practice. If the
5169 committee strongly favors an iterator-based signature, this
5170 could be simply changed. I think that the minimum requirement
5171 should be a <tt>BidirectionalIterator</tt>, but current implementations
5172 take advantage (at least partially) of the <tt>RandomAccessIterator</tt>
5173 sub interface of the char pointers.
5174 </p>
5175
5176 <p><b>Suggested Resolution:</b></p>
5177
5178 <p><i>[Moved into the proposed resloution]</i></p>
5179
5180
5181
5182 </blockquote>
5183
5184 <p><i>[
5185 2009-07-30 Stephan agrees with Daniel's wording.  Howard places Daniel's wording
5186 in the Proposed Resolution.
5187 ]</i></p>
5188
5189
5190 <p><i>[
5191 2009-10 Santa Cruz:
5192 ]</i></p>
5193
5194
5195 <blockquote>
5196 Move to Review. Chair is anxious to move this to Ready in Pittsburgh.
5197 </blockquote>
5198
5199
5200
5201 <p><b>Proposed resolution:</b></p>
5202
5203 <ol>
5204 <li>
5205 <p>
5206 Change 28.4 [re.syn] as indicated:
5207 </p>
5208
5209 <blockquote><pre>// 28.11.4, function template regex_replace:
5210 template &lt;class OutputIterator, class BidirectionalIterator,
5211           class traits, class charT<ins>, class ST, class SA</ins>&gt;
5212   OutputIterator
5213   regex_replace(OutputIterator out,
5214                 BidirectionalIterator first, BidirectionalIterator last,
5215                 const basic_regex&lt;charT, traits&gt;&amp; e,
5216                 const basic_string&lt;charT<ins>, ST, SA</ins>&gt;&amp; fmt,
5217                 regex_constants::match_flag_type flags =
5218                   regex_constants::match_default);
5219
5220 <ins>
5221 template &lt;class OutputIterator, class BidirectionalIterator,
5222           class traits, class charT&gt;
5223   OutputIterator
5224   regex_replace(OutputIterator out,
5225                 BidirectionalIterator first, BidirectionalIterator last,
5226                 const basic_regex&lt;charT, traits&gt;&amp; e,
5227                 const charT* fmt,
5228                 regex_constants::match_flag_type flags =
5229                   regex_constants::match_default);
5230 </ins>
5231
5232 template &lt;class traits, class charT<ins>, class ST, class SA,
5233           class FST, class FSA</ins>&gt;
5234   basic_string&lt;charT<ins>, ST, SA</ins>&gt;
5235   regex_replace(const basic_string&lt;charT<ins>, ST, SA</ins>&gt;&amp; s,
5236                 const basic_regex&lt;charT, traits&gt;&amp; e,
5237                 const basic_string&lt;charT<ins>, FST, FSA</ins>&gt;&amp; fmt,
5238                 regex_constants::match_flag_type flags =
5239                   regex_constants::match_default);
5240
5241 <ins>
5242 template &lt;class traits, class charT, class ST, class SA&gt;
5243   basic_string&lt;charT, ST, SA&gt;
5244   regex_replace(const basic_string&lt;charT, ST, SA&gt;&amp; s,
5245                 const basic_regex&lt;charT, traits&gt;&amp; e,
5246                 const charT* fmt,
5247                 regex_constants::match_flag_type flags =
5248                   regex_constants::match_default);
5249 </ins>
5250
5251 <ins>
5252 template &lt;class traits, class charT, class ST, class SA&gt;
5253   basic_string&lt;charT&gt;
5254   regex_replace(const charT* s,
5255                 const basic_regex&lt;charT, traits&gt;&amp; e,
5256                 const basic_string&lt;charT, ST, SA&gt;&amp; fmt,
5257                 regex_constants::match_flag_type flags =
5258                   regex_constants::match_default);
5259 </ins>
5260
5261 <ins>
5262 template &lt;class traits, class charT&gt;
5263   basic_string&lt;charT&gt;
5264   regex_replace(const charT* s,
5265                 const basic_regex&lt;charT, traits&gt;&amp; e,
5266                 const charT* fmt,
5267                 regex_constants::match_flag_type flags =
5268                   regex_constants::match_default);
5269 </ins>
5270 </pre></blockquote>
5271 </li>
5272
5273 <li>
5274 <p>
5275 Change 28.10 [re.results]/3, class template <tt>match_results</tt> as
5276 indicated:
5277 </p>
5278
5279 <blockquote><pre><ins>
5280 template &lt;class OutputIter&gt;
5281   OutputIter
5282   format(OutputIter out,
5283          const char_type* fmt_first, const char_type* fmt_last,
5284          regex_constants::match_flag_type flags =
5285            regex_constants::format_default) const;
5286 </ins>
5287
5288 template &lt;class OutputIter<ins>, class ST, class SA</ins>&gt;
5289   OutputIter
5290   format(OutputIter out,
5291          const <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>&amp; fmt,
5292          regex_constants::match_flag_type flags =
5293            regex_constants::format_default) const;
5294
5295 <ins>template &lt;class ST, class SA&gt;</ins>
5296   <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>
5297   format(const <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>&amp; fmt,
5298          regex_constants::match_flag_type flags =
5299            regex_constants::format_default) const;
5300
5301 <ins>
5302 string_type
5303 format(const char_type* fmt,
5304        regex_constants::match_flag_type flags =
5305          regex_constants::format_default) const;
5306 </ins>
5307 </pre></blockquote>
5308 </li>
5309
5310 <li>
5311 <p>
5312 Insert at the very beginning of 28.10.4 [re.results.form] the following:
5313 </p>
5314
5315 <blockquote><pre><ins>
5316 template &lt;class OutputIter&gt;
5317   OutputIter
5318   format(OutputIter out,
5319          const char_type* fmt_first, const char_type* fmt_last,
5320          regex_constants::match_flag_type flags =
5321            regex_constants::format_default) const;
5322 </ins>
5323 </pre>
5324 <blockquote>
5325
5326 <p><ins>
5327 1 <i>Requires:</i> The type <tt>OutputIter</tt> shall satisfy the requirements for an
5328 Output Iterator (24.2.2 [output.iterators]).
5329 </ins></p>
5330
5331 <p><ins>
5332 2 <i>Effects:</i> Copies the character sequence <tt>[fmt_first,fmt_last)</tt> to
5333 <tt>OutputIter out</tt>. Replaces each
5334 format specifier or escape sequence in the copied range with either
5335 the character(s) it represents
5336 or the sequence of characters within <tt>*this</tt> to which it refers. The
5337 bitmasks specified in <tt>flags</tt>
5338 determines what format specifiers and escape sequences are recognized.
5339 </ins></p>
5340
5341 <p><ins>
5342 3 <i>Returns:</i> <tt>out</tt>.
5343 </ins></p>
5344 </blockquote>
5345 </blockquote>
5346 </li>
5347
5348 <li>
5349 <p>
5350 Change 28.10.4 [re.results.form], before p. 1 until p. 3 (according to
5351 previous numbering)
5352 as indicated:
5353 </p>
5354
5355 <blockquote><pre>template &lt;class OutputIter<ins>, class ST, class SA</ins>&gt;
5356   OutputIter
5357   format(OutputIter out,
5358          const <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>&amp; fmt,
5359          regex_constants::match_flag_type flags =
5360            regex_constants::format_default) const;
5361 </pre>
5362
5363 <blockquote>
5364 <p>
5365 <del><i>Requires:</i> The type <tt>OutputIter</tt> shall satisfy the requirements for
5366 an Output Iterator (24.2.3).</del>
5367 </p>
5368
5369 <p>
5370 <del><i>Effects:</i> Copies the character sequence <tt>[fmt.begin(),fmt.end())</tt> to
5371 <tt>OutputIter out</tt>. Replaces each
5372 format specifier or escape sequence in <tt>fmt</tt> with either the
5373 character(s) it represents or the sequence of
5374 characters within <tt>*this</tt> to which it refers. The bitmasks specified in
5375 <tt>flags</tt> determines what format
5376 specifiers and escape sequences are recognized.</del>
5377 </p>
5378
5379 <p>
5380 <i>Returns:</i> <tt><del>out</del><ins>format(out, fmt.data(), fmt.data() +
5381 fmt.size(), flags)</ins></tt>.
5382 </p>
5383 </blockquote>
5384 </blockquote>
5385 </li>
5386
5387 <li>
5388 <p>
5389 Change 28.10.4 [re.results.form], before p. 4 until p. 4 (according to
5390 previous numbering) as indicated:
5391 </p>
5392
5393 <blockquote><pre><ins>template &lt;class ST, class SA&gt;</ins>
5394   <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>
5395   format(const <del>string_type</del><ins>basic_string&lt;char_type, ST, SA&gt;</ins>&amp; fmt,
5396          regex_constants::match_flag_type flags =
5397            regex_constants::format_default) const;
5398 </pre>
5399
5400 <blockquote>
5401 <p>
5402 <i>Effects:</i> <del>Returns a copy of the string <tt>fmt</tt>. Replaces each format
5403 specifier or escape sequence
5404 in <tt>fmt</tt> with either the character(s) it represents or the sequence of
5405 characters within <tt>*this</tt> to which
5406 it refers. The bitmasks specified in flags determines what format
5407 specifiers and escape sequences are
5408 recognized.</del> <ins>Constructs an empty string <tt>result</tt> of type
5409 <tt>basic_string&lt;char_type, ST, SA&gt;</tt>,
5410 and calls <tt>format(back_inserter(result), fmt, flags)</tt>.</ins>
5411 </p>
5412
5413 <p>
5414 <ins><i>Returns:</i> <tt>result</tt></ins>
5415 </p>
5416 </blockquote>
5417 </blockquote>
5418 </li>
5419
5420 <li>
5421 <p>
5422 At the end of 28.10.4 [re.results.form] insert as indicated:
5423 </p>
5424
5425 <blockquote><pre><ins>
5426 string_type
5427   format(const char_type* fmt,
5428          regex_constants::match_flag_type flags =
5429            regex_constants::format_default) const;
5430 </ins></pre>
5431
5432 <blockquote>
5433 <p>
5434 <ins><i>Effects:</i> Constructs an empty string <tt>result</tt> of type <tt>string_type</tt>, and calls
5435 <tt>format(back_inserter(result), fmt, fmt +
5436 char_traits&lt;char_type&gt;::length(fmt), flags)</tt>.</ins>
5437 </p>
5438 <p>
5439 <ins><i>Returns:</i> <tt>result</tt></ins>
5440 </p>
5441 </blockquote>
5442 </blockquote>
5443
5444 </li>
5445
5446 <li>
5447 <p>
5448 Change 28.11.4 [re.alg.replace] before p. 1 as indicated:
5449 </p>
5450
5451 <blockquote><pre>template &lt;class OutputIterator, class BidirectionalIterator,
5452           class traits, class charT<ins>, class ST, class SA</ins>&gt;
5453   OutputIterator
5454   regex_replace(OutputIterator out,
5455                 BidirectionalIterator first, BidirectionalIterator last,
5456                 const basic_regex&lt;charT, traits&gt;&amp; e,
5457                 const basic_string&lt;charT<ins>, ST, SA</ins>&gt;&amp; fmt,
5458                 regex_constants::match_flag_type flags =
5459                   regex_constants::match_default);
5460
5461 <ins>
5462 template &lt;class OutputIterator, class BidirectionalIterator,
5463           class traits, class charT&gt;
5464   OutputIterator
5465   regex_replace(OutputIterator out,
5466                 BidirectionalIterator first, BidirectionalIterator last,
5467                 const basic_regex&lt;charT, traits&gt;&amp; e,
5468                 const charT* fmt,
5469                 regex_constants::match_flag_type flags =
5470                   regex_constants::match_default);
5471 </ins></pre>
5472
5473 <blockquote>
5474 <i>Effects:</i> [..]. If any matches are found then, for each such match, if <tt>!(flags &amp;
5475  regex_constants::format_no_copy)</tt> calls <tt>std::copy(m.prefix().first,
5476 m.prefix().second,
5477  out)</tt>, and then calls <tt>m.format(out, fmt, flags)</tt> <ins>for the first
5478 form of the function
5479  and <tt>m.format(out, fmt, fmt + char_traits&lt;charT&gt;::length(fmt), flags)</tt>
5480 for the second
5481  form</ins>. [..].
5482 </blockquote>
5483 </blockquote>
5484 </li>
5485
5486 <li>
5487 <p>
5488 Change 28.11.4 [re.alg.replace] before p. 3 as indicated:
5489 </p>
5490
5491 <blockquote><pre>template &lt;class traits, class charT<ins>, class ST, class SA,
5492           class FST, class FSA</ins>&gt;
5493   basic_string&lt;charT<ins>, ST, SA</ins>&gt;
5494   regex_replace(const basic_string&lt;charT<ins>, ST, SA</ins>&gt;&amp; s,
5495                 const basic_regex&lt;charT, traits&gt;&amp; e,
5496                 const basic_string&lt;charT<ins>, FST, FSA</ins>&gt;&amp; fmt,
5497                 regex_constants::match_flag_type flags =
5498                   regex_constants::match_default);
5499
5500 <ins>
5501 template &lt;class traits, class charT, class ST, class SA&gt;
5502   basic_string&lt;charT, ST, SA&gt;
5503   regex_replace(const basic_string&lt;charT, ST, SA&gt;&amp; s,
5504                 const basic_regex&lt;charT, traits&gt;&amp; e,
5505                 const charT* fmt,
5506                 regex_constants::match_flag_type flags =
5507                   regex_constants::match_default);
5508 </ins></pre>
5509
5510 <blockquote>
5511 <i>Effects:</i> Constructs an empty string <tt>result</tt> of type <tt>basic_string&lt;charT<ins>,
5512 ST, SA</ins>&gt;</tt>, calls <tt>regex_replace(back_inserter(result), s.begin(), s.end(),
5513 e, fmt, flags)</tt>, and then returns <tt>result</tt>.
5514 </blockquote>
5515 </blockquote>
5516 </li>
5517
5518 <li>
5519 <p>
5520 At the end of 28.11.4 [re.alg.replace] add the following new prototype description:
5521 </p>
5522
5523 <blockquote><pre><ins>
5524 template &lt;class traits, class charT, class ST, class SA&gt;
5525   basic_string&lt;charT&gt;
5526   regex_replace(const charT* s,
5527                 const basic_regex&lt;charT, traits&gt;&amp; e,
5528                 const basic_string&lt;charT, ST, SA&gt;&amp; fmt,
5529                 regex_constants::match_flag_type flags =
5530                   regex_constants::match_default);
5531 </ins>
5532
5533 <ins>
5534 template &lt;class traits, class charT&gt;
5535   basic_string&lt;charT&gt;
5536   regex_replace(const charT* s,
5537                 const basic_regex&lt;charT, traits&gt;&amp; e,
5538                 const charT* fmt,
5539                 regex_constants::match_flag_type flags =
5540                   regex_constants::match_default);
5541 </ins></pre>
5542
5543 <blockquote>
5544 <ins>
5545 <i>Effects:</i> Constructs an empty string <tt>result</tt> of type <tt>basic_string&lt;charT&gt;</tt>,
5546 calls <tt>regex_replace(back_inserter(result), s, s +
5547 char_traits&lt;charT&gt;::length(s),
5548 e, fmt, flags)</tt>, and then returns <tt>result</tt>.
5549 </ins>
5550 </blockquote>
5551 </blockquote>
5552 </li>
5553
5554 </ol>
5555
5556
5557
5558
5559
5560
5561
5562 <hr>
5563 <h3><a name="742"></a>742. Enabling <tt>swap</tt> for proxy iterators</h3>
5564 <p><b>Section:</b> 20.2.1 [utility.arg.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
5565  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2007-10-10  <b>Last modified:</b> 2009-11-08</p>
5566 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#utility.arg.requirements">active issues</a> in [utility.arg.requirements].</p>
5567 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#utility.arg.requirements">issues</a> in [utility.arg.requirements].</p>
5568 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
5569 <p><b>Discussion:</b></p>
5570 <p>
5571 This issue was split from <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a>. <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#672">672</a> now just
5572 deals with changing the requirements of <tt>T</tt> in the <tt>Swappable</tt>
5573 requirement from <tt>CopyConstructible</tt> and <tt>CopyAssignable</tt> to
5574 <tt>MoveConstructible</tt> and <tt>MoveAssignable</tt>.
5575 </p>
5576
5577 <p>
5578 This issue seeks to widen the <tt>Swappable</tt> requirement to support proxy iterators.  Here
5579 is example code:
5580 </p>
5581
5582 <blockquote><pre>namespace Mine {
5583
5584 template &lt;class T&gt;
5585 struct proxy {...};
5586
5587 template &lt;class T&gt;
5588 struct proxied_iterator
5589 {
5590    typedef T value_type;
5591    typedef proxy&lt;T&gt; reference;
5592    reference operator*() const;
5593    ...
5594 };
5595
5596 struct A
5597 {
5598    // heavy type, has an optimized swap, maybe isn't even copyable or movable, just swappable
5599    void swap(A&amp;);
5600    ...
5601 };
5602
5603 void swap(A&amp;, A&amp;);
5604 void swap(proxy&lt;A&gt;, A&amp;);
5605 void swap(A&amp;, proxy&lt;A&gt;);
5606 void swap(proxy&lt;A&gt;, proxy&lt;A&gt;);
5607
5608 }  // Mine
5609
5610 ...
5611
5612 Mine::proxied_iterator&lt;Mine::A&gt; i(...)
5613 Mine::A a;
5614 <b>swap(*i1, a);</b>
5615 </pre></blockquote>
5616
5617 <p>
5618 The key point to note in the above code is that in the call to <tt>swap</tt>, <tt>*i1</tt>
5619 and <tt>a</tt> are different types (currently types can only be <tt>Swappable</tt> with the
5620 same type).  A secondary point is that to support proxies, one must be able to pass rvalues
5621 to <tt>swap</tt>.  But note that I am not stating that the general purpose <tt>std::swap</tt>
5622 should accept rvalues!  Only that overloaded <tt>swap</tt>s, as in the example above, be allowed
5623 to take rvalues.
5624 </p>
5625
5626 <p>
5627 That is, no standard library code needs to change.  We simply need to have a more flexible
5628 definition of <tt>Swappable</tt>.
5629 </p>
5630
5631 <p><i>[
5632 Bellevue:
5633 ]</i></p>
5634
5635
5636 <blockquote>
5637 <p>
5638 While we believe Concepts work will define a swappable concept, we
5639 should still resolve this issue if possible to give guidance to the
5640 Concepts work.
5641 </p>
5642 <p>
5643 Would an ambiguous swap function in two namespaces found by ADL break
5644 this wording? Suggest that the phrase "valid expression" means such a
5645 pair of types would still not be swappable.
5646 </p>
5647 <p>
5648 Motivation is proxy-iterators, but facility is considerably more
5649 general. Are we happy going so far?
5650 </p>
5651 <p>
5652 We think this wording is probably correct and probably an improvement on
5653 what's there in the WP. On the other hand, what's already there in the
5654 WP is awfully complicated. Why do we need the two bullet points? They're
5655 too implementation-centric. They don't add anything to the semantics of
5656 what swap() means, which is there in the post-condition. What's wrong
5657 with saying that types are swappable if you can call swap() and it
5658 satisfies the semantics of swapping?
5659 </p>
5660 </blockquote>
5661
5662 <p><i>[
5663 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
5664 ]</i></p>
5665
5666
5667 <p><i>[
5668 2009-10 Santa Cruz:
5669 ]</i></p>
5670
5671
5672 <blockquote>
5673 Leave as Open. Dave to provide wording.
5674 </blockquote>
5675
5676 <p><i>[
5677 2009-11-08 Howard adds:
5678 ]</i></p>
5679
5680
5681 <blockquote>
5682 Updated wording to sync with
5683 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n3000.pdf">N3000</a>.
5684 Also this issue is very closely related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>.
5685 </blockquote>
5686
5687
5688
5689 <p><b>Proposed resolution:</b></p>
5690 <p>
5691 Change 20.2.1 [utility.arg.requirements]:
5692 </p>
5693
5694 <blockquote>
5695
5696 <p>
5697 -1- The template definitions in the C++ Standard Library refer to various
5698 named requirements whose details are set out in tables 31-38. In these
5699 tables, <tt>T</tt> <ins>and <tt>V</tt> are</ins> <del>is a</del> type<ins>s</ins> to be supplied by a C++ program
5700 instantiating a template; <tt>a</tt>, <tt>b</tt>, and <tt>c</tt> are
5701 values of type <tt>const T</tt>; <tt>s</tt> and <tt>t</tt> are modifiable
5702 lvalues of type <tt>T</tt>; <tt>u</tt> is a value of type (possibly
5703 <tt>const</tt>) <tt>T</tt>; <del>and</del> <tt>rv</tt> is a non-<tt>const</tt>
5704 rvalue of type <tt>T</tt><ins>; <tt>w</tt> is a value of type <tt>T</tt>; and <tt>v</tt> is a value of type <tt>V</tt></ins>.
5705 </p>
5706
5707 <table border="1">
5708 <caption>Table 37: <tt>Swappable</tt> requirements <b>[swappable]</b></caption>
5709 <tbody><tr><th>expression</th><th>Return type</th><th>Post-condition</th></tr>
5710 <tr><td><tt>swap(<del>s</del><ins>w</ins>,<del>t</del><ins>v</ins>)</tt></td><td><tt>void</tt></td>
5711 <td><del><tt>t</tt></del><ins><tt>w</tt></ins> has the value originally
5712 held by <del><tt>u</tt></del><ins><tt>v</tt></ins>, and
5713 <del><tt>u</tt></del><ins><tt>v</tt></ins> has the value originally held
5714 by <del><tt>t</tt></del><ins><tt>w</tt></ins></td></tr>
5715 <tr><td colspan="3">
5716 <p>
5717 The <tt>Swappable</tt> requirement is met by satisfying one or more of the following conditions:
5718 </p>
5719 <ul>
5720 <li>
5721 <tt>T</tt> is <tt>Swappable</tt> if <ins><tt>T</tt> and <tt>V</tt> are
5722 the same type and </ins> <tt>T</tt> satisfies the
5723 <tt>MoveConstructible</tt> requirements (Table
5724 33) and the
5725 <tt>MoveAssignable</tt> requirements (Table
5726 35);
5727 </li>
5728 <li>
5729 <tt>T</tt> is <tt>Swappable</tt> <ins>with <tt>V</tt></ins> if a namespace scope function named
5730 <tt>swap</tt> exists in the same namespace as the definition of
5731 <tt>T</tt> <ins>or <tt>V</tt></ins>, such that the expression
5732 <tt>swap(<del>s</del><ins>w</ins>,<del>t</del> <ins>v</ins>)</tt> is valid and has the
5733 semantics described in this table.
5734 </li>
5735 <li>
5736 <tt>T</tt> is <tt>Swappable</tt> if <tt>T</tt> is an array type whose
5737 element type is <tt>Swappable</tt>.
5738 </li>
5739 </ul>
5740 </td></tr>
5741 </tbody></table>
5742 </blockquote>
5743
5744
5745
5746 <p><b>Rationale:</b></p>
5747 <p><i>[
5748 post San Francisco:
5749 ]</i></p>
5750
5751
5752 <blockquote>
5753 Solved by
5754 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2758.pdf">N2758</a>.
5755 </blockquote>
5756
5757
5758
5759
5760
5761
5762 <hr>
5763 <h3><a name="774"></a>774. Member swap undefined for most containers</h3>
5764 <p><b>Section:</b> 23 [containers] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
5765  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2008-01-14  <b>Last modified:</b> 2009-10-31</p>
5766 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#containers">active issues</a> in [containers].</p>
5767 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#containers">issues</a> in [containers].</p>
5768 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
5769 <p><b>Discussion:</b></p>
5770 <p>
5771 It appears most containers declare but do not define a member-swap
5772 function.
5773 </p>
5774
5775 <p>
5776 This is unfortunate, as all overload the <tt>swap</tt> algorithm to call the
5777 member-swap function!
5778 (required for <tt>swappable</tt> guarantees [Table 37] and Container Requirements
5779 [Table 87])
5780 </p>
5781
5782 <p>
5783 Note in particular that Table 87 gives semantics of <tt>a.swap(b)</tt> as <tt>swap(a,b)</tt>,
5784 yet for all containers we define <tt>swap(a,b)</tt> to call <tt>a.swap(b)</tt> - a circular
5785 definition.
5786 </p>
5787
5788 <p>
5789 A quick survey of clause 23 shows that the following containers provide a
5790 definition for member-swap:
5791 </p>
5792
5793 <blockquote><pre>array
5794 queue
5795 stack
5796 vector
5797 </pre></blockquote>
5798
5799 <p>
5800 Whereas the following declare it, but do not define the semantics:
5801 </p>
5802
5803 <blockquote><pre>deque
5804 list
5805 map
5806 multimap
5807 multiset
5808 priority_queue
5809 set
5810 unordered_map
5811 unordered_multi_map
5812 unordered_multi_set
5813 unordered_set
5814 </pre></blockquote>
5815
5816 <p>
5817 Suggested resolution:
5818 </p>
5819 <blockquote>
5820 Provide a definition for each of the affected containers...
5821 </blockquote>
5822
5823 <p><i>[
5824 Bellevue:
5825 ]</i></p>
5826
5827
5828 <blockquote>
5829 Move to Open and ask Alisdair to provide wording.
5830 </blockquote>
5831
5832 <p><i>[
5833 2009-07 Frankfurt:
5834 ]</i></p>
5835
5836
5837 <blockquote>
5838 Daniel to provide wording.
5839 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2590.pdf">N2590</a>
5840 is no longer applicable.
5841 </blockquote>
5842
5843 <p><i>[
5844 2009-07-28 Daniel provided wording.
5845 ]</i></p>
5846
5847
5848 <blockquote>
5849 <ol>
5850 <li>
5851 It assumes that the proposed resolution for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#883">883</a> is applied,
5852 which breaks the circularity of definition between member
5853 <tt>swap</tt> and free <tt>swap</tt>.
5854 </li>
5855
5856 <li>
5857 It uses the notation of the pre-concept allocator trait
5858 <tt>allocator_propagation_map</tt>, which might be renamed after the
5859 next refactoring phase of generalized allocators.
5860 </li>
5861
5862 <li>
5863 It requires that compare objects, key equal functions and
5864 hash functions in containers are swapped via unqualified free
5865 <tt>swap</tt> according to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#594">594</a>.
5866 </li>
5867 </ol>
5868 </blockquote>
5869
5870 <p><i>[
5871 2009-09-30 Daniel adds:
5872 ]</i></p>
5873
5874
5875 <blockquote>
5876 The outcome of this issue should be considered with the outcome of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1198">1198</a> both in style and in content (e.g. bullet 9 suggests to
5877 define the semantic of <tt>void
5878 priority_queue::swap(priority_queue&amp;)</tt> in terms of the member
5879 <tt>swap</tt> of the container).
5880 </blockquote>
5881
5882 <p><i>[
5883 2009-10 Santa Cruz:
5884 ]</i></p>
5885
5886
5887 <blockquote>
5888 Looked at, but took no action on as it overlaps too much with
5889 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2982.pdf">N2982</a>.
5890 Waiting for a new draft WP.
5891 </blockquote>
5892
5893 <p><i>[
5894 2009-10 Santa Cruz:
5895 ]</i></p>
5896
5897
5898 <blockquote>
5899 Leave as open. Pablo to provide wording.
5900 </blockquote>
5901
5902 <p><i>[
5903 2009-10-26 Pablo updated wording.  Here is the wording he replaced:
5904 ]</i></p>
5905
5906
5907 <blockquote class="note">
5908 <ol>
5909 <li>
5910 <p>
5911 Add a new Throws clause just after X [allocator.propagation.map]/5:
5912 </p>
5913
5914 <blockquote><pre>static void swap(Alloc&amp; a, Alloc&amp; b);
5915 </pre>
5916 <blockquote>
5917 <p>
5918 <i>Effects:</i> [..]
5919 </p>
5920
5921 <p>
5922 <ins><i>Throws:</i> Nothing.</ins>
5923 </p>
5924 </blockquote>
5925 </blockquote>
5926 <p><i>[
5927 This exception requirement is added, such that it's combination with the
5928 general container requirements of
5929 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>
5930 [container.requirements.general]/9
5931 make it unambiguously clear that the following descriptions of "swaps the
5932 allocators" have the following meaning: (a) This swap is done by calling
5933 <tt>allocator_propagation_map&lt;allocator_type&gt;::swap</tt> and (b) This allocator
5934 swap does never propagate an exception
5935 ]</i></p>
5936
5937 </li>
5938
5939 <li>
5940 <p>
5941 Change 23.2.4.1 [associative.reqmts.except]/3 as indicated:
5942 </p>
5943
5944 <blockquote>
5945 For associative containers, no <tt>swap</tt> function throws an exception unless that
5946 exception is thrown by the <del>copy constructor or copy assignment
5947 operator</del>
5948 <ins><tt>swap</tt></ins> of the container's <tt>Pred</tt> object<ins>s</ins><del> (if any)</del>.
5949 </blockquote>
5950 </li>
5951
5952 <li>
5953 <p>
5954 Change 23.2.5.1 [unord.req.except]/3 as indicated:
5955 </p>
5956
5957 <blockquote>
5958 For unordered associative containers, no <tt>swap</tt> function throws an
5959 exception unless
5960 that exception is thrown by the <del>copy constructor or copy
5961 assignment operator</del>
5962 <ins><tt>swap</tt></ins> of the container's <tt>Hash</tt> or <tt>Pred</tt> object<ins>s,
5963 respectively</ins><del> (if any)</del>.
5964 </blockquote>
5965 </li>
5966
5967 <li>
5968 <p>
5969 Insert a new paragraph just after 23.3 [sequences]/1:
5970 </p>
5971
5972 <blockquote>
5973 <ins>In addition to being available via inclusion of the <tt>&lt;algorithm&gt;</tt> header,
5974 the <tt>swap</tt> function templates in 25.3.3 [alg.swap] are also available when the
5975 header <tt>&lt;queue&gt;</tt> is included.</ins>
5976 </blockquote>
5977
5978 <p><i>[
5979 There is a new issue in process that will suggest a minimum header for <tt>swap</tt>
5980 and <tt>move</tt>. If this one is provided, this text can be removed and the header
5981 dependency should be added to <tt>&lt;queue&gt;</tt>
5982 ]</i></p>
5983
5984
5985 </li>
5986
5987 <li>
5988 <p>
5989 Add one further clause at the end of 23.3.1.2 [array.special]:
5990 </p>
5991 <p><i>[This part is added, because otherwise <tt>array::swap</tt> would otherwise
5992 contradict the
5993 general contract of 23.2.1 [container.requirements.general] p. 10 b. 5]</i></p>
5994
5995
5996 <blockquote>
5997 <ins><i>Throws:</i> Nothing, unless one of the element-wise <tt>swap</tt> calls throws
5998 an exception.</ins>
5999 </blockquote>
6000 </li>
6001
6002 <li>
6003 <ol type="a">
6004 <li>
6005 <p>
6006 In 23.3.2 [deque], class template deque synopsis change as indicated:
6007 </p>
6008 <blockquote><pre>void swap(deque<del>&lt;T,Alloc&gt;</del>&amp;);
6009 </pre></blockquote>
6010 </li>
6011
6012 <li>
6013 <p>
6014 At the end of 23.3.2.3 [deque.modifiers] add as indicated:
6015 </p>
6016
6017 <blockquote><pre><ins>void swap(deque&amp; x);</ins>
6018 </pre>
6019 <blockquote>
6020 <p>
6021 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6022 with that of <tt>x</tt>.</ins>
6023 </p>
6024 <p>
6025 <ins><i>Complexity:</i> Constant time.</ins>
6026 </p>
6027 </blockquote>
6028 </blockquote>
6029 </li>
6030 </ol>
6031 </li>
6032
6033 <li>
6034 <ol type="a">
6035 <li>
6036 <p>
6037 In 23.3.3 [forwardlist], class template <tt>forward_list</tt> synposis change as indicated:
6038 </p>
6039
6040 <blockquote><pre>void swap(forward_list<del>&lt;T,Allocator&gt;</del>&amp;);
6041 </pre></blockquote>
6042 </li>
6043
6044 <li>
6045 <p>
6046 At the end of 23.3.3.4 [forwardlist.modifiers] add as indicated:
6047 </p>
6048
6049 <blockquote><pre><ins>void swap(forward_list&amp; x);</ins>
6050 </pre>
6051 <blockquote>
6052 <p>
6053 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6054 with that of <tt>x</tt>.</ins>
6055 </p>
6056 <p>
6057 <ins><i>Complexity:</i> Constant time.</ins>
6058 </p>
6059 </blockquote>
6060 </blockquote>
6061 </li>
6062 </ol>
6063 </li>
6064
6065 <li>
6066 <ol type="a">
6067 <li>
6068 <p>
6069 In 23.3.4 [list], class template <tt>list</tt> synopsis change as indicated:
6070 </p>
6071
6072 <blockquote><pre>void swap(list<del>&lt;T,Allocator&gt;</del>&amp;);
6073 </pre></blockquote>
6074 </li>
6075
6076 <li>
6077 <p>
6078 At the end of 23.3.4.3 [list.modifiers] add as indicated:
6079 </p>
6080
6081 <blockquote><pre><ins>void swap(list&amp; x);</ins>
6082 </pre>
6083
6084 <blockquote>
6085 <p>
6086 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6087 with that of <tt>x</tt>.</ins>
6088 </p>
6089
6090 <p>
6091 <ins><i>Complexity:</i> Constant time.</ins>
6092 </p>
6093 </blockquote>
6094 </blockquote>
6095 </li>
6096 </ol>
6097 </li>
6098
6099 <li>
6100 <p>
6101 At the end of 23.3.5.2.2 [priqueue.members] add a new prototype description:
6102 </p>
6103
6104 <blockquote><pre><ins>void swap(priority_queue&amp; q);</ins>
6105 </pre>
6106 <blockquote>
6107 <p>
6108 <ins><i>Requires:</i> <tt>Compare</tt> shall satisfy the <tt>Swappable</tt> requirements
6109 ( [swappable]).</ins>
6110 </p>
6111
6112 <p><i>[
6113 This requirement is added to ensure that even a user defined <tt>swap</tt>
6114 which is found by
6115 ADL for <tt>Compare</tt> satisfies the <tt>Swappable</tt> requirements
6116 ]</i></p>
6117
6118
6119 <p>
6120 <ins><i>Effects:</i> <tt>this-&gt;c.swap(q.c); swap(this-&gt;comp, q.comp);</tt></ins>
6121 </p>
6122 <p>
6123 <ins><i>Throws:</i> What and if <tt>c.swap(q.c)</tt> and <tt>swap(comp, q.comp)</tt> throws.</ins>
6124 </p>
6125 </blockquote>
6126 </blockquote>
6127 <p><i>[
6128 This part is added, because otherwise <tt>priority_queue::swap</tt> would otherwise
6129 contradict the general contract of 23.2.1 [container.requirements.general] p. 10 b. 5
6130 ]</i></p>
6131
6132 </li>
6133
6134 <li>
6135 <ol type="a">
6136 <li>
6137 <p>
6138 In 23.3.6 [vector], class template <tt>vector</tt> synopsis change as indicated:
6139 </p>
6140
6141 <blockquote><pre>void swap(vector<del>&lt;T,Allocator&gt;</del>&amp;);
6142 </pre></blockquote>
6143 </li>
6144
6145 <li>
6146 <p>
6147 Change 23.3.6.2 [vector.capacity]/8 as indicated:
6148 </p>
6149
6150 <blockquote><pre>void swap(vector<del>&lt;T,Allocator&gt;</del>&amp; x);
6151 </pre>
6152
6153 <blockquote>
6154 <i>Effects:</i> Exchanges the contents and <tt>capacity()</tt> <ins>and swaps the
6155 allocators</ins>
6156 of <tt>*this</tt> with that of <tt>x</tt>.
6157 </blockquote>
6158 </blockquote>
6159 </li>
6160 </ol>
6161 </li>
6162
6163 <li>
6164 <p>
6165 Insert a new paragraph just before 23.4 [associative]/1:
6166 </p>
6167
6168 <blockquote>
6169 <ins>In addition to being available via inclusion of the <tt>&lt;algorithm&gt;</tt> header,
6170 the <tt>swap</tt> function templates in 25.3.3 [alg.swap] are also available when any of the
6171 headers <tt>&lt;map&gt;</tt> or <tt>&lt;set&gt;</tt> are included.</ins>
6172 </blockquote>
6173 </li>
6174
6175 <li>
6176 <ol type="a">
6177 <li>
6178 <p>
6179 In 23.4.1 [map], class template <tt>map</tt> synopsis change as indicated:
6180 </p>
6181
6182 <blockquote><pre>void swap(map<del>&lt;Key,T,Compare,Allocator&gt;</del>&amp;);
6183 </pre></blockquote>
6184 </li>
6185
6186 <li>
6187 <p>
6188 At the end of 23.4.1.3 [map.modifiers] add as indicated:
6189 </p>
6190
6191 <blockquote><pre><ins>void swap(map&amp; x);</ins>
6192 </pre>
6193
6194 <blockquote>
6195 <p>
6196 <ins><i>Requires:</i> Compare shall satisfy the <tt>Swappable</tt> requirements
6197 ( [swappable]).</ins>
6198 </p>
6199
6200 <p><i>[
6201 This requirement is added to ensure that even a user defined <tt>swap</tt>
6202 which is found by ADL for <tt>Compare</tt> satisfies the <tt>Swappable</tt>
6203 requirements
6204 ]</i></p>
6205
6206
6207 <p>
6208 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6209 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the comparison objects
6210 of <tt>*this</tt> and <tt>x</tt>.</ins>
6211 </p>
6212
6213 <p>
6214 <ins><i>Complexity:</i> Constant time</ins>
6215 </p>
6216 </blockquote>
6217 </blockquote>
6218 </li>
6219 </ol>
6220 </li>
6221
6222 <li>
6223 <ol type="a">
6224 <li>
6225 <p>
6226 In 23.4.2 [multimap], class template <tt>multimap</tt> synopsis change as indicated:
6227 </p>
6228
6229 <blockquote><pre>void swap(multimap<del>&lt;Key,T,Compare,Allocator&gt;</del>&amp;);
6230 </pre></blockquote>
6231 </li>
6232
6233 <li>
6234 <p>
6235 At the end of 23.4.2.2 [multimap.modifiers] add as indicated:
6236 </p>
6237
6238 <blockquote><pre><ins>void swap(multimap&amp; x);</ins>
6239 </pre>
6240
6241 <blockquote>
6242 <p>
6243 <ins><i>Requires:</i> Compare shall satisfy the <tt>Swappable</tt> requirements
6244 ( [swappable]).</ins>
6245 </p>
6246 <p>
6247 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6248 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the comparison objects
6249 of <tt>*this</tt> and <tt>x</tt>.</ins>
6250 </p>
6251 <p>
6252 <ins><i>Complexity:</i> Constant time</ins>
6253 </p>
6254 </blockquote>
6255 </blockquote>
6256 </li>
6257 </ol>
6258 </li>
6259
6260 <li>
6261 <ol type="a">
6262 <li>
6263 <p>
6264 In 23.4.3 [set], class template <tt>set</tt> synopsis change as indicated:
6265 </p>
6266
6267 <blockquote><pre>void swap(set<del>&lt;Key,Compare,Allocator&gt;</del>&amp;);
6268 </pre></blockquote>
6269 </li>
6270
6271 <li>
6272 <p>
6273 After section 23.4.3.1 [set.cons] add a new section <ins><tt>set</tt> modifiers
6274  [set.modifiers]</ins>
6275 and add the following paragraphs:
6276 </p>
6277
6278 <blockquote><pre><ins>void swap(set&amp; x);</ins>
6279 </pre>
6280
6281 <blockquote>
6282 <p>
6283 <ins><i>Requires:</i> Compare shall satisfy the <tt>Swappable</tt> requirements
6284 ( [swappable]).</ins>
6285 </p>
6286
6287 <p>
6288 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6289 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the comparison objects
6290 of <tt>*this</tt> and <tt>x</tt>.</ins>
6291 </p>
6292
6293 <p>
6294 <ins>Complexity: Constant time</ins>
6295 </p>
6296 </blockquote>
6297 </blockquote>
6298 </li>
6299 </ol>
6300 </li>
6301
6302 <li>
6303 <ol type="a">
6304 <li>
6305 <p>
6306 In 23.4.4 [multiset], class template <tt>multiset</tt> synosis, change as indicated:
6307 </p>
6308
6309 <blockquote><pre>void swap(multiset<del>&lt;Key,Compare,Allocator&gt;</del>&amp;);
6310 </pre></blockquote>
6311 </li>
6312
6313 <li>
6314 <p>
6315 After section 23.4.4.1 [multiset.cons] add a new section <ins><tt>multiset</tt> modifiers
6316  [multiset.modifiers]</ins> and add the following paragraphs:
6317 </p>
6318
6319 <blockquote><pre><ins>void swap(multiset&amp; x);</ins>
6320 </pre>
6321
6322 <blockquote>
6323 <p>
6324 <ins><i>Requires:</i> Compare shall satisfy the <tt>Swappable</tt> requirements
6325 ( [swappable]).</ins>
6326 </p>
6327
6328 <p>
6329 <ins><i>Effects:</i> Exchanges the contents and swaps the allocators of <tt>*this</tt>
6330 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the comparison objects
6331 of <tt>*this</tt> and <tt>x</tt>.</ins>
6332 </p>
6333
6334 <p>
6335 <ins><i>Complexity:</i> Constant time</ins>
6336 </p>
6337 </blockquote>
6338 </blockquote>
6339 </li>
6340 </ol>
6341 </li>
6342
6343 <li>
6344 <p>
6345 Insert a new paragraph just before 23.5 [unord]/1:
6346 </p>
6347
6348 <blockquote>
6349 <ins>In addition to being available via inclusion of the <tt>&lt;algorithm&gt;</tt> header,
6350 the <tt>swap</tt> function templates in 25.3.3 [alg.swap] are also available when any of the
6351 headers <tt>&lt;unordered_map&gt;</tt> or <tt>&lt;unordered_set&gt;</tt> are included.</ins>
6352 </blockquote>
6353
6354 </li>
6355
6356 <li>
6357 <p>
6358 After section 23.5.1.2 [unord.map.elem] add a new section <ins>unordered_map
6359 modifiers  [unord.map.modifiers]</ins> and add the following paragraphs:
6360 </p>
6361
6362 <blockquote><pre><ins>void swap(unordered_map&amp; x);</ins>
6363 </pre>
6364
6365 <blockquote>
6366 <p>
6367 <ins><i>Requires:</i> <tt>Hash</tt> and <tt>Pred</tt> shall satisfy the <tt>Swappable</tt> requirements
6368 ( [swappable]).</ins>
6369 </p>
6370
6371 <p><i>[
6372 This requirement is added to ensure that even a user defined <tt>swap</tt>
6373 which is found by ADL for <tt>Hash</tt> and <tt>Pred</tt> satisfies the <tt>Swappable</tt>
6374 requirements
6375 ]</i></p>
6376
6377
6378 <p>
6379 <ins><i>Effects:</i> Exchanges the contents and hash policy and swaps the
6380 allocators of <tt>*this</tt>
6381 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the <tt>Pred</tt> objects
6382 and an unqualified <tt>swap</tt> of the <tt>Hash</tt> objects of <tt>*this</tt> and <tt>x</tt>.</ins>
6383 </p>
6384
6385 <p>
6386 <ins><i>Complexity:</i> Constant time</ins>
6387 </p>
6388 </blockquote>
6389 </blockquote>
6390 </li>
6391
6392 <li>
6393 <p>
6394 After section 23.5.2.1 [unord.multimap.cnstr] add a new section
6395 <ins>unordered_multimap
6396 modifiers  [unord.multimap.modifiers]</ins> and add the following paragraphs:
6397 </p>
6398
6399 <blockquote><pre><ins>void swap(unordered_multimap&amp; x);</ins>
6400 </pre>
6401
6402 <blockquote>
6403 <p>
6404 <ins><i>Requires:</i> <tt>Hash</tt> and <tt>Pred</tt> shall satisfy the <tt>Swappable</tt> requirements
6405 ( [swappable]).</ins>
6406 </p>
6407
6408 <p>
6409 <ins><i>Effects:</i> Exchanges the contents and hash policy and swaps the
6410 allocators of <tt>*this</tt>
6411 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the <tt>Pred</tt> objects
6412 and an unqualified <tt>swap</tt> of the <tt>Hash</tt> objects of <tt>*this</tt> and <tt>x</tt></ins>
6413 </p>
6414 <p>
6415 <ins><i>Complexity:</i> Constant time</ins>
6416 </p>
6417 </blockquote>
6418 </blockquote>
6419 </li>
6420
6421 <li>
6422 <p>
6423 After section 23.5.3.1 [unord.set.cnstr] add a new section
6424 <ins>unordered_set modifiers
6425  [unord.set.modifiers]</ins> and add the following paragraphs:
6426 </p>
6427
6428 <blockquote><pre><ins>void swap(unordered_set&amp; x);</ins>
6429 </pre>
6430
6431 <blockquote>
6432 <p>
6433 <ins><i>Requires:</i> <tt>Hash</tt> and <tt>Pred</tt> shall satisfy the <tt>Swappable</tt> requirements
6434 ( [swappable]).</ins>
6435 </p>
6436
6437 <p>
6438 <ins><i>Effects:</i> Exchanges the contents and hash policy and swaps the
6439 allocators of <tt>*this</tt>
6440 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the <tt>Pred</tt> objects
6441 and an unqualified <tt>swap</tt> of the <tt>Hash</tt> objects of <tt>*this</tt> and <tt>x</tt></ins>
6442 </p>
6443
6444 <p>
6445 <ins><i>Complexity:</i> Constant time</ins>
6446 </p>
6447 </blockquote>
6448 </blockquote>
6449 </li>
6450
6451 <li>
6452 <p>
6453 After section 23.5.4.1 [unord.multiset.cnstr] add a new section
6454 <ins>unordered_multiset
6455 modifiers  [unord.multiset.modifiers]</ins> and add the following paragraphs:
6456 </p>
6457
6458 <blockquote><pre><ins>void swap(unordered_multiset&amp; x);</ins>
6459 </pre>
6460
6461 <blockquote>
6462 <p>
6463 <ins><i>Requires:</i> <tt>Hash</tt> and <tt>Pred</tt> shall satisfy the <tt>Swappable</tt> requirements
6464 ( [swappable]).</ins>
6465 </p>
6466
6467 <p>
6468 <ins><i>Effects:</i> Exchanges the contents and hash policy and swaps the
6469 allocators of <tt>*this</tt>
6470 with that of <tt>x</tt>, followed by an unqualified <tt>swap</tt> of the <tt>Pred</tt> objects
6471 and an unqualified <tt>swap</tt> of the <tt>Hash</tt> objects of <tt>*this</tt> and <tt>x</tt></ins>
6472 </p>
6473 <p>
6474 <ins><i>Complexity:</i> Constant time</ins>
6475 </p>
6476 </blockquote>
6477 </blockquote>
6478 </li>
6479
6480 </ol>
6481
6482 </blockquote>
6483
6484 <p><i>[
6485 2009-10-30 Pablo and Daniel updated wording.
6486 ]</i></p>
6487
6488
6489
6490
6491 <p><b>Proposed resolution:</b></p>
6492
6493 <p><i>[
6494 This resolution is based on the September 2009 WP,
6495 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>,
6496 except that it
6497 assumes that
6498 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2982.pdf">N2982</a>
6499 and issues <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#883">883</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1232">1232</a> have already been applied.  Note in
6500 particular that Table 91 in
6501 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>
6502 is refered to as Table 90 because
6503 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2982.pdf">N2982</a>
6504 removed the old Table 90.  This resolution also addresses issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#431">431</a>.
6505 ]</i></p>
6506
6507 <p>
6508 In 23.2.1 [container.requirements.general], replace the a.swap(b) row in table 90,
6509 "container requirements" (was table 91 before the application of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2982.pdf">N2982</a> to the
6510 WP):
6511 </p>
6512 <blockquote>
6513 <table border="1">
6514   <tbody><tr>
6515     <td><code>a.swap(b)</code></td>
6516     <td><code>void</code></td>
6517     <td>&nbsp;&nbsp;&nbsp;</td>
6518     <td><code><del>swap(a,b)</del><ins>Exchange the contents of <tt>a</tt> and <tt>b</tt>.</ins></code></td>
6519     <td>(Note A)</td>
6520   </tr>
6521   <tr>
6522     <td><ins><code>swap(a,b)</code></ins></td>
6523     <td><ins><code>void</code></ins></td>
6524     <td><code>&nbsp;&nbsp;&nbsp;</code></td>
6525     <td><ins><code>a.swap(b)</code></ins></td>
6526     <td><ins>(Note A)</ins></td>
6527   </tr>
6528 </tbody></table>
6529 </blockquote>
6530 <p>
6531 Modify the notes immediately following Table 90 in
6532 23.2.1 [container.requirements.general] as follows (The wording below is after the
6533 application of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2982.pdf">N2982</a> to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>.  The editor might also want to combine Notes
6534 A and B into one.):
6535 </p>
6536 <blockquote><p>
6537 Notes: the algorithms<del> swap(),</del> equal() and lexicographical_compare()
6538 are defined in Clause 25.  Those entries marked "(Note A)" or "(Note B)"
6539 <del>should</del> have <ins>linear complexity for array and</ins> constant
6540 complexity <ins>for all other standard containers</ins>.
6541 </p></blockquote>
6542 <p>
6543 In 23.2.1 [container.requirements.general], after paragraph 9, add:
6544 </p>
6545 <blockquote><p><ins>
6546 The expression <code>a.swap(b)</code>, for containers <code>a</code>
6547 and <code>b</code> of a standard container type other than <code>array</code>,
6548 exchanges the values of <code>a</code> and <code>b</code> without invoking any
6549 move, copy, or swap operations on the individual container elements.
6550 Any <code>Compare</code>, <code>Pred</code>, or <code>Hash</code> function
6551 objects belonging to <code>a</code> and <code>b</code> shall satisfy
6552 the <code>Swappable</code> requirements and are exchanged by unqualified calls
6553 to non-member <code>swap</code>.  If
6554 <code>allocator_traits&lt;allocator_type&gt;::propagate_on_container_swap::value
6555 == true</code>, then the allocators of <code>a</code> and <code>b</code> are
6556 also exchanged using an unqualified call to non-member <code>swap</code>.
6557 Otherwise, the behavior is undefined unless <code>a.get_allocator() ==
6558 b.get_allocator()</code>.  Each iterator refering to an element in one
6559 container before the swap shall refer to the same element in the other
6560 container after the swap.  It is unspecified whether an iterator with
6561 value <code>a.end()</code> before the swap will have
6562 value <code>b.end()</code> after the swap.  In addition to being available via
6563 inclusion of the <code>&lt;utility&gt;</code> header, the <code>swap</code>
6564 function template in 25.3.3 [alg.swap] is also available within the definition of
6565 every standard container's <code>swap</code> function.
6566 </ins></p></blockquote>
6567 <p><i>[
6568 Note to the editor: Paragraph 2 starts with a sentence fragment,
6569 clearly from an editing or source-control error.
6570 ]</i></p>
6571
6572 <p>
6573 Modify 23.2.4.1 [associative.reqmts.except] as follows:
6574 </p>
6575 <blockquote>
6576 <p>
6577 <b>23.2.4.1 Exception safety guarantees 23.2.4.1 [associative.reqmts.except]</b>
6578 </p>
6579 <p>
6580 For associative containers, no <code>clear()</code> function throws an
6581 exception. <code>erase(k)</code> does not throw an exception unless that
6582 exception is thrown by the
6583 container's <code><del>Pred</del><ins>Compare</ins></code> object (if any).
6584 </p>
6585 <p>
6586 For associative containers, if an exception is thrown by any operation from
6587 within an <code>insert()</code> function inserting a single element,
6588 the <code>insert()</code> function has no effect.
6589 </p>
6590 <p>
6591 For associative containers, no <code>swap</code> function throws an exception
6592 unless that exception is thrown by the <del>copy constructor
6593 or copy assignment operator</del><ins>swap</ins> of the
6594 container's <code><del>Pred</del><ins>Compare</ins></code> object (if any).
6595 </p></blockquote>
6596 <p>
6597 Modify 23.2.5.1 [unord.req.except], paragraph 3 as follows:
6598 </p>
6599 <blockquote><p>
6600 For unordered associative containers, no <code>swap</code> function throws an
6601 exception unless that exception is thrown by the <del>copy constructor or copy
6602 assignment operator</del><ins>swap</ins> of the container's <code>Hash</code>
6603 or <code>Pred</code> object (if any).
6604 </p></blockquote>
6605 <p>
6606 Modify section 23.3.1.2 [array.special]:
6607 </p>
6608 <blockquote>
6609 <p>
6610 <b>array specialized algorithms 23.3.1.2 [array.special]</b>
6611 </p>
6612 <p>
6613 <code>template &lt;class T, size_t N&gt; void swap(array&lt;T,N&gt;&amp; x,array&lt;T,N&gt;&amp; y);</code>
6614 </p>
6615 <blockquote>
6616 <p>
6617 <i>Effects:</i> <code><del>swap_ranges(x.begin(), x.end(), y.begin() );</del><ins>x.swap(y);</ins></code>
6618 </p>
6619 </blockquote>
6620 </blockquote>
6621 <p>
6622 Add a new section after 23.3.1.5 [array.fill] (Note to the editor: array::fill make use
6623 of a concept requirement that must be removed or changed to text.):
6624 </p>
6625 <blockquote>
6626 <p>
6627 <ins><b>array::swap [array.swap]</b></ins>
6628 </p>
6629 <p>
6630 <ins><code>void swap(array&amp; y);</code></ins>
6631 </p>
6632 <blockquote>
6633 <p><ins>
6634 <i>Effects:</i> <code>swap_ranges(this-&gt;begin(), this-&gt;end(), y.begin() );</code>
6635 </ins></p>
6636 <p><ins>
6637 <i>Throws:</i> Nothing unless one of the element-wise swap calls throws an
6638 exception.
6639 </ins></p>
6640 <p><ins>
6641 [<i>Note</i>: Unlike other containers' <code>swap</code> functions,
6642 <code>array::swap</code> takes linear, not constant, time, may exit via an
6643 exception, and does not cause iterators to become associated with the other
6644 container. &#8212; <i>end note</i>]
6645 </ins></p>
6646 </blockquote>
6647 </blockquote>
6648
6649 <p>
6650 Insert a new paragraph just after 23.3.5 [container.adaptors]/1:
6651 </p>
6652 <blockquote><p><ins>
6653 For container adaptors, no <code>swap</code> function throws an exception
6654 unless that exception is thrown by the swap of the
6655 adaptor's <code>Container</code> or <code>Compare</code> object (if any).
6656 </ins></p></blockquote>
6657
6658
6659
6660
6661
6662
6663
6664
6665
6666
6667
6668
6669 <hr>
6670 <h3><a name="780"></a>780. <tt>std::merge()</tt> specification incorrect/insufficient</h3>
6671 <p><b>Section:</b> 25.4.4 [alg.merge] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
6672  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-01-25  <b>Last modified:</b> 2009-10-22</p>
6673 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
6674 <p><b>Discussion:</b></p>
6675 <p>
6676 Though issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#283">283</a> has fixed many open issues, it seems that some are still open:
6677 </p>
6678
6679 <p>
6680 Both 25.3.4 [lib.alg.merge] in 14882:2003 and 25.4.4 [alg.merge] in N2461
6681 have no Requires element and the Effects element contains some requirements,
6682 which is probably editorial. Worse is that:
6683 </p>
6684
6685 <ul>
6686 <li>
6687 no assignment requirements are specified (neither implicit nor explicit).
6688 </li>
6689
6690 <li>
6691 the effects clause just speaks of "merges", which is badly worded
6692 near to a circular definition.
6693 </li>
6694
6695 <li>
6696 p. 2 mentions a range <tt>[first, last)</tt>, which is not defined by the
6697 function arguments or otherwise.
6698 </li>
6699
6700 <li>
6701 p. 2 says "according to the ordering defined by comp" which is both
6702 incomplete (because
6703 this excludes the first variant with &lt;) and redundant (because the
6704 following subordinate
6705 clause mentions comp again)
6706 </li>
6707 </ul>
6708
6709 <p><i>[
6710 Post Summit Alisdair adds:
6711 ]</i></p>
6712
6713
6714 <blockquote>
6715 <p>
6716 Suggest:
6717 </p>
6718 <blockquote>
6719 (where <tt>last</tt> is equal to <tt>next(result, distance(first1, last1) +
6720 distance(first2, last2))</tt>, such that resulting range will be sorted in
6721 non-decreasing order; that is, for every iterator <tt>i</tt> in <tt>[result,last)</tt> other
6722 than <tt>result</tt>, the condition <tt>*i &lt; *prev(i)</tt> or, respectively, <tt>comp(*i,
6723 *prev(i))</tt> will be <tt>false</tt>.
6724 </blockquote>
6725
6726 <p>
6727 Note that this might still not be technically accurate in the case of
6728 <tt>InputIterators</tt>, depending on other resolutions working their way through the
6729 system (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1011">1011</a>).
6730 </p>
6731 </blockquote>
6732
6733 <p><i>[
6734 Post Summit Daniel adds:
6735 ]</i></p>
6736
6737
6738 <blockquote>
6739 If we want to use <tt>prev</tt> and <tt>next</tt> here (Note: <tt>merge</tt>
6740 is sufficiently satisfied with <tt>InputIterator</tt>) we should instead *add* more to
6741 25 [algorithms]/6, but I can currently not propose any good wording for this.
6742 </blockquote>
6743
6744 <p><i>[
6745 Batavia (2009-05):
6746 ]</i></p>
6747
6748 <blockquote>
6749 <p>
6750 Pete points out the existing wording in [algorithms]/4
6751 that permits the use of + in algorithm specifications.
6752 </p>
6753 <p>
6754 Alisdair points out that that wording may not apply to input iterators.
6755 </p>
6756 <p>
6757 Move to Review.
6758 </p>
6759 </blockquote>
6760
6761 <p><i>[
6762 2009-07 Frankfurt:
6763 ]</i></p>
6764
6765
6766 <blockquote>
6767 Move to Ready.
6768 </blockquote>
6769
6770 <p><i>[
6771 2009-08-23 Daniel reopens:
6772 ]</i></p>
6773
6774
6775 <blockquote>
6776 <p>
6777 The proposed wording must be rephrased, because the part
6778 </p>
6779
6780 <blockquote>
6781 for every iterator <tt>i</tt> in <tt>[result,last)</tt> other than <tt>result</tt>, the condition
6782 <tt>*i &lt; *(i - 1)</tt> or, respectively, <tt>comp(*i, *(i - 1))</tt> will be <tt>false</tt>"
6783 </blockquote>
6784
6785 <p>
6786 isn't meaningful, because the range <tt>[result,last)</tt> is that of a pure
6787 <tt>OutputIterator</tt>, which is not <em>readable</em> in general.
6788 </p>
6789
6790 <p><i>[Howard:  Proposed wording updated by Daniel, status moved from Ready to Review.]</i></p>
6791
6792 </blockquote>
6793
6794 <p><i>[
6795 2009-10 Santa Cruz:
6796 ]</i></p>
6797
6798
6799 <blockquote>
6800 <p>
6801 Matt has some different words to propose.  Those words have been moved into
6802 the proposed wording section, and the original proposed wording now appears
6803 here:
6804 </p>
6805 <blockquote>
6806 <p>
6807 In 25.4.4 [alg.merge] replace p.1+ 2:
6808 </p>
6809
6810 <blockquote>
6811 <p>
6812 <i>Effects:</i> <del>Merges</del><ins>Copies all the elements of the</ins>
6813 two sorted ranges
6814 <tt>[first1,last1)</tt> and <tt>[first2,last2)</tt> into the range <tt>[result,result +
6815 (last1 - first1) + (last2 - first2))</tt>
6816 <ins>, such that resulting range will be sorted in non-decreasing
6817 order; that is for every
6818 pair of iterators <tt>i</tt> and <tt>j</tt> of either input ranges, where <tt>*i</tt> was copied
6819 to the output range
6820 before <tt>*j</tt> was copied to the output range, the condition <tt>*j &lt; *i</tt> or,
6821 respectively, <tt>comp(*j, *i)</tt>
6822 will be <tt>false</tt>.</ins>
6823 </p>
6824
6825 <p>
6826 <ins><i>Requires:</i></ins>The resulting range shall not overlap with either
6827 of the original ranges.
6828 <del>The list will be sorted in non-decreasing order according to the
6829 ordering defined by
6830 <tt>comp</tt>; that is, for every iterator <tt>i</tt> in <tt>[first,last)</tt> other than <tt>first</tt>,
6831 the condition <tt>*i &lt; *(i - 1)</tt> or
6832 <tt>comp(*i, *(i - 1))</tt> will be <tt>false</tt>.</del>
6833 </p>
6834 </blockquote>
6835 </blockquote>
6836 </blockquote>
6837
6838
6839
6840 <p><b>Proposed resolution:</b></p>
6841 <p><del>
6842 <i>Effects:</i> Merges two sorted ranges <tt>[first1,last1)</tt> and
6843 <tt>[first2,last2)</tt> into the range <tt>[result, result + (last1 -
6844 first1) + (last2 - first2))</tt>.
6845 </del></p>
6846 <p><ins>
6847 <i>Effects:</i> Copies all the elements of the two sorted ranges
6848 <tt>[first1,last1)</tt> and <tt>[first2,last2)</tt> into the range
6849 <tt>[result, result_last)</tt>, where <tt>result_last</tt> is <tt>result
6850 + (last1 - first1) + (last2 - first2)</tt>, such that the resulting
6851 range satisfies <tt>is_sorted(result, result_last)</tt> or
6852 <tt>is_sorted(result, result_last, comp)</tt>, respectively.
6853 </ins></p>
6854
6855 <p>
6856 <ins><i>Requires:</i></ins> The resulting range shall not overlap with
6857 either of the original ranges.  <del>The list will be sorted in
6858 non-decreasing order according to the ordering defined by <tt>comp</tt>;
6859 that is, for every iterator <tt>i</tt> in <tt>[first,last)</tt> other
6860 than <tt>first</tt>, the condition <tt>*i &lt; *(i - 1)</tt> or
6861 <tt>comp(*i, *(i - 1))</tt> will be <tt>false</tt>.</del>
6862 </p>
6863
6864
6865
6866
6867
6868
6869 <hr>
6870 <h3><a name="801"></a>801. <tt>tuple</tt> and <tt>pair</tt> trivial members</h3>
6871 <p><b>Section:</b> 20.5 [tuple] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
6872  <b>Submitter:</b> Lawrence Crowl <b>Opened:</b> 2008-02-18  <b>Last modified:</b> 2009-10-23</p>
6873 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#tuple">issues</a> in [tuple].</p>
6874 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
6875 <p><b>Discussion:</b></p>
6876 <p>
6877 Classes with trivial special member functions are inherently more
6878 efficient than classes without such functions.  This efficiency is
6879 particularly pronounced on modern ABIs that can pass small classes
6880 in registers.  Examples include value classes such as complex numbers
6881 and floating-point intervals.  Perhaps more important, though, are
6882 classes that are simple collections, like <tt>pair</tt> and <tt>tuple</tt>.  When the
6883 parameter types of these classes are trivial, the <tt>pair</tt>s and <tt>tuple</tt>s
6884 themselves can be trivial, leading to substantial performance wins.
6885 </p>
6886 <p>
6887 The current working draft make specification of trivial functions
6888 (where possible) much easer through <tt>default</tt>ed and <tt>delete</tt>d functions.
6889 As long as the semantics of defaulted and deleted functions match
6890 the intended semantics, specification of defaulted and deleted
6891 functions will yield more efficient programs.
6892 </p>
6893 <p>
6894 There are at least two cases where specification of an explicitly
6895 defaulted function may be desirable.
6896 </p>
6897 <p>
6898 First, the <tt>std::pair</tt> template has a non-trivial default constructor,
6899 which prevents static initialization of the pair even when the
6900 types are statically initializable.  Changing the definition to
6901 </p>
6902
6903 <blockquote><pre>pair() = default;
6904 </pre></blockquote>
6905
6906 <p>
6907 would enable such initialization.  Unfortunately, the change is
6908 not semantically neutral in that the current definition effectively
6909 forces value initialization whereas the change would not value
6910 initialize in some contexts.
6911 </p>
6912
6913 <p>
6914 ** Does the committee confirm that forced value initialization
6915 was the intent?  If not, does the committee wish to change the
6916 behavior of <tt>std::pair</tt> in C++0x?
6917 </p>
6918 <p>
6919 Second, the same default constructor issue applies to <tt>std::tuple</tt>.
6920 Furthermore, the <tt>tuple</tt> copy constructor is current non-trivial,
6921 which effectively prevents passing it in registers.  To enable
6922 passing <tt>tuples</tt> in registers, the copy constructor should be
6923 make explicitly <tt>default</tt>ed.  The new declarations are:
6924 </p>
6925
6926 <blockquote><pre>tuple() = default;
6927 tuple(const tuple&amp;) = default;
6928 </pre></blockquote>
6929
6930 <p>
6931 This changes is not implementation neutral.  In particular, it
6932 prevents implementations based on pointers to the parameter
6933 types.  It does however, permit implementations using the
6934 parameter types as bases.
6935 </p>
6936 <p>
6937 ** How does the committee wish to trade implementation
6938 efficiency versus implementation flexibility?
6939 </p>
6940
6941 <p><i>[
6942 Bellevue:
6943 ]</i></p>
6944
6945
6946 <blockquote>
6947 <p>
6948 General agreement; the first half of the issue is NAD.
6949 </p>
6950 <p>
6951 Before voting on the second half, it was agreed that a "Strongly Favor"
6952 vote meant support for trivial tuples (assuming usual requirements met),
6953 even at the expense of other desired qualities. A "Weakly Favor" vote
6954 meant support only if not at the expense of other desired qualities.
6955 </p>
6956 <p>
6957 Concensus: Go forward, but not at expense of other desired qualities.
6958 </p>
6959 <p>
6960 It was agreed to Alisdair should fold this work in with his other
6961 pair/tuple action items, above, and that issue 801 should be "open", but
6962 tabled until Alisdair's proposals are disposed of.
6963 </p>
6964 </blockquote>
6965
6966 <p><i>[
6967 2009-05-27 Daniel adds:
6968 ]</i></p>
6969
6970
6971 <blockquote>
6972 This is partly solved by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1117">1117</a>.
6973 </blockquote>
6974
6975 <p><i>[
6976 2009-07 Frankfurt:
6977 ]</i></p>
6978
6979
6980 <blockquote>
6981 Wait for dust to settle from fixing exception safety problem
6982 with rvalue refs.
6983 </blockquote>
6984
6985 <p><i>[
6986 2009-07-20 Alisdair adds:
6987 ]</i></p>
6988
6989
6990 <blockquote>
6991 <p>
6992 Basically, this issue is what should we do with the default constructor
6993 for pairs and tuples of trivial types.  The motivation of the issue was
6994 to force static initialization rather than dynamic initialization, and
6995 was rejected in the case of pair as it would change the meaning of
6996 existing programs.  The advice was "do the best we can" for tuple
6997 without changing existing meaning.
6998 </p>
6999
7000 <p>
7001 Frankfurt seems to simply wait and see the resolution on no-throw move
7002 constructors, which (I believe) is only tangentially related to this
7003 issue, but as good as any to defer until Santa Cruz.
7004 </p>
7005
7006 <p>
7007 Looking again now, I think constant (static) initialization for pair can
7008 be salvaged by making the default construct constexpr.  I have a
7009 clarification from Core that this is intended to work, even if the
7010 constructor is not trivial/constexpr, so long as no temporaries are
7011 implied in the process (even if elided).
7012 </p>
7013 </blockquote>
7014
7015 <p><i>[
7016 2009-10 Santa Cruz:
7017 ]</i></p>
7018
7019
7020 <blockquote>
7021 Leave as open. Alisdair to provide wording.
7022 </blockquote>
7023
7024
7025
7026 <p><b>Proposed resolution:</b></p>
7027 <p>
7028 </p>
7029
7030
7031
7032
7033
7034 <hr>
7035 <h3><a name="811"></a>811. <tt>pair</tt> of pointers no longer works with literal 0</h3>
7036 <p><b>Section:</b> 20.3.4 [pairs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7037  <b>Submitter:</b> Doug Gregor <b>Opened:</b> 2008-03-14  <b>Last modified:</b> 2009-10-23</p>
7038 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#pairs">active issues</a> in [pairs].</p>
7039 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#pairs">issues</a> in [pairs].</p>
7040 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7041 <p><b>Discussion:</b></p>
7042 <blockquote><pre>#include &lt;utility&gt;
7043
7044 int main()
7045 {
7046    std::pair&lt;char *, char *&gt; p (0,0);
7047 }
7048 </pre></blockquote>
7049
7050 <p>
7051 I just got a bug report about that, because it's valid C++03, but not
7052 C++0x. The important realization, for me, is that the emplace
7053 proposal---which made <tt>push_back</tt> variadic, causing the <tt>push_back(0)</tt>
7054 issue---didn't cause this break in backward compatibility. The break
7055 actually happened when we added this pair constructor as part of adding
7056 rvalue references into the language, long before variadic templates or
7057 emplace came along:
7058 </p>
7059
7060 <blockquote><pre>template&lt;class U, class V&gt; pair(U&amp;&amp; x, V&amp;&amp; y);
7061 </pre></blockquote>
7062
7063 <p>
7064 Now, concepts will address this issue by constraining that <tt>pair</tt>
7065 constructor to only <tt>U</tt>'s and <tt>V</tt>'s that can properly construct "first" and
7066 "second", e.g. (from
7067 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2322.pdf">N2322</a>):
7068 </p>
7069
7070 <blockquote><pre>template&lt;class U , class V &gt;
7071 requires Constructible&lt;T1, U&amp;&amp;&gt; &amp;&amp; Constructible&lt;T2, V&amp;&amp;&gt;
7072 pair(U&amp;&amp; x , V&amp;&amp; y );
7073 </pre></blockquote>
7074
7075 <p><i>[
7076 San Francisco:
7077 ]</i></p>
7078
7079
7080 <blockquote>
7081 <p>
7082 Suggested to resolve using pass-by-value for that case.
7083 </p>
7084 <p>
7085 Side question: Should pair interoperate with tuples? Can construct a
7086 tuple of a pair, but not a pair from a two-element tuple.
7087 </p>
7088 <p>
7089 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#885">885</a>.
7090 </p>
7091 </blockquote>
7092
7093 <p><i>[
7094 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
7095 ]</i></p>
7096
7097
7098 <p><i>[
7099 2009-10 Santa Cruz:
7100 ]</i></p>
7101
7102
7103 <blockquote>
7104 Leave as open. Howard to provide wording.
7105 </blockquote>
7106
7107
7108
7109 <p><b>Proposed resolution:</b></p>
7110 <p>
7111 </p>
7112
7113
7114 <p><b>Rationale:</b></p>
7115 <p><i>[
7116 San Francisco:
7117 ]</i></p>
7118
7119
7120 <blockquote>
7121 Solved by
7122 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2770.pdf">N2770</a>.
7123 </blockquote>
7124
7125
7126
7127
7128
7129
7130
7131 <hr>
7132 <h3><a name="815"></a>815. <tt>std::function</tt> and <tt>reference_closure</tt> do not use perfect forwarding</h3>
7133 <p><b>Section:</b> 20.7.15.2.4 [func.wrap.func.inv] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7134  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2008-03-16  <b>Last modified:</b> 2009-10-23</p>
7135 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7136 <p><b>Discussion:</b></p>
7137 <p>
7138 <tt>std::function</tt> and <tt>reference_closure</tt> should use "perfect forwarding" as
7139 described in the rvalue core proposal.
7140 </p>
7141
7142 <p><i>[
7143 Sophia Antipolis:
7144 ]</i></p>
7145
7146
7147 <blockquote>
7148 According to Doug Gregor, as far as <tt>std::function</tt> is concerned, perfect
7149 forwarding can not be obtained because of type erasure. Not everyone
7150 agreed with this diagnosis of forwarding.
7151 </blockquote>
7152
7153 <p><i>[
7154 2009-05-01 Howard adds:
7155 ]</i></p>
7156
7157
7158 <blockquote>
7159 <p>
7160 Sebastian Gesemann brought to my attention that the <tt>CopyConstructible</tt>
7161 requirement on <tt>function</tt>'s <tt>ArgTypes...</tt> is an unnecessary
7162 restriction.
7163 </p>
7164
7165 <blockquote><pre>template&lt;Returnable R, <b>CopyConstructible</b>... ArgTypes&gt;
7166 class function&lt;R(ArgTypes...)&gt;
7167 ...
7168 </pre></blockquote>
7169
7170 <p>
7171 On further investigation, this complaint seemed to be the same
7172 issue as this one.  I believe the reason <tt>CopyConstructible</tt> was put
7173 on <tt>ArgTypes</tt> in the first place was because of the nature of the
7174 <i>invoke</i> member:
7175 </p>
7176
7177 <blockquote><pre>template&lt;class R, class ...ArgTypes&gt;
7178 R
7179 function&lt;R(ArgTypes...)&gt;::operator()(ArgTypes... arg) const
7180 {
7181     if (f_ == 0)
7182         throw bad_function_call();
7183     return (*f_)(arg...);
7184 }
7185 </pre></blockquote>
7186
7187 <p>
7188 However now with rvalue-refs, "by value" no longer implies <tt>CopyConstructible</tt>
7189 (as Sebastian correctly points out).  If rvalue arguments are supplied, <tt>MoveConstructible</tt>
7190 is sufficient.  Furthermore, the constraint need not be applied in <tt>function</tt>
7191 if I understand correctly.  Rather the client must apply the proper constraints
7192 at the call site.  Therefore, at the very least, I recommend that <tt>CopyConstructible</tt>
7193 be removed from the template class <tt>function</tt>.
7194 </p>
7195
7196 <p>
7197 Furthermore we need to mandate that the <i>invoker</i> is coded as:
7198 </p>
7199
7200 <blockquote><pre>template&lt;class R, class ...ArgTypes&gt;
7201 R
7202 function&lt;R(ArgTypes...)&gt;::operator()(ArgTypes... arg) const
7203 {
7204     if (f_ == 0)
7205         throw bad_function_call();
7206     return (*f_)(<b>std::forward&lt;ArgTypes&gt;(</b>arg<b>)</b>...);
7207 }
7208 </pre></blockquote>
7209
7210 <p>
7211 Note that <tt>ArgTypes&amp;&amp;</tt> (the "perfect forwarding signature") is not
7212 appropriate here as this is not a deduced context for <tt>ArgTypes</tt>.  Instead
7213 the client's arguments must implicitly convert to the non-deduced <tt>ArgType</tt>
7214 type.  Catching these arguments by value makes sense to enable decay.
7215 </p>
7216
7217 <p>
7218 Next <tt>forward</tt> is used to move the <tt>ArgTypes</tt> as efficiently as
7219 possible, and also with minimum requirements (not <tt>CopyConstructible</tt>)
7220 to the type-erased functor.  For object types, this will be a <tt>move</tt>.  For
7221 reference type <tt>ArgTypes</tt>, this will be a copy.  The end result <em>must</em> be
7222 that the following is a valid program:
7223 </p>
7224
7225 <blockquote><pre>#include &lt;functional&gt;
7226 #include &lt;memory&gt;
7227 #include &lt;cassert&gt;
7228
7229 std::unique_ptr&lt;int&gt;
7230 f(std::unique_ptr&lt;int&gt; p, int&amp; i)
7231 {
7232     ++i;
7233     return std::move(p);
7234 }
7235
7236 int main()
7237 {
7238     int i = 2;
7239     std::function&lt;std::unique_ptr&lt;int&gt;(std::unique_ptr&lt;int&gt;,
7240                                        int&amp;&gt; g(f);
7241     std::unique_ptr&lt;int&gt; p = g(std::unique_ptr&lt;int&gt;(new int(1)), i);
7242     assert(*p == 1);
7243     assert(i == 3);
7244 }
7245 </pre></blockquote>
7246
7247 <p><i>[
7248 Tested in pre-concepts rvalue-ref-enabled compiler.
7249 ]</i></p>
7250
7251
7252 <p>
7253 In the example above, the first <tt>ArgType</tt> is <tt>unique_ptr&lt;int&gt;</tt>
7254 and the second <tt>ArgType</tt> is <tt>int&amp;</tt>.  Both <em>must</em> work!
7255 </p>
7256
7257 </blockquote>
7258
7259 <p><i>[
7260 2009-05-27 Daniel adds:
7261 ]</i></p>
7262
7263
7264 <blockquote>
7265 <p>
7266 in the 2009-05-01 comment of above mentioned issue Howard
7267 </p>
7268
7269 <ol type="a">
7270 <li>
7271 Recommends to replace the <tt>CopyConstructible</tt> requirement by a
7272 <tt>MoveConstructible</tt> requirement
7273 </li>
7274 <li>
7275 Says: "Furthermore, the constraint need not be applied in <tt>function</tt> if I
7276 understand correctly. Rather the client must apply the proper constraints
7277 at the call site"
7278 </li>
7279 </ol>
7280 <p>
7281 I'm fine with (a), but I think comment (b) is incorrect, at least in the
7282 sense I read these sentences. Let's look at Howard's example code:
7283 </p>
7284
7285 <blockquote><pre>function&lt;R(ArgTypes...)&gt;::operator()(ArgTypes... arg) const
7286 {
7287    if (f_ == 0)
7288        throw bad_function_call();
7289    return (*f_)(std::forward&lt;ArgTypes&gt;(arg)...);
7290 }
7291 </pre></blockquote>
7292
7293 <p>
7294 In the constrained scope of this <tt>operator()</tt> overload the expression
7295 "<tt>(*f_)(std::forward&lt;ArgTypes&gt;(arg)...)</tt>" must be valid. How can it
7296 do so, if <tt>ArgTypes</tt> aren't at least <tt>MoveConstructible</tt>?
7297 </p>
7298 </blockquote>
7299
7300 <p><i>[
7301 2009-07 Frankfurt:
7302 ]</i></p>
7303
7304
7305 <blockquote>
7306 Leave this open and wait until concepts are removed from the Working
7307 Draft so that we know how to write the proposed resolution in terms of
7308 diffs to otherwise stable text.
7309 </blockquote>
7310
7311 <p><i>[
7312 2009-10 Santa Cruz:
7313 ]</i></p>
7314
7315
7316 <blockquote>
7317 Leave as open. Howard to provide wording. Howard welcomes any help.
7318 </blockquote>
7319
7320
7321
7322 <p><b>Proposed resolution:</b></p>
7323 <p>
7324 </p>
7325
7326
7327
7328
7329
7330 <hr>
7331 <h3><a name="816"></a>816. Should <tt>bind()</tt>'s returned functor have a nofail copy ctor when <tt>bind()</tt> is nofail?</h3>
7332 <p><b>Section:</b> 20.7.11.1.3 [func.bind.bind] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
7333  <b>Submitter:</b> Stephan T. Lavavej <b>Opened:</b> 2008-02-08  <b>Last modified:</b> 2009-11-07</p>
7334 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#func.bind.bind">active issues</a> in [func.bind.bind].</p>
7335 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#func.bind.bind">issues</a> in [func.bind.bind].</p>
7336 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
7337 <p><b>Discussion:</b></p>
7338 <p>
7339 Library Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a> notes that <tt>bind(f, t1, ..., tN)</tt>
7340 should be nofail when <tt>f, t1, ..., tN</tt> have nofail copy ctors.
7341 </p>
7342 <p>
7343 However, no guarantees are provided for the copy ctor of the functor
7344 returned by <tt>bind()</tt>.  (It's guaranteed to have a copy ctor, which can
7345 throw implementation-defined exceptions: <tt>bind()</tt> returns a forwarding
7346 call wrapper, TR1 3.6.3/2.  A forwarding call wrapper is a call wrapper,
7347 TR1 3.3/4.  Every call wrapper shall be CopyConstructible, TR1 3.3/4.
7348 Everything without an exception-specification may throw
7349 implementation-defined exceptions unless otherwise specified, C++03
7350 17.4.4.8/3.)
7351 </p>
7352 <p>
7353 Should the nofail guarantee requested by Library Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#527">527</a> be extended
7354 to cover both calling <tt>bind()</tt> and copying the returned functor?
7355 </p>
7356
7357 <p><i>[
7358 Howard adds:
7359 ]</i></p>
7360
7361
7362 <blockquote>
7363 <tt>tuple</tt> construction should probably have a similar guarantee.
7364 </blockquote>
7365
7366 <p><i>[
7367 San Francisco:
7368 ]</i></p>
7369
7370
7371 <blockquote>
7372 Howard to provide wording.
7373 </blockquote>
7374
7375 <p><i>[
7376 Post Summit, Anthony provided wording.
7377 ]</i></p>
7378
7379
7380 <p><i>[
7381 Batavia (2009-05):
7382 ]</i></p>
7383
7384 <blockquote>
7385 Part of all of this issue appears to be rendered moot
7386 by the proposed resolution to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a> (q.v.).
7387 We recommend the issues be considered simultaneously
7388 (or possibly even merged)
7389 to ensure there is no overlap.
7390 Move to Open, and likewise for issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>.
7391 </blockquote>
7392
7393 <p><i>[
7394 2009-07 Frankfurt:
7395 ]</i></p>
7396
7397
7398 <blockquote>
7399 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a> (see below). Leave Open.
7400 </blockquote>
7401
7402 <p><i>[
7403 2009-10 Santa Cruz:
7404 ]</i></p>
7405
7406
7407 <blockquote>
7408 Move to Ready. Decoupling from issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#817">817</a>.
7409 </blockquote>
7410
7411
7412
7413 <p><b>Proposed resolution:</b></p>
7414 <p>
7415 Add a new sentence to the end of paragraphs 2 and 4 of 20.7.11.1.3 [func.bind.bind]:
7416 </p>
7417
7418 <blockquote>
7419 <p>
7420 -2- <i>Returns:</i> A forwarding call wrapper <tt>g</tt> with a weak result type (20.6.2). The effect of <tt>g(u1, u2,
7421 ..., uM)</tt> shall be <tt><i>INVOKE</i>(f, v1, v2, ..., vN, Callable&lt;F cv,V1, V2, ..., VN&gt;::result_type)</tt>, where <i>cv</i>
7422 represents the <i>cv</i>-qualifiers of <tt>g</tt> and the values and types of the bound arguments
7423 <tt>v1, v2, ..., vN</tt> are determined as specified below.
7424 <ins>The copy constructor and move constructor of the forwarding call wrapper shall throw an
7425 exception if and only if the corresponding constructor of <tt>F</tt> or any of the types
7426 in <tt>BoundArgs...</tt> throw an exception.</ins>
7427 </p>
7428 <p>...</p>
7429 <p>
7430 -5- <i>Returns:</i> A forwarding call wrapper <tt>g</tt> with a nested type <tt>result_type</tt> defined as a synonym
7431 for <tt>R</tt>. The effect of <tt>g(u1, u2, ..., uM)</tt> shall be <tt><i>INVOKE</i>(f, v1, v2, ..., vN, R)</tt>, where the
7432 values and types of the bound arguments <tt>v1, v2, ..., vN</tt> are determined as specified below.
7433 <ins>The copy constructor and move constructor of the forwarding call wrapper shall throw an
7434 exception if and only if the corresponding constructor of <tt>F</tt> or any of the types
7435 in <tt>BoundArgs...</tt> throw an exception.</ins>
7436 </p>
7437
7438 </blockquote>
7439
7440
7441
7442
7443
7444 <hr>
7445 <h3><a name="817"></a>817. <tt>bind</tt> needs to be moved</h3>
7446 <p><b>Section:</b> 20.7.11.1.3 [func.bind.bind] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7447  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2008-03-17  <b>Last modified:</b> 2009-11-08</p>
7448 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#func.bind.bind">active issues</a> in [func.bind.bind].</p>
7449 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#func.bind.bind">issues</a> in [func.bind.bind].</p>
7450 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7451 <p><b>Discussion:</b></p>
7452 <p><b>Addresses US 72, JP 38 and DE 21</b></p>
7453
7454 <p>
7455 The functor returned by <tt>bind()</tt> should have a move constructor that
7456 requires only move construction of its contained functor and bound arguments.
7457 That way move-only functors can be passed to objects such as <tt>thread</tt>.
7458 </p>
7459 <p>
7460 This issue is related to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#816">816</a>.
7461 </p>
7462
7463 <p>
7464 US 72:
7465 </p>
7466
7467 <blockquote>
7468 <tt>bind</tt> should support move-only functors and bound arguments.
7469 </blockquote>
7470
7471 <p>
7472 JP 38:
7473 </p>
7474
7475 <blockquote>
7476 <p>
7477 add the move requirement for bind's return type.
7478 </p>
7479 <p>
7480 For example, assume following <tt>th1</tt> and <tt>th2</tt>,
7481 </p>
7482
7483 <blockquote><pre>void f(vector&lt;int&gt; v) { }
7484
7485 vector&lt;int&gt; v{ ... };
7486 thread th1([v]{ f(v); });
7487 thread th2(bind(f, v));
7488 </pre></blockquote>
7489
7490 <p>
7491 When function object are set to thread, <tt>v</tt> is moved to <tt>th1</tt>'s lambda
7492 expression in a Move Constructor of lambda expression because <tt>th1</tt>'s lambda
7493 expression has a Move Constructor. But <tt>bind</tt> of <tt>th2</tt>'s
7494 return type doesn't have the requirement of Move, so it may not
7495 moved but copied.
7496 </p>
7497 <p>
7498 Add the requirement of move to get rid of this useless copy.
7499 </p>
7500 <p>
7501 And also, add the <tt>MoveConstructible</tt> as well as <tt>CopyConstructible</tt>.
7502 </p>
7503 </blockquote>
7504
7505 <p>
7506 DE 21
7507 </p>
7508
7509 <blockquote>
7510 The specification for bind claims twice that "the values and types for
7511 the bound arguments v1, v2, ..., vN are determined as specified below".
7512 No such specification appears to exist.
7513 </blockquote>
7514
7515 <p><i>[
7516 San Francisco:
7517 ]</i></p>
7518
7519
7520 <blockquote>
7521 Howard to provide wording.
7522 </blockquote>
7523
7524 <p><i>[
7525 Post Summit Alisdair and Howard provided wording.
7526 ]</i></p>
7527
7528
7529 <blockquote>
7530 <p>
7531 Several issues are being combined in this resolution.  They are all touching the
7532 same words so this is an attempt to keep one issue from stepping on another, and
7533 a place to see the complete solution in one place.
7534 </p>
7535
7536 <ol>
7537 <li>
7538 <tt>bind</tt> needs to be "moved".
7539 </li>
7540 <li>
7541 20.7.11.1.3 [func.bind.bind]/p3, p6 and p7 were accidently removed from N2798.
7542 </li>
7543 <li>
7544 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a> argues for a way to pass by &amp;&amp; for
7545 efficiency but retain the decaying behavior of pass by value for the
7546 <tt>thread</tt> constructor.  That same solution is applicable here.
7547 </li>
7548 </ol>
7549 </blockquote>
7550
7551 <p><i>[
7552 Batavia (2009-05):
7553 ]</i></p>
7554
7555 <blockquote>
7556 <p>
7557 We were going to recommend moving this issue to Tentatively Ready
7558 until we noticed potential overlap with issue 816 (q.v.).
7559 </p>
7560 <p>
7561 Move to Open,
7562 and recommend both issues be considered together
7563 (and possibly merged).
7564 </p>
7565 </blockquote>
7566
7567 <p><i>[
7568 2009-07 Frankfurt:
7569 ]</i></p>
7570
7571
7572 <blockquote>
7573 The proposed resolution uses concepts. Leave Open.
7574 </blockquote>
7575
7576 <p><i>[
7577 2009-10 Santa Cruz:
7578 ]</i></p>
7579
7580
7581 <blockquote>
7582 Leave as Open. Howard to provide deconceptified wording.
7583 </blockquote>
7584
7585 <p><i>[
7586 2009-11-07 Howard updates wording.
7587 ]</i></p>
7588
7589
7590
7591
7592 <p><b>Proposed resolution:</b></p>
7593 <p>
7594 Change 20.7 [function.objects] p2:
7595 </p>
7596
7597 <blockquote><pre>template&lt;class F<del>n</del>, class... <del>Types</del> <ins>BoundArgs</ins>&gt;
7598   <i>unspecified</i> bind(F<del>n</del><ins>&amp;&amp;</ins>, <del>Types</del> <ins>BoundArgs&amp;&amp;</ins>...);
7599 template&lt;class R, class F<del>n</del>, class... <del>Types</del> <ins>BoundArgs</ins>&gt;
7600   <i>unspecified</i> bind(F<del>n</del><ins>&amp;&amp;</ins>, <del>Types</del> <ins>BoundArgs&amp;&amp;</ins>...);
7601 </pre></blockquote>
7602
7603 <p>
7604 Change 20.7.11.1.3 [func.bind.bind]:
7605 </p>
7606
7607 <blockquote>
7608 <p><ins>
7609 Within this clause:
7610 </ins></p>
7611
7612 <ul>
7613 <li><ins>
7614 Let <tt>FD</tt> be a synonym for the type <tt>decay&lt;F&gt;::type</tt>.
7615 </ins></li>
7616 <li><ins>
7617 Let <tt>fd</tt> be an lvalue of type <tt>FD</tt> constructed from
7618 <tt>std::forward&lt;F&gt;(f)</tt>.
7619 </ins></li>
7620 <li><ins>
7621 Let <tt>Ti</tt> be a synonym for the i<sup><i>th</i></sup> type in the
7622 parameter pack <tt>BoundArgs</tt>.
7623 </ins></li>
7624 <li><ins>
7625 Let <tt>TiD</tt> be a synonym for the type <tt>decay&lt;Ti&gt;::type</tt>.
7626 </ins></li>
7627 <li><ins>
7628 Let <tt>ti</tt> be the i<sup><i>th</i></sup> argument in <tt>bound_args</tt>.
7629 </ins></li>
7630 <li><ins>
7631 Let <tt>tid</tt> be an lvalue of type <tt>TiD</tt> constructed from
7632 <tt>std::forward&lt;Ti&gt;(ti)</tt>.
7633 </ins></li>
7634 </ul>
7635
7636 <pre>template&lt;class F, class... BoundArgs&gt;
7637   <i>unspecified</i> bind(F<ins>&amp;&amp;</ins> f, BoundArgs<ins>&amp;&amp;</ins>... bound_args);
7638 </pre>
7639
7640 <blockquote>
7641 <p>
7642 -1- <i>Requires:</i>
7643 <ins><tt>is_constructible&lt;FD, F&gt;::value</tt>
7644 shall be <tt>true</tt>.</ins>
7645 <ins><tt>is_constructible&lt;TiD, Ti&gt;::value</tt>
7646 shall be <tt>true</tt></ins>.
7647 <del><tt>F</tt> and each <tt>Ti</tt> in
7648 <tt>BoundArgs</tt> shall be CopyConstructible.</del>
7649 <tt><i>INVOKE</i>(f<ins>d</ins>, w1, w2, ..., wN)</tt> (20.7.2 [func.require]) shall be a valid expression for some values
7650 <i>w1, w2, ..., wN</i>, where <tt>N == sizeof...(bound_args)</tt>.
7651 </p>
7652 <p>
7653 -2- <i>Returns:</i> A forwarding call wrapper <tt>g</tt> with a weak
7654 result type (20.7.2 [func.require]). The effect of <tt>g(u1, u2,
7655 ..., uM)</tt> shall be <tt><i>INVOKE</i>(f<ins>d</ins>, v1, v2, ..., vN,
7656 result_of&lt;F<ins>D</ins> <i>cv</i> (V1, V2, ..., VN)&gt;::type)</tt>, where
7657 <i>cv</i> represents the <i>cv</i>-qualifiers of <tt>g</tt> and the
7658 values and types of the bound arguments <tt>v1, v2, ..., vN</tt> are
7659 determined as specified below.
7660 <ins>The copy constructor and move constructor of the forwarding call wrapper shall throw an
7661 exception if and only if the corresponding constructor of <tt>FD</tt> or any of the types
7662 <tt>TiD</tt> throw an exception.</ins>
7663 </p>
7664 <p>
7665 -3- <i>Throws:</i> Nothing unless the <del>copy</del>
7666 construct<ins>ion</ins><del>or</del> of
7667 <tt><del>F</del><ins>fd</ins></tt> or of one of the <ins>values
7668 <tt>tid</tt></ins> <del>types in the <tt>BoundArgs...</tt> pack
7669 expansion</del> throws an exception.
7670 </p>
7671 <p>
7672 <ins>
7673 <i>Remarks:</i> The <i>unspecified</i> return type shall be
7674 <tt>MoveConstructible</tt>.  If all of <tt>FD</tt> and <tt>TiD</tt> are
7675 <tt>CopyConstructible</tt> then the <i>unspecified</i> return type shall
7676 be <tt>CopyConstructible</tt>. [<i>Note:</i> This implies that all of
7677 <tt>FD</tt> and <tt>TiD</tt> shall be <tt>MoveConstructible</tt> &#8212;
7678 <i>end note</i>]
7679 </ins>
7680 </p>
7681 </blockquote>
7682
7683 <pre>template&lt;class R, class F, class... BoundArgs&gt;
7684   <i>unspecified</i> bind(F<ins>&amp;&amp;</ins> f, BoundArgs<ins>&amp;&amp;</ins>... bound_args);
7685 </pre>
7686
7687 <blockquote>
7688 <p>
7689 -4- <i>Requires:</i>
7690 <ins><tt>is_constructible&lt;FD, F&gt;::value</tt>
7691 shall be <tt>true</tt>.</ins>
7692 <ins><tt>is_constructible&lt;TiD, Ti&gt;::value</tt>
7693 shall be <tt>true</tt></ins>.
7694 <del><tt>F</tt> and each <tt>Ti</tt> in
7695 <tt>BoundArgs</tt> shall be CopyConstructible.</del>
7696 <tt><i>INVOKE</i>(f<ins>d</ins>, w1,
7697 w2, ..., wN)</tt> shall be a valid expression for some values <i>w1, w2,
7698 ..., wN</i>, where <tt>N == sizeof...(bound_args)</tt>.
7699 </p>
7700 <p>
7701 -5- <i>Returns:</i> A forwarding call wrapper <tt>g</tt> with a nested
7702 type <tt>result_type</tt> defined as a synonym for <tt>R</tt>. The
7703 effect of <tt>g(u1, u2, ..., uM)</tt> shall be <tt><i>INVOKE</i>(f<ins>d</ins>, v1,
7704 v2, ..., vN, R)</tt>, where the values and types of the bound arguments
7705 <tt>v1, v2, ..., vN</tt> are determined as specified below.
7706 <ins>The copy constructor and move constructor of the forwarding call wrapper shall throw an
7707 exception if and only if the corresponding constructor of <tt>FD</tt> or any of the types
7708 <tt>TiD</tt> throw an exception.</ins>
7709 </p>
7710 <p>
7711 -6- <i>Throws:</i> Nothing unless the <del>copy</del>
7712 construct<ins>ion</ins><del>or</del> of
7713 <tt><del>F</del><ins>fd</ins></tt> or of one of the <ins>values
7714 <tt>tid</tt></ins> <del>types in the <tt>BoundArgs...</tt> pack
7715 expansion</del> throws an exception.
7716 </p>
7717 <p>
7718 <ins>
7719 <i>Remarks:</i> The <i>unspecified</i> return type shall be
7720 <tt>MoveConstructible</tt>.  If all of <tt>FD</tt> and <tt>TiD</tt> are
7721 <tt>CopyConstructible</tt> then the <i>unspecified</i> return type shall
7722 be <tt>CopyConstructible</tt>. [<i>Note:</i> This implies that all of
7723 <tt>FD</tt> and <tt>TiD</tt> shall be <tt>MoveConstructible</tt> &#8212;
7724 <i>end note</i>]
7725 </ins>
7726 </p>
7727 </blockquote>
7728
7729 <p>
7730 -7- The values of the <i>bound arguments</i> <tt>v1, v2, ..., vN</tt> and
7731 their corresponding types <tt>V1, V2, ..., VN</tt> depend on the type<ins>s
7732 <tt>TiD</tt> derived from</ins>
7733 <del>of the corresponding argument <tt>ti</tt> in <tt>bound_args</tt> of type
7734 <tt>Ti</tt> in <tt>BoundArgs</tt> in</del>
7735 the call to <tt>bind</tt> and the
7736 <i>cv</i>-qualifiers <i>cv</i> of the call wrapper <tt>g</tt> as
7737 follows:
7738 </p>
7739
7740 <ul>
7741 <li>
7742 if <tt><del>ti</del> <ins>TiD</ins></tt> <del>is of</del> <ins>has</ins>
7743 type <tt>reference_wrapper&lt;T&gt;</tt> the argument is
7744 <tt>ti<ins>d</ins>.get()</tt> and its type <tt>Vi</tt> is
7745 <tt>T&amp;</tt>;
7746 </li>
7747 <li>
7748 if the value of
7749 <tt>std::is_bind_expression&lt;Ti<ins>D</ins>&gt;::value</tt> is
7750 <tt>true</tt> the argument is <tt>ti<ins>d</ins>(u1, u2, ..., uM)</tt>
7751 and its type <tt>Vi</tt> is <tt>result_of&lt;Ti<ins>D</ins> <i>cv</i>
7752 (U1&amp;, U2&amp;, ..., UM&amp;)&gt;::type</tt>;
7753 </li>
7754 <li>
7755 if the value <tt>j</tt> of
7756 <tt>std::is_placeholder&lt;Ti<ins>D</ins>&gt;::value</tt> is not zero
7757 the argument is <tt>std::forward&lt;Uj&gt;(uj)</tt> and its type
7758 <tt>Vi</tt> is <tt>Uj&amp;&amp;</tt>;
7759 </li>
7760 <li>
7761 otherwise the value is <tt>ti<ins>d</ins></tt> and its type <tt>Vi</tt>
7762 is <tt>Ti<ins>D</ins> <i>cv</i> &amp;</tt>.
7763 </li>
7764 </ul>
7765
7766 </blockquote>
7767
7768
7769
7770
7771
7772
7773 <hr>
7774 <h3><a name="819"></a>819. rethrow_if_nested</h3>
7775 <p><b>Section:</b> 18.8.6 [except.nested] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7776  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2008-03-25  <b>Last modified:</b> 2009-10-23</p>
7777 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#except.nested">active issues</a> in [except.nested].</p>
7778 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#except.nested">issues</a> in [except.nested].</p>
7779 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7780 <p><b>Discussion:</b></p>
7781 <p>
7782 Looking at the wording I submitted for <tt>rethrow_if_nested</tt>, I don't think I
7783 got it quite right.
7784 </p>
7785
7786 <p>
7787 The current wording says:
7788 </p>
7789
7790 <blockquote>
7791 <pre>template &lt;class E&gt; void rethrow_if_nested(const E&amp; e);
7792 </pre>
7793 <blockquote>
7794 <p>
7795 <i>Effects:</i> Calls <tt>e.rethrow_nested()</tt> only if <tt>e</tt>
7796 is publicly derived from <tt>nested_exception</tt>.
7797 </p>
7798 </blockquote>
7799 </blockquote>
7800
7801 <p>
7802 This is trying to be a bit subtle, by requiring <tt>e</tt> (not <tt>E</tt>) to be publicly
7803 derived from <tt>nested_exception</tt> the idea is that a <tt>dynamic_cast</tt> would be
7804 required to be sure.  Unfortunately, if <tt>e</tt> is dynamically but not statically
7805 derived from <tt>nested_exception</tt>, <tt>e.rethrow_nested()</tt> is ill-formed.
7806 </p>
7807
7808 <p><i>[
7809 San Francisco:
7810 ]</i></p>
7811
7812
7813 <blockquote>
7814 Alisdair was volunteered to provide wording.
7815 </blockquote>
7816
7817 <p><i>[
7818 2009-10 Santa Cruz:
7819 ]</i></p>
7820
7821
7822 <blockquote>
7823 Leave as Open. Alisdair to provide wording.
7824 </blockquote>
7825
7826
7827
7828 <p><b>Proposed resolution:</b></p>
7829
7830
7831
7832
7833
7834 <hr>
7835 <h3><a name="834"></a>834. Unique_ptr::pointer requirements underspecified</h3>
7836 <p><b>Section:</b> 20.8.14.2 [unique.ptr.single] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7837  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-05-14  <b>Last modified:</b> 2009-10-26</p>
7838 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unique.ptr.single">active issues</a> in [unique.ptr.single].</p>
7839 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unique.ptr.single">issues</a> in [unique.ptr.single].</p>
7840 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7841 <p><b>Discussion:</b></p>
7842 <p>
7843 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#673">673</a> (including recent updates by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#821">821</a>) proposes a useful
7844 extension point for <tt>unique_ptr</tt> by granting support for an optional
7845 <tt>deleter_type::pointer</tt> to act as pointer-like replacement for <tt>element_type*</tt>
7846 (In the following: <tt>pointer</tt>).
7847 </p>
7848 <p>
7849 Unfortunately no requirements are specified for the type <tt>pointer</tt> which has
7850 impact on at least two key features of <tt>unique_ptr</tt>:
7851 </p>
7852
7853 <ol>
7854 <li>Operational fail-safety.</li>
7855 <li>(Well-)Definedness of expressions.</li>
7856 </ol>
7857
7858 <p>
7859 <tt>Unique_ptr</tt> specification makes great efforts to require that essentially *all*
7860 operations cannot throw and therefore adds proper wording to the affected
7861 operations of the deleter as well. If user-provided <tt>pointer</tt>-emulating types
7862 ("smart pointers") will be allowed, either *all* throw-nothing clauses have to
7863 be replaced by weaker "An exception is thrown only if <tt>pointer</tt>'s {op} throws
7864 an exception"-clauses or it has to be said explicitly that all used
7865 operations of
7866 <tt>pointer</tt> are required *not* to throw. I understand the main focus of <tt>unique_ptr</tt>
7867 to be as near as possible to the advantages of native pointers which cannot
7868 fail and thus strongly favor the second choice. Also, the alternative position
7869 would make it much harder to write safe and simple template code for
7870 <tt>unique_ptr</tt>. Additionally, I assume that a general statement need to be given
7871 that all of the expressions of <tt>pointer</tt> used to define semantics are required to
7872 be well-formed and well-defined (also as back-end for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#762">762</a>).
7873 </p>
7874
7875 <p><i>[
7876 Sophia Antipolis:
7877 ]</i></p>
7878
7879
7880 <blockquote>
7881 <p>
7882 Howard: We maybe need a core concept <tt>PointerLike</tt>, but we don't need the
7883 arithmetic (see <tt>shared_ptr</tt> vs. <tt>vector&lt;T&gt;::iterator</tt>.
7884 </p>
7885 <p>
7886 Howard will go through and enumerate the individual requirements wrt. <tt>pointer</tt> for each member function.
7887 </p>
7888 </blockquote>
7889
7890 <p><i>[
7891 2009-07 Frankfurt:
7892 ]</i></p>
7893
7894
7895 <blockquote>
7896 Move to Ready.
7897 </blockquote>
7898
7899 <p><i>[
7900 2009-10-15 Alisdair pulls from Ready:
7901 ]</i></p>
7902
7903
7904 <blockquote>
7905 <p>
7906 I hate to pull an issue out of Ready status, but I don't think 834 is
7907 fully baked yet.
7908 </p>
7909
7910 <p>
7911 For reference the proposed resolution is to add the following words:
7912 </p>
7913
7914 <blockquote>
7915 <tt>unique_ptr&lt;T, D&gt;::pointer</tt>'s operations shall be
7916 well-formed, shall have well defined behavior, and shall not throw
7917 exceptions.
7918 </blockquote>
7919
7920 <p>
7921 This leaves me with a big question : which operations?
7922 </p>
7923
7924 <p>
7925 Are all pointer operations required to be nothrow, including operations
7926 that have nothing to do with interactions with <tt>unique_ptr</tt>?  This was
7927 much simpler with concepts where we could point to operations within a
7928 certain concept, and so nail down the interactions.
7929 </p>
7930 </blockquote>
7931
7932 <p><i>[
7933 2009-10-15 Daniel adds:
7934 ]</i></p>
7935
7936
7937 <blockquote>
7938 I volunteer to prepare a more fine-grained solution, but I would like
7939 to ask for feedback that helps me doing so. If this question is asked
7940 early in the meeting I might be able to fix it within the week, but I
7941 cannot promise that now.
7942 </blockquote>
7943
7944 <p><i>[
7945 2009-10 Santa Cruz:
7946 ]</i></p>
7947
7948
7949 <blockquote>
7950 Leave in open. Daniel to provide wording as already suggested.
7951 </blockquote>
7952
7953
7954
7955
7956 <p><b>Proposed resolution:</b></p>
7957 <p>
7958 Add the following sentence just at the end of the newly proposed
7959 20.8.14.2 [unique.ptr.single]/p. 3:
7960 </p>
7961
7962 <blockquote>
7963 <tt>unique_ptr&lt;T, D&gt;::pointer</tt>'s operations shall be well-formed, shall have well
7964 defined behavior, and shall not throw exceptions.
7965 </blockquote>
7966
7967
7968
7969
7970
7971 <hr>
7972 <h3><a name="835"></a>835. tying two streams together (correction to DR 581)</h3>
7973 <p><b>Section:</b> 27.5.4.2 [basic.ios.members] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
7974  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2008-05-17  <b>Last modified:</b> 2009-10-20</p>
7975 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#basic.ios.members">active issues</a> in [basic.ios.members].</p>
7976 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#basic.ios.members">issues</a> in [basic.ios.members].</p>
7977 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
7978 <p><b>Discussion:</b></p>
7979        <p>
7980
7981 The fix for
7982 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#581">581</a>,
7983 now integrated into the working paper, overlooks a couple of minor
7984 problems.
7985
7986        </p>
7987        <p>
7988
7989 First, being an unformatted function once again, <code>flush()</code>
7990 is required to create a sentry object whose constructor must, among
7991 other things, flush the tied stream. When two streams are tied
7992 together, either directly or through another intermediate stream
7993 object, flushing one will also cause a call to <code>flush()</code> on
7994 the other tied stream(s) and vice versa, ad infinitum. The program
7995 below demonstrates the problem.
7996
7997        </p>
7998        <p>
7999
8000 Second, as Bo Persson notes in his
8001 comp.lang.c++.moderated <a href="http://groups.google.com/group/comp.lang.c++.moderated/tree/browse_frm/thread/f2187794e9cc036d/305df31dc583054a">post</a>,
8002 for streams with the <code>unitbuf</code> flag set such
8003 as <code>std::stderr</code>, the destructor of the sentry object will
8004 again call <code>flush()</code>. This seems to create an infinite
8005 recursion for <code>std::cerr &lt;&lt; std::flush;</code>
8006
8007        </p>
8008        <blockquote>
8009            <pre>#include &lt;iostream&gt;
8010
8011 int main ()
8012 {
8013    std::cout.tie (&amp;std::cerr);
8014    std::cerr.tie (&amp;std::cout);
8015    std::cout &lt;&lt; "cout\n";
8016    std::cerr &lt;&lt; "cerr\n";
8017 }
8018 </pre>
8019 </blockquote>
8020
8021 <p><i>[
8022 Batavia (2009-05):
8023 ]</i></p>
8024
8025 <blockquote>
8026 We agree with the proposed resolution.
8027 Move to Review.
8028 </blockquote>
8029
8030 <p><i>[
8031 2009-05-26 Daniel adds:
8032 ]</i></p>
8033
8034
8035 <blockquote>
8036 <p>
8037 I think that the most recently suggested change in
8038 27.7.2.4 [ostream::sentry] need some further word-smithing. As
8039 written, it would make the behavior undefined, if under
8040 conditions when <tt>pubsync()</tt> should be called, but when
8041 in this scenario <tt>os.rdbuf()</tt> returns 0.
8042 </p>
8043 <p>
8044 This case is explicitly handled in <tt>flush()</tt> and needs to be
8045 taken care of. My suggested fix is:
8046 </p>
8047
8048 <blockquote>
8049 If <tt>((os.flags() &amp; ios_base::unitbuf) &amp;&amp; !uncaught_exception()</tt>
8050 <ins><tt>&amp;&amp; os.rdbuf() != 0</tt></ins>) is true, calls <del><tt>os.flush()</tt></del>
8051 <ins><tt>os.rdbuf()-&gt;pubsync()</tt></ins>.
8052 </blockquote>
8053
8054 <p>
8055 Two secondary questions are:
8056 </p>
8057
8058 <ol>
8059 <li>
8060 Should <tt>pubsync()</tt> be invoked in any case or shouldn't a
8061 base requirement for this trial be that <tt>os.good() == true</tt>
8062 as required in the original <tt>flush()</tt> case?
8063 </li>
8064 <li>
8065 Since <tt>uncaught_exception()</tt> is explicitly tested, shouldn't
8066 a return value of -1 of <tt>pubsync()</tt> produce <tt>setstate(badbit)</tt>
8067 (which may throw <tt>ios_base::failure</tt>)?
8068 </li>
8069 </ol>
8070 </blockquote>
8071
8072 <p><i>[
8073 2009-07 Frankfurt:
8074 ]</i></p>
8075
8076
8077 <blockquote>
8078 <p>
8079 Daniel volunteered to modify the proposed resolution to address his two questions.
8080 </p>
8081 <p>
8082 Move back to Open.
8083 </p>
8084 </blockquote>
8085
8086 <p><i>[
8087 2009-07-26 Daniel provided wording.  Moved to Review.
8088 ]</i></p>
8089
8090
8091 <p><i>[
8092 2009-10-13 Daniel adds:
8093 ]</i></p>
8094
8095
8096 <blockquote>
8097 This proposed wording is written to match the outcome
8098 of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#397">397</a>.
8099 </blockquote>
8100
8101 <p><i>[
8102 2009 Santa Cruz:
8103 ]</i></p>
8104
8105
8106 <blockquote>
8107 Move to Open. Martin to propose updated wording that will also resolve
8108 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#397">397</a> consistently.
8109 </blockquote>
8110
8111
8112
8113 <p><b>Proposed resolution:</b></p>
8114 <p><i>[
8115 based on
8116 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>
8117 numbering
8118 ]</i></p>
8119
8120
8121 <ol>
8122 <li>
8123 <p>
8124 Just before 27.5.4.2 [basic.ios.members]/2 insert a new paragraph:
8125 </p>
8126
8127 <blockquote>
8128 <ins><i>Requires:</i> If <tt>(tiestr != 0)</tt> is <tt>true</tt>, <tt>tiestr</tt> must not be reachable
8129 by traversing the linked list of tied stream objects starting from
8130 <tt>tiestr-&gt;tie()</tt>.</ins>
8131 </blockquote>
8132 </li>
8133
8134 <li>
8135 <p>
8136 Change 27.7.2.4 [ostream::sentry]/4 as indicated:
8137 </p>
8138
8139 <blockquote>
8140 If <tt>((os.flags() &amp; ios_base::unitbuf) &amp;&amp; !uncaught_exception()<ins>&amp;&amp;
8141 os.good()</ins>)</tt> is <tt>true</tt>, calls <del><tt>os.flush()</tt></del>
8142 <ins><tt>os.rdbuf()-&gt;pubsync()</tt>. If that function returns -1 sets
8143 <tt>badbit</tt> in <tt>os.rdstate()</tt> without propagating an exception</ins>.
8144 </blockquote>
8145 </li>
8146
8147 </ol>
8148
8149
8150
8151
8152
8153
8154 <hr>
8155 <h3><a name="836"></a>836.
8156        effects of <code>money_base::space</code> and
8157        <code>money_base::none</code> on <code>money_get</code>
8158    </h3>
8159 <p><b>Section:</b> 22.4.6.1.2 [locale.money.get.virtuals] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
8160  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2008-05-17  <b>Last modified:</b> 2009-10-21</p>
8161 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#locale.money.get.virtuals">issues</a> in [locale.money.get.virtuals].</p>
8162 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
8163 <p><b>Duplicate of:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#670">670</a></p>
8164 <p><b>Discussion:</b></p>
8165
8166        <p>
8167
8168 In paragraph 2, 22.4.6.1.2 [locale.money.get.virtuals] specifies the following:
8169
8170        </p>
8171        <blockquote>
8172
8173 Where <code>space</code> or <code>none</code> appears in the format
8174 pattern, except at the end, optional white space (as recognized
8175 by <code>ct.is</code>) is consumed after any required space.
8176
8177        </blockquote>
8178        <p>
8179
8180 This requirement can be (and has been) interpreted two mutually
8181 exclusive ways by different readers. One possible interpretation
8182 is that:
8183
8184        </p>
8185        <blockquote>
8186            <ol>
8187                <li>
8188
8189 where <code>money_base::space</code> appears in the format, at least
8190 one space is required, and
8191
8192                </li>
8193                <li>
8194
8195 where <code>money_base::none</code> appears in the format, space is
8196 allowed but not required.
8197
8198                </li>
8199            </ol>
8200        </blockquote>
8201        <p>
8202
8203 The other is that:
8204
8205        </p>
8206        <blockquote>
8207
8208 where either <code>money_base::space</code> or <code>money_base::none</code> appears in the format, white space is optional.
8209
8210        </blockquote>
8211
8212 <p><i>[
8213 San Francisco:
8214 ]</i></p>
8215
8216
8217 <blockquote>
8218 Martin will revise the proposed resolution.
8219 </blockquote>
8220
8221 <p><i>[
8222 2009-07 Frankfurt:
8223 ]</i></p>
8224
8225
8226 <blockquote>
8227 <p>
8228 There is a noun missing from the proposed resolution. It's not clear
8229 that the last sentence would be helpful, even if the word were not
8230 missing:
8231 </p>
8232 <blockquote>
8233 In either case, any required MISSINGWORD followed by all optional whitespace (as recognized by ct.is()) is consumed.
8234 </blockquote>
8235 <p>
8236 Strike this sentence and move to Review.
8237 </p>
8238
8239 <p><i>[
8240 Howard: done.
8241 ]</i></p>
8242
8243 </blockquote>
8244
8245 <p><i>[
8246 2009-10 Santa Cruz:
8247 ]</i></p>
8248
8249
8250 <blockquote>
8251 Move to Ready.
8252 </blockquote>
8253
8254
8255
8256    <p><b>Proposed resolution:</b></p>
8257        <p>
8258
8259 I propose to change the text to make it clear that the first
8260 interpretation is intended, that is, to make following change to
8261 22.4.6.1.2 [locale.money.get.virtuals], p2:
8262
8263        </p>
8264
8265        <blockquote>
8266
8267 When <code><ins>money_base::</ins>space</code>
8268 or <code><ins>money_base::</ins>none</code> appears <ins>as the last
8269 element </ins>in the format pattern, <del>except at the end, optional
8270 white space (as recognized by <code>ct.is</code>) is consumed after
8271 any required space.</del> <ins>no white space is consumed. Otherwise,
8272 where <code>money_base::space</code> appears in any of the initial
8273 elements of the format pattern, at least one white space character is
8274 required. Where <code>money_base::none</code> appears in any of the
8275 initial elements of the format pattern, white space is allowed but not
8276 required.</ins>
8277 If <code>(str.flags() &amp; str.showbase)</code> is <code>false</code>, ...
8278
8279        </blockquote>
8280
8281
8282
8283
8284 <hr>
8285 <h3><a name="854"></a>854. <tt>default_delete</tt> converting constructor underspecified</h3>
8286 <p><b>Section:</b> 20.8.14.1.1 [unique.ptr.dltr.dflt] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
8287  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2008-06-18  <b>Last modified:</b> 2009-10-21</p>
8288 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
8289 <p><b>Discussion:</b></p>
8290 <p>
8291 No relationship between <tt>U</tt> and <tt>T</tt> in the converting constructor for <tt>default_delete</tt> template.
8292 </p>
8293 <p>
8294 Requirements: <tt>U*</tt> is convertible to <tt>T*</tt> and <tt>has_virtual_destructor&lt;T&gt;</tt>;
8295 the latter should also become a concept.
8296 </p>
8297 <p>
8298 Rules out cross-casting.
8299 </p>
8300 <p>
8301 The requirements for <tt>unique_ptr</tt> conversions should be the same as those on the deleter.
8302 </p>
8303
8304 <p><i>[
8305 Howard adds 2008-11-26:
8306 ]</i></p>
8307
8308
8309 <blockquote>
8310 <p>
8311 I believe we need to be careful to not outlaw the following use case, and
8312 I believe the current proposed wording
8313 (<tt>requires Convertible&lt;U*, T*&gt; &amp;&amp; HasVirtualDestructor&lt;T&gt;</tt>) does so:
8314 </p>
8315
8316 <blockquote><pre>#include &lt;memory&gt;
8317
8318 int main()
8319 {
8320     std::unique_ptr&lt;int&gt; p1(new int(1));
8321     std::unique_ptr&lt;const int&gt; p2(move(p1));
8322     int i = *p2;
8323 <font color="#c80000">//    *p2 = i;  // should not compile</font>
8324 }
8325 </pre></blockquote>
8326
8327 <p>
8328 I've removed "<tt>&amp;&amp; HasVirtualDestructor&lt;T&gt;</tt>" from the
8329 <tt>requires</tt> clause in the proposed wording.
8330 </p>
8331
8332 </blockquote>
8333
8334 <p><i>[
8335 Post Summit:
8336 ]</i></p>
8337
8338
8339 <blockquote>
8340 <p>
8341 Alisdair: This issue has to stay in review pending a paper constraining
8342 <tt>unique_ptr</tt>.
8343 </p>
8344 <p>
8345 Consensus: We agree with the resolution, but <tt>unique_ptr</tt> needs
8346 to be constrained, too.
8347 </p>
8348 <p>
8349 Recommend Keep in Review.
8350 </p>
8351 </blockquote>
8352
8353 <p><i>[
8354 Batavia (2009-05):
8355 ]</i></p>
8356
8357 <blockquote>
8358 Keep in Review status for the reasons cited.
8359 </blockquote>
8360
8361 <p><i>[
8362 2009-07 Frankfurt:
8363 ]</i></p>
8364
8365
8366 <blockquote>
8367 <p>
8368 The proposed resolution uses concepts. Howard needs to rewrite the
8369 proposed resolution.
8370 </p>
8371 <p>
8372 Move back to Open.
8373 </p>
8374 </blockquote>
8375
8376 <p><i>[
8377 2009-07-26 Howard provided rewritten proposed wording and moved to Review.
8378 ]</i></p>
8379
8380
8381 <p><i>[
8382 2009-10 Santa Cruz:
8383 ]</i></p>
8384
8385
8386 <blockquote>
8387 Move to Ready.
8388 </blockquote>
8389
8390
8391
8392 <p><b>Proposed resolution:</b></p>
8393 <p>
8394 Add after 20.8.14.1.1 [unique.ptr.dltr.dflt], p1:
8395 </p>
8396
8397 <blockquote><pre>template &lt;class U&gt; default_delete(const default_delete&lt;U&gt;&amp; other);
8398 </pre>
8399 <blockquote>
8400 <p>
8401 -1- <i>Effects:</i> ...
8402 </p>
8403 <p><ins>
8404 <i>Remarks:</i> This constructor shall participate in overload resolution
8405 if and only if <tt>U*</tt> is implicitly convertible to <tt>T*</tt>.
8406 </ins></p>
8407 </blockquote>
8408 </blockquote>
8409
8410
8411
8412
8413
8414
8415 <hr>
8416 <h3><a name="860"></a>860. Floating-Point State</h3>
8417 <p><b>Section:</b> 26 [numerics] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
8418  <b>Submitter:</b> Lawrence Crowl <b>Opened:</b> 2008-06-23  <b>Last modified:</b> 2009-10-23</p>
8419 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#numerics">issues</a> in [numerics].</p>
8420 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
8421 <p><b>Discussion:</b></p>
8422 <p>
8423 There are a number of functions that affect the floating point state.
8424 These function need to be thread-safe, but I'm unsure of the right
8425 approach in the standard, as we inherit them from C.
8426 </p>
8427
8428 <p><i>[
8429 San Francisco:
8430 ]</i></p>
8431
8432
8433 <blockquote>
8434 <p>
8435 Nick: I think we already say that these functions do not introduce data
8436 races; see 17.6.5.6/20
8437 </p>
8438 <p>
8439 Pete: there's more to it than not introducing data races; are these
8440 states maintained per thread?
8441 </p>
8442 <p>
8443 Howard: 21.5/14 says that strtok and strerror are not required to avoid
8444 data races, and 20.9/2 says the same about asctime, gmtime, ctime, and
8445 gmtime.
8446 </p>
8447 <p>
8448 Nick: POSIX has a list of not-safe functions. All other functions are
8449 implicitly thread safe.
8450 </p>
8451 <p>
8452 Lawrence is to form a group between meetings to attack this issue. Nick
8453 and Tom volunteered to work with Lawrence.
8454 </p>
8455 <p>
8456 Move to Open.
8457 </p>
8458 </blockquote>
8459
8460 <p><i>[
8461 Post Summit:
8462 ]</i></p>
8463
8464
8465 <blockquote>
8466 <p>
8467 Hans: Sane oses seem ok. Sensible thing is implementable and makes sense.
8468 </p>
8469 <p>
8470 Nick: Default wording seems to cover this? Hole in POSIX, these
8471 functions need to be added to list of thread-unsafe functions.
8472 </p>
8473 <p>
8474 Lawrence: Not sufficient, not "thread-safe" per our definition, but
8475 think of state as a thread-local variable. Need something like "these
8476 functions only affect state in the current thread."
8477 </p>
8478 <p>
8479 Hans: Suggest the following wording: "The floating point environment is
8480 maintained per-thread."
8481 </p>
8482 <p>
8483 Walter: Any other examples of state being thread safe that are not
8484 already covered elsewhere?
8485 </p>
8486 <p>
8487 Have thread unsafe functions paper which needs to be updated. Should
8488 just fold in 26.3 [cfenv] functions.
8489 </p>
8490 <p>
8491 Recommend Open. Lawrence instead suggests leaving it open until we have
8492 suitable wording that may or may not include the thread local
8493 commentary.
8494 </p>
8495 </blockquote>
8496
8497 <p><i>[
8498 2009-09-23 Hans provided wording.
8499 ]</i></p>
8500
8501
8502 <blockquote>
8503 If I understand the history correctly, Nick, as the Posix liaison,
8504 should probably get a veto on this, since I think it came from Posix (?)
8505 via WG14 and should probably really be addressed there (?).  But I think
8506 we are basically in agreement that there is no other sane way to do
8507 this, and hence we don't have to worry too much about stepping on toes.
8508 As far as I can tell, this same issue also exists in the latest Posix
8509 standard (?).
8510 </blockquote>
8511
8512 <p><i>[
8513 2009-10 Santa Cruz:
8514 ]</i></p>
8515
8516
8517 <blockquote>
8518 Moved to Ready.
8519 </blockquote>
8520
8521
8522
8523 <p><b>Proposed resolution:</b></p>
8524 <p>
8525 Add at the end of 26.3.1 [cfenv.syn]:
8526 </p>
8527
8528 <blockquote>
8529 <p>
8530 2 The header defines all functions, types, and macros the same as C99 7.6.
8531 </p>
8532
8533 <p><ins>
8534 A separate floating point environment shall be maintained for each
8535 thread.  Each function accesses the environment corresponding to its
8536 calling thread.
8537 </ins></p>
8538 </blockquote>
8539
8540
8541
8542
8543
8544 <hr>
8545 <h3><a name="861"></a>861. Incomplete specification of EqualityComparable for std::forward_list</h3>
8546 <p><b>Section:</b> 23.2 [container.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
8547  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-06-24  <b>Last modified:</b> 2009-10-24</p>
8548 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.requirements">active issues</a> in [container.requirements].</p>
8549 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.requirements">issues</a> in [container.requirements].</p>
8550 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
8551 <p><b>Discussion:</b></p>
8552 <p>
8553 Table 89, Container requirements, defines <tt>operator==</tt> in terms of the container
8554 member function <tt>size()</tt> and the algorithm <tt>std::equal</tt>:
8555 </p>
8556
8557 <blockquote>
8558 <tt>==</tt> is an equivalence relation. <tt>a.size() == b.size() &amp;&amp;
8559 equal(a.begin(), a.end(), b.begin()</tt>
8560 </blockquote>
8561
8562 <p>
8563 The new container <tt>forward_list</tt> does not provide a <tt>size</tt> member function
8564 by design but does provide <tt>operator==</tt> and <tt>operator!=</tt> without specifying it's semantic.
8565 </p>
8566 <p>
8567 Other parts of the (sequence) container requirements do also depend on
8568 <tt>size()</tt>, e.g. <tt>empty()</tt>
8569 or <tt>clear()</tt>, but this issue explicitly attempts to solve the missing
8570 <tt>EqualityComparable</tt> specification,
8571 because of the special design choices of <tt>forward_list</tt>.
8572 </p>
8573 <p>
8574 I propose to apply one of the following resolutions, which are described as:
8575 </p>
8576
8577 <ol type="A">
8578 <li>
8579 Provide a definition, which is optimal for this special container without
8580 previous size test. This choice prevents two <tt>O(N)</tt> calls of <tt>std::distance()</tt>
8581 with the corresponding container ranges and instead uses a special
8582 <tt>equals</tt> implementation which takes two container ranges instead of 1 1/2.
8583 </li>
8584 <li>
8585 The simple fix where the usual test is adapted such that <tt>size()</tt> is replaced
8586 by <tt>distance</tt> with corresponding performance disadvantages.
8587 </li>
8588 </ol>
8589 <p>
8590 Both proposal choices are discussed, the preferred choice of the author is
8591 to apply (A).
8592 </p>
8593
8594 <p><i>[
8595 San Francisco:
8596 ]</i></p>
8597
8598
8599 <blockquote>
8600 <p>
8601 There's an Option C: change the requirements table to use distance().
8602 </p>
8603 <p>
8604 LWG found Option C acceptable.
8605 </p>
8606 <p>
8607 Martin will draft the wording for Option C.
8608 </p>
8609 </blockquote>
8610
8611 <p><i>[
8612 post San Francisco:
8613 ]</i></p>
8614
8615
8616 <blockquote>
8617 Martin provided wording for Option C.
8618 </blockquote>
8619
8620 <p><i>[
8621 2009-07 Frankfurt
8622 ]</i></p>
8623
8624
8625 <blockquote>
8626 <p>
8627 Other operational semantics (see, for example, Tables 82 and 83) are
8628 written in terms of a container's size() member. Daniel to update
8629 proposed resolution C.
8630 </p>
8631 <p><i>[
8632 Howard: Commented out options A and B.
8633 ]</i></p>
8634
8635 </blockquote>
8636
8637 <p><i>[
8638 2009-07-26 Daniel updated proposed resolution C.
8639 ]</i></p>
8640
8641
8642 <p><i>[
8643 2009-10 Santa Cruz:
8644 ]</i></p>
8645
8646
8647 <blockquote>
8648 Mark NAD Editorial. Addressed by
8649 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2986.pdf">N2986</a>.
8650 </blockquote>
8651
8652 <p><i>[
8653 2009-10 Santa Cruz:
8654 ]</i></p>
8655
8656
8657 <blockquote>
8658 Reopened.
8659 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2986.pdf">N2986</a>
8660 was rejected in full committee on procedural grounds.
8661 </blockquote>
8662
8663
8664
8665 <p><b>Proposed resolution:</b></p>
8666
8667
8668 <p>
8669 Option (C):
8670 </p>
8671 <blockquote>
8672
8673 <p><i>[
8674 The changes are relative to
8675 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>
8676 but concept-free.
8677 ]</i></p>
8678
8679
8680 <ol>
8681 <li>
8682 <p>
8683 In 23.2.1 [container.requirements.general] change Table 80 -- Container requirements as indicated:
8684 </p>
8685
8686 <ol type="a">
8687 <li>
8688 <p>
8689 Change the text in the Assertion/note column in the row for "<tt>X u</tt>;"
8690 as follows:
8691 </p>
8692
8693 <blockquote>
8694 post: <tt>u.<del>size() == 0</del><ins>empty() == true</ins></tt>
8695 </blockquote>
8696 </li>
8697
8698 <li>
8699 <p>
8700 Change the text in the Assertion/note column in the row for "<tt>X();</tt>"
8701 as follows:
8702 </p>
8703
8704 <blockquote>
8705 <tt>X().<del>size() == 0</del><ins>empty() == true</ins></tt>
8706 </blockquote>
8707 </li>
8708
8709 <li>
8710 <p>
8711 Change the text in the Operational Semantics column in the row for
8712 "<tt>a == b</tt>" as follows:
8713 </p>
8714 <blockquote>
8715 <tt>==</tt> is an equivalence relation.
8716 <tt><del>a.size()</del><ins>distance(a.begin(), a.end())</ins> ==
8717    <del>b.size()</del><ins>distance(b.begin(), b.end())</ins> &amp;&amp;
8718 equal(a.begin(), a.end(), b.begin())</tt>
8719 </blockquote>
8720 </li>
8721
8722 <li>
8723 <p>
8724 Change the text in the Operational Semantics column in the row for
8725 "<tt>a.size()</tt>" as follows:
8726 </p>
8727
8728 <blockquote>
8729 <tt><del>a.end() - a.begin()</del><ins>distance(a.begin(), a.end())</ins></tt>
8730 </blockquote>
8731 </li>
8732
8733 <li>
8734 <p>
8735 Change the text in the Operational Semantics column in the row for
8736 "<tt>a.max_size()</tt>" as follows:
8737 </p>
8738
8739 <blockquote>
8740 <tt><del>size()</del><ins>distance(begin(), end())</ins></tt> of the largest
8741 possible container
8742 </blockquote>
8743 </li>
8744
8745 <li>
8746 <p>
8747 Change the text in the Operational Semantics column in the row for
8748 "<tt>a.empty()</tt>" as follows:
8749 </p>
8750
8751 <blockquote>
8752 <tt><del>a.size() == 0</del><ins>a.begin() == a.end()</ins></tt>
8753 </blockquote>
8754 </li>
8755 </ol>
8756 </li>
8757
8758 <li>
8759 <p>
8760 In 23.2.1 [container.requirements.general] change Table 82 -- Allocator-aware container requirements as indicated:
8761 </p>
8762
8763 <ol type="a">
8764 <li>
8765 <p>
8766 Change the text in the Assertion/note column in the row for "<tt>X() /
8767 X u;</tt>" as follows:
8768 </p>
8769
8770 <blockquote>
8771 <i>Requires:</i> <tt>A</tt> is <tt>DefaultConstructible</tt> post: <tt><del>u.size() ==
8772 0</del><ins>u.empty() == true</ins></tt>, <tt>get_allocator() == A()</tt>
8773 </blockquote>
8774 </li>
8775
8776 <li>
8777 <p>
8778 Change the text in the Assertion/note column in the row for "<tt>X(m) /
8779 X u(m);</tt>" as follows:
8780 </p>
8781
8782 <blockquote>
8783 post: <tt><del>u.size() == 0</del><ins>u.empty() == true</ins>,
8784 get_allocator() == m</tt>
8785 </blockquote>
8786 </li>
8787 </ol>
8788 </li>
8789
8790 <li>
8791 <p>
8792 In 23.2.3 [sequence.reqmts] change Table 83 -- Sequence container requirements as indicated:
8793 </p>
8794
8795 <ol type="a">
8796 <li>
8797 <p>
8798 Change the text in the Assertion/note column in the row for "<tt>X(n,
8799 t) / X a(n, t)</tt>" as follows:
8800 </p>
8801
8802 <blockquote>
8803 post: <tt><del>size()</del><ins>distance(begin(), end())</ins> == n [..]</tt>
8804 </blockquote>
8805 </li>
8806
8807 <li>
8808 <p>
8809 Change the text in the Assertion/note column in the row for "<tt>X(i,
8810 j) / X a(i, j)</tt>" as follows:
8811 </p>
8812
8813 <blockquote>
8814 [..] post: <del><tt>size() ==</tt> distance between <tt>i</tt> and
8815 <tt>j</tt></del><ins><tt>distance(begin(), end()) == distance(i, j)</tt></ins> [..]
8816 </blockquote>
8817 </li>
8818
8819 <li>
8820 <p>
8821 Change the text in the Assertion/note column in the row for
8822 "<tt>a.clear()</tt>" as follows:
8823 </p>
8824 <blockquote>
8825 <tt><ins>a.</ins>erase(<ins>a.</ins>begin(), <ins>a.</ins>end())</tt> post:
8826 <tt><del>size() == 0</del><ins>a.empty() == true</ins></tt>
8827 </blockquote>
8828 </li>
8829 </ol>
8830 </li>
8831
8832 <li>
8833 <p>
8834 In 23.2.4 [associative.reqmts] change Table 85 -- Associative container requirements as indicated:
8835 </p>
8836
8837 <p><i>[
8838 Not every occurrence of <tt>size()</tt> was replaced, because all current
8839 associative containers
8840 have a <tt>size</tt>. The following changes ensure consistency regarding the
8841 semantics of "<tt>erase</tt>"
8842 for all tables and adds some missing objects
8843 ]</i></p>
8844
8845
8846 <ol type="a">
8847 <li>
8848 <p>
8849 Change the text in the Complexity column in the row for
8850 "<tt>a.insert(i, j)</tt>" as follows:
8851 </p>
8852 <blockquote>
8853 <tt>N log(<ins>a.</ins>size() + N)</tt> <del>(<tt>N</tt> is the distance from <tt>i</tt> to
8854 <tt>j</tt>)</del><ins> where <tt>N == distance(i, j)</tt></ins>
8855 </blockquote>
8856 </li>
8857
8858 <li>
8859 <p>
8860 Change the text in the Complexity column in the row for
8861 "<tt>a.erase(k)</tt>" as follows:
8862 </p>
8863 <blockquote>
8864 <tt>log(<ins>a.</ins>size()) + <ins>a.</ins>count(k)</tt>
8865 </blockquote>
8866 </li>
8867
8868 <li>
8869 <p>
8870 Change the text in the Complexity column in the row for
8871 "<tt>a.erase(q1, q2)</tt>" as follows:
8872 </p>
8873
8874 <blockquote>
8875 <tt>log(<ins>a.</ins>size()) + N</tt> where <tt>N</tt> <del>is the distance from <tt>q1</tt>
8876 to <tt>q2</tt></del>
8877    <ins><tt>== distance(q1, q2)</tt></ins>.
8878 </blockquote>
8879 </li>
8880
8881 <li>
8882 <p>
8883 Change the text in the Assertion/note column in the row for
8884 "<tt>a.clear()</tt>" as follows:
8885 </p>
8886
8887 <blockquote>
8888 <tt><ins>a.</ins>erase(a.begin(),a.end())</tt> post: <tt><del>size() ==
8889 0</del><ins>a.empty() == true</ins></tt>
8890 </blockquote>
8891 </li>
8892
8893 <li>
8894 <p>
8895 Change the text in the Complexity column in the row for "<tt>a.clear()</tt>"
8896 as follows:
8897 </p>
8898
8899 <blockquote>
8900 linear in <tt><ins>a.</ins>size()</tt>
8901 </blockquote>
8902 </li>
8903
8904 <li>
8905 <p>
8906 Change the text in the Complexity column in the row for
8907 "<tt>a.count(k)</tt>" as follows:
8908 </p>
8909
8910 <blockquote>
8911 <tt>log(<ins>a.</ins>size()) + <ins>a.</ins>count(k)</tt>
8912 </blockquote>
8913 </li>
8914 </ol>
8915 </li>
8916
8917 <li>
8918 <p>
8919 In 23.2.5 [unord.req] change Table 87 -- Unordered associative container requirements as indicated:
8920 </p>
8921 <p><i>[
8922 The same rational as for Table 85 applies here
8923 ]</i></p>
8924
8925
8926 <ol type="a">
8927 <li>
8928 <p>
8929 Change the text in the Assertion/note column in the row for
8930 "<tt>a.clear()</tt>" as follows:
8931 </p>
8932
8933 <blockquote>
8934 [..] Post: <tt>a.<del>size() == 0</del><ins>empty() == true</ins></tt>
8935 </blockquote>
8936 </li>
8937 </ol>
8938 </li>
8939 </ol>
8940
8941
8942 </blockquote>
8943
8944
8945
8946
8947
8948 <hr>
8949 <h3><a name="865"></a>865. More algorithms that throw away information</h3>
8950 <p><b>Section:</b> 25.3.6 [alg.fill], 25.3.7 [alg.generate] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
8951  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-07-13  <b>Last modified:</b> 2009-10-23</p>
8952 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
8953 <p><b>Discussion:</b></p>
8954 <p>
8955 In regard to library defect <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#488">488</a> I found some more algorithms which
8956 unnecessarily throw away information. These are typically algorithms,
8957 which sequentially write into an <tt>OutputIterator</tt>, but do not return the
8958 final value of this output iterator. These cases are:
8959 </p>
8960
8961 <ol>
8962 <li>
8963 <pre>template&lt;class OutputIterator, class Size, class T&gt;
8964 void fill_n(OutputIterator first, Size n, const T&amp; value);</pre></li>
8965
8966 <li>
8967 <pre>template&lt;class OutputIterator, class Size, class Generator&gt;
8968 void generate_n(OutputIterator first, Size n, Generator gen);</pre></li>
8969 </ol>
8970 <p>
8971 In both cases the minimum requirements on the iterator are
8972 <tt>OutputIterator</tt>, which means according to the requirements of
8973 24.2.2 [output.iterators]/2 that only single-pass iterations are guaranteed.
8974 So, if users of <tt>fill_n</tt> and <tt>generate_n</tt> have *only* an <tt>OutputIterator</tt>
8975 available, they have no chance to continue pushing further values
8976 into it, which seems to be a severe limitation to me.
8977 </p>
8978
8979 <p><i>[
8980 Post Summit Daniel "conceptualized" the wording.
8981 ]</i></p>
8982
8983
8984 <p><i>[
8985 Batavia (2009-05):
8986 ]</i></p>
8987
8988 <blockquote>
8989 <p>
8990 Alisdair likes the idea, but has concerns about the specific wording
8991 about the returns clauses.
8992 </p>
8993 <p>
8994 Alan notes this is a feature request.
8995 </p>
8996 <p>
8997 Bill notes we have made similar changes to other algorithms.
8998 </p>
8999 <p>
9000 Move to Open.
9001 </p>
9002 </blockquote>
9003
9004 <p><i>[
9005 2009-07 Frankfurt
9006 ]</i></p>
9007
9008
9009 <blockquote>
9010 We have a consensus for moving forward on this issue, but Daniel needs
9011 to deconceptify it.
9012 </blockquote>
9013
9014 <p><i>[
9015 2009-07-25 Daniel provided non-concepts wording.
9016 ]</i></p>
9017
9018
9019 <p><i>[
9020 2009-10 Santa Cruz:
9021 ]</i></p>
9022
9023
9024 <blockquote>
9025 Moved to Ready.
9026 </blockquote>
9027
9028
9029
9030 <p><b>Proposed resolution:</b></p>
9031
9032 <ol>
9033 <li>
9034 <p>
9035 Replace the current declaration of <tt>fill_n</tt> in 25 [algorithms]/2, header
9036 <tt>&lt;algorithm&gt;</tt> synopsis and in 25.3.6 [alg.fill] by
9037 </p>
9038
9039 <blockquote><pre>template&lt;class OutputIterator, class Size, class T&gt;
9040   <del>void</del><ins>OutputIterator</ins> fill_n(OutputIterator first, Size n, const T&amp; value);
9041 </pre></blockquote>
9042
9043 <p>
9044 Just after the effects clause add a new returns clause saying:
9045 </p>
9046
9047 <blockquote>
9048 <ins><i>Returns:</i> For <tt>fill_n</tt> and positive <tt>n</tt>, returns <tt>first + n</tt>. Otherwise
9049 returns <tt>first</tt> for <tt>fill_n</tt>.</ins>
9050 </blockquote>
9051 </li>
9052
9053 <li>
9054 <p>
9055 Replace the current declaration of <tt>generate_n</tt> in 25 [algorithms]/2,
9056 header <tt>&lt;algorithm&gt;</tt> synopsis and in 25.3.7 [alg.generate] by
9057 </p>
9058
9059 <blockquote><pre>template&lt;class OutputIterator, class Size, class Generator&gt;
9060   <del>void</del><ins>OutputIterator</ins> generate_n(OutputIterator first, Size n, Generator gen);
9061 </pre></blockquote>
9062
9063 <p>
9064 Just after the effects clause add a new returns clause saying:
9065 </p>
9066
9067 <blockquote>
9068 <ins>For <tt>generate_n</tt> and positive <tt>n</tt>, returns <tt>first + n</tt>. Otherwise
9069 returns <tt>first</tt> for <tt>generate_n</tt>.</ins>
9070 </blockquote>
9071 </li>
9072 </ol>
9073
9074
9075
9076
9077
9078
9079
9080 <hr>
9081 <h3><a name="868"></a>868. default construction and value-initialization</h3>
9082 <p><b>Section:</b> 23 [containers] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
9083  <b>Submitter:</b> Alberto Ganesh Barbati <b>Opened:</b> 2008-07-22  <b>Last modified:</b> 2009-10-20</p>
9084 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#containers">active issues</a> in [containers].</p>
9085 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#containers">issues</a> in [containers].</p>
9086 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
9087 <p><b>Discussion:</b></p>
9088 <p>
9089 The term "default constructed" is often used in wording that predates
9090 the introduction of the concept of value-initialization. In a few such
9091 places the concept of value-initialization is more correct than the
9092 current wording (for example when the type involved can be a built-in)
9093 so a replacement is in order. Two of such places are already covered by
9094 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#867">867</a>. This issue deliberately addresses the hopefully
9095 non-controversial changes in the attempt of being approved more quickly.
9096 A few other occurrences (for example in <tt>std::tuple</tt>,
9097 <tt>std::reverse_iterator</tt> and <tt>std::move_iterator</tt>) are left to separate
9098 issues. For <tt>std::reverse_iterator</tt>, see also issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#408">408</a>. This issue is
9099 related with issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#724">724</a>.
9100 </p>
9101
9102 <p><i>[
9103 San Francisco:
9104 ]</i></p>
9105
9106
9107 <blockquote>
9108 <p>
9109 The list provided in the proposed resolution is not complete. James
9110 Dennett will review the library and provide a complete list and will
9111 double-check the vocabulary.
9112 </p>
9113 <p>
9114 This issue relates to Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#886">886</a> tuple construction
9115 </p>
9116 </blockquote>
9117
9118 <p><i>[
9119 2009-07 Frankfurt
9120 ]</i></p>
9121
9122
9123 <blockquote>
9124 <p>
9125 The proposed resolution is incomplete.
9126 </p>
9127 <p>
9128 Move to Tentatively NAD Future. Howard will contact Ganesh for wording.
9129 If wording is forthcoming, Howard will move it back to Review.
9130 </p>
9131 </blockquote>
9132
9133 <p><i>[
9134 2009-07-18 Ganesh updated the proposed wording.
9135 ]</i></p>
9136
9137
9138 <blockquote>
9139 <p>
9140 Howard:  Moved back to Review.  Note that 20.2.1 [utility.arg.requirements]
9141 refers to a section that is not in the current working paper, but does refer to
9142 a section that we expect to reappear after the de-concepts merge.  This was a
9143 point of confusion we did not recognize when we reviewed this issue in Frankfurt.
9144 </p>
9145 <p>
9146 Howard:  Ganesh also includes a survey of places in the WP surveyed for changes
9147 of this nature and purposefully <em>not</em> treated:
9148 </p>
9149
9150 <blockquote>
9151 <p>
9152 Places where changes are <u>not</u> being
9153 proposed
9154 </p>
9155 <p>
9156 In the following paragraphs, we are not proposing changes because
9157 it's not clear whether we actually prefer value-initialization over
9158 default-initialization (now partially covered by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1012">1012</a>):
9159 </p>
9160 <ul>
9161     <li><p>20.8.14.2.1 [unique.ptr.single.ctor] para 3 e 7</p></li>
9162     <li><p>24.5.1.3.1 [reverse.iter.cons] para 1</p></li>
9163     <li><p>24.5.3.3.1 [move.iter.op.const] para 1</p></li>
9164 </ul>
9165 <p>In the following paragraphs, the expression "default
9166 constructed" need not be changed, because the relevant type does
9167 not depend on a template parameter and has a user-provided
9168 constructor:</p>
9169 <ul>
9170     <li><p> [func.referenceclosure.invoke] para 12, type:
9171     reference_closure</p></li>
9172     <li><p>30.3.1.2 [thread.thread.constr] para 30, type: thread</p></li>
9173     <li><p>30.3.1.5 [thread.thread.member] para 52, type: thread_id</p></li>
9174     <li><p>30.3.2 [thread.thread.this], para 1, type: thread_id</p></li>
9175 </ul>
9176 </blockquote>
9177
9178 </blockquote>
9179
9180 <p><i>[
9181 2009-08-18 Daniel adds:
9182 ]</i></p>
9183
9184
9185 <blockquote>
9186 <p>
9187 I have no objections against the currently suggested changes, but I
9188 also cross-checked
9189 with the list regarding intentionally excluded changes, and from this
9190 I miss the discussion
9191 of
9192 </p>
9193
9194 <ol>
9195 <li>
9196 <p>
9197 21.4.1 [string.require]/2:
9198 </p>
9199
9200 <blockquote>
9201 "[..] The <tt>Allocator</tt> object used shall be a copy of the <tt>Allocator&gt;</tt>
9202 object passed to the <tt>basic_string</tt> object's
9203 constructor or, if the constructor does not take an <tt>Allocator</tt>
9204 argument, a copy of a default-constructed
9205 <tt>Allocator</tt> object."
9206 </blockquote>
9207 </li>
9208
9209 <li>
9210 <p>
9211 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>,
9212 X [rand.req.eng], Table 109, expression "<tt>T()</tt>":
9213 </p>
9214 <blockquote>
9215 Pre-/post-condition: "Creates an engine with the same initial state as
9216 all other default-constructed engines of type <tt>X</tt>."
9217 </blockquote>
9218
9219 <p>
9220 as well as in 26.5.5 [rand.predef]/1-9 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>), 26.5.7.1 [rand.util.seedseq]/3, 27.7.1.1.1 [istream.cons]/3, 27.7.2.2 [ostream.cons]/9 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>), 28.13 [re.grammar]/2, 30.3.1.4 [thread.thread.assign]/1 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>),
9221 </p>
9222 <p><i>[
9223 Candidates for the "the expression "default constructed" need not be
9224 changed" list
9225 ]</i></p>
9226
9227
9228 <p>
9229 I'm fine, if these would be added to the intentionally exclusion list,
9230 but mentioning them makes it
9231 easier for other potential reviewers to decide on the relevance or
9232 not-relevance of them for this issue.
9233 </p>
9234 </li>
9235
9236 <li>
9237 <p>
9238 I suggest to remove the reference of  [func.referenceclosure.invoke]
9239 in the "it's not clear" list, because
9240 this component does no longer exist.
9241 </p>
9242 </li>
9243
9244 <li>
9245 <p>
9246 I also suggest to add a short comment that all paragraphs in the
9247 resolution whether they refer to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a> or to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a> numbering, because e.g. "Change 23.3.2.1 [deque.cons] para 5" is an <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a> coordinate, while "Change 23.3.2.2 [deque.capacity] para 1" is an <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a> coordinate. Even better would be to use one default document
9248 for the numbering (probably <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>) and mention special cases (e.g. "Change 20.2.1 [utility.arg.requirements] para 2" as referring to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a> numbering).
9249 </p>
9250 </li>
9251 </ol>
9252
9253 </blockquote>
9254
9255 <p><i>[
9256 2009-08-18 Alisdair adds:
9257 ]</i></p>
9258
9259
9260 <blockquote>
9261 <p>
9262 I strongly believe the term "default constructed" should not appear in
9263 the library clauses unless we very clearly define a meaning for it, and
9264 I am not sure what that would be.
9265 </p>
9266
9267 <p>
9268 In those cases where we do not want to replace "default constructed"
9269 with "vale initialized" we should be using "default initialized".  If we
9270 have a term that could mean either, we reduce portability of programs.
9271 </p>
9272
9273 <p>
9274 I have not done an exhaustive review to clarify if that is a vendor
9275 freedom we have reason to support (e.g. value-init in debug,
9276 default-init in release) so I may yet be convinced that LWG has reason
9277 to define this new term of art, but generally C++ initialization is
9278 confusing enough without supporting further ill-defined terms.
9279 </p>
9280 </blockquote>
9281
9282 <p><i>[
9283 2009-10 Santa Cruz:
9284 ]</i></p>
9285
9286
9287 <blockquote>
9288 Move to Ready.
9289 </blockquote>
9290
9291
9292
9293 <p><b>Proposed resolution:</b></p>
9294 <p>
9295 Change 20.2.1 [utility.arg.requirements] para 2:
9296 </p>
9297
9298 <blockquote>
9299 In general, a default constructor is
9300 not required. Certain container class member function signatures
9301 specify <del>the default constructor</del><ins>T()</ins>
9302 as a default argument. T() shall be a well-defined expression (8.5)
9303 if one of those signatures is called using the default argument
9304 (8.3.6).
9305 </blockquote>
9306
9307 <p>
9308 Change 23.3.2.1 [deque.cons] para 5:
9309 </p>
9310
9311 <blockquote>
9312 <i>Effects:</i> Constructs a deque with n
9313 <del>default constructed</del><ins>value-initialized</ins>
9314 elements.
9315 </blockquote>
9316
9317 <p>
9318 Change 23.3.2.2 [deque.capacity] para 1:
9319 </p>
9320
9321 <blockquote>
9322 <i>Effects:</i> If sz &lt; size(), equivalent
9323 to erase(begin() + sz, end());. If size() &lt; sz, appends sz -
9324 size() <del>default
9325 constructed</del><ins>value-initialized</ins>
9326 elements to the sequence.
9327 </blockquote>
9328
9329 <p>
9330 Change 23.3.3.1 [forwardlist.cons] para 5:
9331 </p>
9332
9333 <blockquote>
9334 <i>Effects:</i> Constructs a forward_list object with n <del>default
9335 constructed</del><ins>value-initialized</ins>
9336 elements.
9337 </blockquote>
9338
9339 <p>
9340 Change 23.3.3.4 [forwardlist.modifiers] para 21:
9341 </p>
9342
9343 <blockquote>
9344 <i>Effects:</i> [...] For the first signature
9345 the inserted elements are <del>default
9346 constructed</del><ins>value-initialized</ins>,
9347 and for the second signature they are copies of c.
9348 </blockquote>
9349
9350 <p>
9351 Change 23.3.4.1 [list.cons] para 5:
9352 </p>
9353
9354 <blockquote>
9355 <i>Effects:</i> Constructs a list with n <del>default
9356 constructed</del><ins>value-initialized</ins>
9357 elements.
9358 </blockquote>
9359
9360 <p>
9361 Change 23.3.4.2 [list.capacity] para 15:
9362 </p>
9363
9364 <blockquote>
9365 <i>Effects:</i> If sz &lt; size(), equivalent
9366 to list&lt;T&gt;::iterator it = begin(); advance(it, sz); erase(it,
9367 end());. If size() &lt; sz, appends sz - size() <del>default
9368 constructed</del><ins>value-initialized</ins>
9369 elements to the sequence.
9370 </blockquote>
9371
9372 <p>
9373 Change 23.3.6.1 [vector.cons] para 3:
9374 </p>
9375
9376 <blockquote>
9377 <i>Effects:</i> Constructs a vector with n
9378 <del>default constructed</del><ins>value-initialized</ins>
9379 elements.
9380 </blockquote>
9381
9382 <p>
9383 Change 23.3.6.2 [vector.capacity] para 24:
9384 </p>
9385
9386 <blockquote>
9387 <i>Effects:</i> If sz &lt; size(), equivalent
9388 to erase(begin() + sz, end());. If size() &lt; sz, appends sz -
9389 size() <del>default
9390 constructed</del><ins>value-initialized</ins>
9391 elements to the sequence.
9392 </blockquote>
9393
9394
9395
9396
9397
9398
9399 <hr>
9400 <h3><a name="870"></a>870. Do unordered containers not support function pointers for predicate/hasher?</h3>
9401 <p><b>Section:</b> 23.2.5 [unord.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
9402  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-08-17  <b>Last modified:</b> 2009-10-20</p>
9403 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
9404 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
9405 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
9406 <p><b>Discussion:</b></p>
9407 <p>
9408 Good ol' associative containers allow both function pointers and
9409 function objects as feasible
9410 comparators, as described in 23.2.4 [associative.reqmts]/2:
9411 </p>
9412
9413 <blockquote>
9414 Each associative container is parameterized on <tt>Key</tt> and an ordering
9415 relation <tt>Compare</tt> that
9416 induces a strict weak ordering (25.3) on elements of Key. [..]. The
9417 object of type <tt>Compare</tt> is
9418 called the comparison object of a container. This comparison object
9419 may be a pointer to
9420 function or an object of a type with an appropriate function call operator.[..]
9421 </blockquote>
9422
9423 <p>
9424 The corresponding wording for unordered containers is not so clear,
9425 but I read it to disallow
9426 function pointers for the hasher and I miss a clear statement for the
9427 equality predicate, see
9428 23.2.5 [unord.req]/3+4+5:
9429 </p>
9430
9431 <blockquote>
9432 <p>
9433 Each unordered associative container is parameterized by <tt>Key</tt>, by a
9434 function object <tt>Hash</tt> that
9435 acts as a hash function for values of type <tt>Key</tt>, and by a binary
9436 predicate <tt>Pred</tt> that induces an
9437 equivalence relation on values of type <tt>Key</tt>.[..]
9438 </p>
9439 <p>
9440 A hash function is a function object that takes a single argument of
9441 type <tt>Key</tt> and returns a
9442 value of type <tt>std::size_t</tt>.
9443 </p>
9444 <p>
9445 Two values <tt>k1</tt> and <tt>k2</tt> of type <tt>Key</tt> are considered equal if the
9446 container's equality function object
9447 returns <tt>true</tt> when passed those values.[..]
9448 </p>
9449 </blockquote>
9450
9451 <p>
9452 and table 97 says in the column "assertion...post-condition" for the
9453 expression X::hasher:
9454 </p>
9455
9456 <blockquote>
9457 <tt>Hash</tt> shall be a unary function object type such that the expression
9458 <tt>hf(k)</tt> has type <tt>std::size_t</tt>.
9459 </blockquote>
9460
9461 <p>
9462 Note that 20.7 [function.objects]/1 defines as "Function objects are
9463 objects with an <tt>operator()</tt> defined.[..]"
9464 </p>
9465 <p>
9466 Does this restriction exist by design or is it an oversight? If an
9467 oversight, I suggest that to apply
9468 the following
9469 </p>
9470
9471 <p><i>[
9472 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
9473 ]</i></p>
9474
9475
9476 <p><i>[
9477 2009-10 Santa Cruz:
9478 ]</i></p>
9479
9480
9481 <blockquote>
9482 Ask Daniel to provide proposed wording that: makes it explicit that
9483 function pointers are function objects at the beginning of 20.7 [function.objects]; fixes the "requirements" for typedefs in
9484 20.7.5 [refwrap] to instead state that the function objects
9485 defined in that clause have these typedefs, but not that these typedefs
9486 are requirements on function objects; remove the wording that explicitly
9487 calls out that associative container comparators may be function
9488 pointers.
9489 </blockquote>
9490
9491
9492
9493 <p><b>Proposed resolution:</b></p>
9494 <p>
9495 In 23.2.5 [unord.req]/3, just after the second sentence which is written as
9496 </p>
9497
9498 <blockquote>
9499 Additionally, <tt>unordered_map</tt> and <tt>unordered_multimap</tt> associate an
9500 arbitrary mapped type <tt>T</tt> with the <tt>Key</tt>.
9501 </blockquote>
9502
9503 <p>
9504 add one further sentence:
9505 </p>
9506
9507 <blockquote>
9508 Both <tt>Hash</tt> and <tt>Pred</tt> may be pointers to function or objects of a type
9509 with an appropriate function call operator.
9510 </blockquote>
9511
9512 <p>
9513 [Note1: Since the detailed requirements for <tt>Pred</tt> and <tt>Hash</tt> are given in
9514 p.4 and p.5, it an alternative resolution
9515 would be to insert a new paragraph just after p.5, which contains the
9516 above proposed sentence]
9517 </p>
9518 <p>
9519 [Note2: I do not propose a change of above quoted element in table 97,
9520 because the mis-usage of the
9521 notion of "function object" seems already present in the standard at
9522 several places, even if it includes
9523 function pointers, see e.g. 25 [algorithms]/7. The important point is
9524 that in those places a statement is
9525 given that the actually used symbol, like "Predicate" applies for
9526 function pointers as well]
9527 </p>
9528
9529
9530 <p><b>Rationale:</b></p>
9531 <p><i>[
9532 San Francisco:
9533 ]</i></p>
9534
9535
9536 <blockquote>
9537 This is fixed by
9538 <a href="" ref="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2776.pdf">N2776</a>.
9539 </blockquote>
9540
9541
9542
9543
9544
9545
9546 <hr>
9547 <h3><a name="871"></a>871. Iota's requirements on T are too strong</h3>
9548 <p><b>Section:</b> 26.7.5 [numeric.iota] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
9549  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-08-20  <b>Last modified:</b> 2009-10-22</p>
9550 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
9551 <p><b>Discussion:</b></p>
9552 <p>
9553 According to the recent WP
9554 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2691.pdf">N2691</a>,
9555 26.7.5 [numeric.iota]/1, the requires clause
9556 of <tt>std::iota</tt> says:
9557 </p>
9558
9559 <blockquote>
9560 <tt>T</tt> shall meet the requirements of <tt>CopyConstructible</tt> and <tt>Assignable</tt> types, and
9561 shall be convertible to <tt>ForwardIterator</tt>'s value type.[..]
9562 </blockquote>
9563
9564 <p>
9565 Neither <tt>CopyConstructible</tt> nor <tt>Assignable</tt> is needed, instead <tt>MoveConstructible</tt>
9566 seems to be the correct choice. I guess the current wording resulted as an
9567 artifact from comparing it with similar numerical algorithms like <tt>accumulate</tt>.
9568 </p>
9569
9570 <p>
9571 Note: If this function will be conceptualized, the here proposed
9572 <tt>MoveConstructible</tt>
9573 requirement can be removed, because this is an implied requirement of
9574 function arguments, see
9575 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2710.pdf">N2710</a>/[temp.req.impl]/3, last bullet.
9576 </p>
9577
9578 <p><i>[
9579 post San Francisco:
9580 ]</i></p>
9581
9582
9583 <blockquote>
9584 Issue pulled by author prior to review.
9585 </blockquote>
9586
9587 <p><i>[
9588 2009-07-30 Daniel reopened:
9589 ]</i></p>
9590
9591
9592 <blockquote>
9593 with the absence of concepts, this issue (closed) is valid again and I
9594 suggest to reopen it.
9595 I also revised by proposed resolution based on N2723 wording:
9596 </blockquote>
9597
9598 <p><i>[
9599 2009-10 Santa Cruz:
9600 ]</i></p>
9601
9602
9603 <blockquote>
9604 Change 'convertible' to 'assignable', Move To Ready.
9605 </blockquote>
9606
9607
9608
9609 <p><b>Proposed resolution:</b></p>
9610 <p>
9611 Change the first sentence of 26.7.5 [numeric.iota]/1:
9612 </p>
9613
9614 <blockquote>
9615 <i>Requires:</i> <tt>T</tt> shall <del>meet the requirements of <tt>CopyConstructible</tt> and
9616 <tt>Assignable</tt> types, and shall</del> be
9617 assignable to <tt>ForwardIterator</tt>'s value type. [..]
9618 </blockquote>
9619
9620
9621
9622
9623
9624
9625
9626
9627 <hr>
9628 <h3><a name="872"></a>872. <tt>move_iterator::operator[]</tt> has wrong return type</h3>
9629 <p><b>Section:</b> 24.5.3.3.12 [move.iter.op.index] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
9630  <b>Submitter:</b> Doug Gregor <b>Opened:</b> 2008-08-21  <b>Last modified:</b> 2009-10-23</p>
9631 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
9632 <p><b>Discussion:</b></p>
9633 <p>
9634 <tt>move_iterator</tt>'s <tt>operator[]</tt> is declared as:
9635 </p>
9636
9637 <blockquote><pre>reference operator[](difference_type n) const;
9638 </pre></blockquote>
9639
9640 <p>
9641 This has the same problem that <tt>reverse_iterator</tt>'s <tt>operator[]</tt> used to
9642 have: if the underlying iterator's <tt>operator[]</tt> returns a proxy, the
9643 implicit conversion to <tt>value_type&amp;&amp;</tt> could end up referencing a temporary
9644 that has already been destroyed. This is essentially the same issue that
9645 we dealt with for <tt>reverse_iterator</tt> in DR <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#386">386</a>.
9646 </p>
9647
9648 <p><i>[
9649 2009-07-28 Reopened by Alisdair.  No longer solved by concepts.
9650 ]</i></p>
9651
9652
9653 <p><i>[
9654 2009-08-15 Howard adds:
9655 ]</i></p>
9656
9657
9658 <blockquote>
9659 I recommend closing this as  a duplicate of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1051">1051</a> which addresses
9660 this issue for both <tt>move_iterator</tt> and <tt>reverse_iterator</tt>.
9661 </blockquote>
9662
9663 <p><i>[
9664 2009-10 Santa Cruz:
9665 ]</i></p>
9666
9667
9668 <blockquote>
9669 Move to Ready. Note that if <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1051">1051</a> is reopened, it may yield a
9670 better resolution, but <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1051">1051</a> is currently marked NAD.
9671 </blockquote>
9672
9673
9674
9675 <p><b>Proposed resolution:</b></p>
9676 <p>
9677 In 24.5.3.1 [move.iterator] and 24.5.3.3.12 [move.iter.op.index], change the declaration of
9678 <tt>move_iterator</tt>'s <tt>operator[]</tt> to:
9679 </p>
9680
9681 <blockquote><pre><del>reference</del> <ins><i>unspecified</i></ins> operator[](difference_type n) const;
9682 </pre></blockquote>
9683
9684
9685
9686 <p><b>Rationale:</b></p>
9687 <p><i>[
9688 San Francisco:
9689 ]</i></p>
9690
9691
9692 <blockquote>
9693 NAD Editorial, see
9694 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2777.pdf">N2777</a>.
9695 </blockquote>
9696
9697
9698
9699
9700 <hr>
9701 <h3><a name="885"></a>885. pair assignment</h3>
9702 <p><b>Section:</b> 20.3.4 [pairs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
9703  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2008-09-15  <b>Last modified:</b> 2009-07-17</p>
9704 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#pairs">active issues</a> in [pairs].</p>
9705 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#pairs">issues</a> in [pairs].</p>
9706 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
9707 <p><b>Discussion:</b></p>
9708 <blockquote><pre>20.2.3 pairs
9709 Missing assignemnt operator:
9710 template&lt;class U , class V&gt;
9711   requires CopyAssignable&lt;T1, U&gt; &amp;&amp; CopyAssignable&lt;T2, V&gt;
9712     pair&amp; operator=(pair&lt;U , V&gt; const &amp; p );
9713 </pre></blockquote>
9714
9715 <p>
9716 Well, that's interesting. This assignment operator isn't in the
9717 current working paper, either. Perhaps we deemed it acceptable to
9718 build a temporary of type <tt>pair</tt> from <tt>pair&lt;U, V&gt;</tt>, then move-assign
9719 from that temporary?
9720 </p>
9721 <p>
9722 It sounds more like an issue waiting to be opened, unless you want to plug
9723 it now.  As written we risk moving from lvalues.
9724 </p>
9725
9726 <p><i>[
9727 San Francisco:
9728 ]</i></p>
9729
9730
9731 <blockquote>
9732 <p>
9733 Would be NAD if better ctors fixed it.
9734 </p>
9735 <p>
9736 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#811">811</a>.
9737 </p>
9738 </blockquote>
9739
9740 <p><i>[
9741 post San Francisco:
9742 ]</i></p>
9743
9744
9745 <blockquote>
9746 Possibly NAD Editorial, solved by
9747 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2770.pdf">N2770</a>.
9748 </blockquote>
9749
9750 <p><i>[
9751 2009-05-25 Alisdair adds:
9752 ]</i></p>
9753
9754
9755 <blockquote>
9756 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#885">885</a> was something I reported while reviewing the library concepts
9757 documents ahead of San Francisco.  The missing operator was added as part of
9758 the paper adopted at that meeting
9759 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2770.pdf">N2770</a>)
9760 and I can confirm this operator is
9761 present in the current working paper.  I recommend NAD.
9762 </blockquote>
9763
9764 <p><i>[
9765 2009-07 Frankfurt
9766 ]</i></p>
9767
9768
9769 <blockquote>
9770 We agree with the intent, but we need to wait for the dust to settle on concepts.
9771 </blockquote>
9772
9773
9774
9775 <p><b>Proposed resolution:</b></p>
9776 <p>
9777 </p>
9778
9779
9780
9781
9782
9783 <hr>
9784 <h3><a name="887"></a>887. issue with condition::wait_...</h3>
9785 <p><b>Section:</b> 30.5.1 [thread.condition.condvar] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
9786  <b>Submitter:</b> Lawrence Crowl <b>Opened:</b> 2008-09-15  <b>Last modified:</b> 2009-10-26</p>
9787 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvar">active issues</a> in [thread.condition.condvar].</p>
9788 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvar">issues</a> in [thread.condition.condvar].</p>
9789 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
9790 <p><b>Discussion:</b></p>
9791 <p>
9792 The Posix/C++ working group has identified an inconsistency between
9793 Posix and the C++ working draft in that Posix requires the clock to be
9794 identified at creation, whereas C++ permits identifying the clock at the
9795 call to wait.  The latter cannot be implemented with the former.
9796 </p>
9797
9798 <p><i>[
9799 San Francisco:
9800 ]</i></p>
9801
9802
9803 <blockquote>
9804 <p>
9805 Howard recommends NAD with the following explanation:
9806 </p>
9807
9808 <p>
9809 The intent of the current wording is for the <tt>condtion_variable::wait_until</tt>
9810 be able to handle user-defined clocks as well as clocks the system knows about.
9811 This can be done by providing overloads for the known clocks, and another
9812 overload for unknown clocks which synchs to a known clock before waiting.
9813 For example:
9814 </p>
9815
9816 <blockquote><pre>template &lt;class Duration&gt;
9817 bool
9818 condition_variable::wait_until(unique_lock&lt;mutex&gt;&amp; lock,
9819                                const chrono::time_point&lt;chrono::system_clock, Duration&gt;&amp; abs_time)
9820 {
9821     using namespace chrono;
9822     nanoseconds d = __round_up&lt;nanoseconds&gt;(abs_time.time_since_epoch());
9823     __do_timed_wait(lock.mutex()-&gt;native_handle(), time_point&lt;system_clock, nanoseconds&gt;(d));
9824     return system_clock::now() &lt; abs_time;
9825 }
9826
9827 template &lt;class Clock, class Duration&gt;
9828 bool
9829 condition_variable::wait_until(unique_lock&lt;mutex&gt;&amp; lock,
9830                                const chrono::time_point&lt;Clock, Duration&gt;&amp; abs_time)
9831 {
9832     using namespace chrono;
9833     system_clock::time_point    s_entry = system_clock::now();
9834     typename Clock::time_point  c_entry = Clock::now();
9835     nanoseconds dn = __round_up&lt;nanoseconds&gt;(abs_time.time_since_epoch() -
9836                                               c_entry.time_since_epoch());
9837     __do_timed_wait(lock.mutex()-&gt;native_handle(), s_entry + dn);
9838     return Clock::now() &lt; abs_time;
9839 }
9840 </pre></blockquote>
9841
9842 <p>
9843 In the above example, <tt>system_clock</tt> is the only clock which the underlying
9844 condition variable knows how to deal with.  One overload just passes that clock
9845 through.  The second overload (approximately) converts the unknown clock into
9846 a <tt>system_clock  time_point</tt> prior to passing it down to the native
9847 condition variable.
9848 </p>
9849
9850 <p>
9851 On Posix systems vendors are free to add implementation defined constructors which
9852 take a clock.  That clock can be stored in the condition_variable, and converted
9853 to (or not as necessary) as shown above.
9854 </p>
9855
9856 <p>
9857 If an implementation defined constructor takes a clock (for example), then part
9858 of the semantics for that implementation defined ctor might include that a
9859 <tt>wait_until</tt> using a clock other than the one constructed with results
9860 in an error (exceptional condition) instead of a conversion to the stored clock.
9861 Such a design is up to the vendor as once an implementation defined ctor is used,
9862 the vendor is free to specifiy the behavior of waits and/or notifies however
9863 he pleases (when the cv is constructed in an implementation defined manner).
9864 </p>
9865 </blockquote>
9866
9867 <p><i>[
9868 Post Summit:
9869 ]</i></p>
9870
9871
9872 <blockquote>
9873 <p>
9874 "POSIX people will review the proposed NAD resolution at their upcoming NY
9875 meeting.
9876 </p>
9877
9878 <p>
9879 See the minutes at: <a href="http://wiki.dinkumware.com/twiki/bin/view/Posix/POSIX-CppBindingWorkingGroupNewYork2009">http://wiki.dinkumware.com/twiki/bin/view/Posix/POSIX-CppBindingWorkingGroupNewYork2009</a>.
9880 </p>
9881 </blockquote>
9882
9883 <p><i>[
9884 2009-07 Frankfurt
9885 ]</i></p>
9886
9887
9888 <blockquote>
9889 Move to NAD.
9890 </blockquote>
9891
9892 <p><i>[
9893 2009-07-18 Detlef reopens the issue:
9894 ]</i></p>
9895
9896
9897 <blockquote>
9898 <p>
9899 On Friday afternoon in Frankfurt is was decided that 887 is NAD.
9900 This decision was mainly based on a sample implementation presented
9901 by Howard that implemented one clock on top of another.
9902 Unfortunately this implementation doesn't work for the probably most
9903 important case where a system has a monotonic clock and a real-time
9904 clock (or "wall time" clock):
9905 </p>
9906 <p>
9907 If the underlying "system_clock" is a monotonic clock, and
9908 the program waits on the real-time clock, and the real-time clock
9909 is set forward, the wait will unblock too late.
9910 </p>
9911
9912 <p>
9913 If the underlying "system_clock" is a real-time clock, and the
9914 program waits on the monotonic clock, and the real-time clock
9915 is set back, the wait again will unblock too late.
9916 </p>
9917
9918 <p>
9919 Sorry that I didn't remember this on Friday, but it was Friday
9920 afternoon after a busy week...
9921 </p>
9922
9923 <p>
9924 So as the decision was made on a wrong asumption, I propose to re-open
9925 the issue.
9926 </p>
9927 </blockquote>
9928
9929 <p><i>[
9930 2009-07-26 Howard adds:
9931 ]</i></p>
9932
9933
9934 <blockquote>
9935 <p>
9936 Detlef correctly argues that <tt>condition_variable::wait_until</tt> could
9937 return "too late" in the context of clocks being adjusted during the wait.  I agree
9938 with his logic.  But I disagree that this makes this interface unimplementable
9939 on POSIX.
9940 </p>
9941
9942 <p>
9943 The POSIX spec also does not guarantee that <tt>pthread_cond_timedwait</tt> does
9944 not return "too late" when clocks are readjusted during the wait.  Indeed, the
9945 POSIX specification lacks any requirements at all concerning how soon
9946 <tt>pthread_cond_timedwait</tt> returns after a time out.  This is evidently a
9947 QOI issue by the POSIX standard.  Here is a quote of the most relevant normative
9948 text concerning <tt>pthread_cond_timedwait</tt> found
9949 <a href="http://www.unix.org/single_unix_specification/">here</a>.
9950 </p>
9951
9952 <blockquote>
9953 The <tt>pthread_cond_timedwait()</tt> function shall be equivalent to
9954 <tt>pthread_cond_wait()</tt>, except that an error is returned if the absolute
9955 time specified by <tt>abstime</tt> passes (that is, system time equals or exceeds
9956 <tt>abstime</tt>) before the condition <tt>cond</tt> is signaled or broadcasted, or if the
9957 absolute time specified by <tt>abstime</tt> has already been passed at the time
9958 of the call.
9959 </blockquote>
9960
9961 <p>
9962 I.e. the POSIX specification speaks of the error code returned in case of a time
9963 out, but not on the timeliness of that return.
9964 </p>
9965
9966 <p>
9967 Might this simply be an oversight, or minor defect in the POSIX specification?
9968 </p>
9969
9970 <p>
9971 I do not believe so.  This same section goes on to say in <em>non-normative</em>
9972 text:
9973 </p>
9974
9975 <blockquote>
9976 For cases when the system clock is advanced discontinuously by an
9977 operator, it is expected that implementations process any timed wait
9978 expiring at an intervening time as if that time had actually occurred.
9979 </blockquote>
9980
9981 <p>
9982 Here is non-normative wording encouraging the implementation to ignore an advancing
9983 underlying clock and subsequently causing an early (spurious) return.  There is
9984 no wording at all which addresses Detlef's example of a "late return".  With
9985 <tt>pthread_cond_timedwait</tt> this would be caused by setting the system clock
9986 backwards.  It seems reasonable to assume, based on the wording that is already
9987 in the POSIX spec, that again, the discontinuously changed clock would be ignored
9988 by <tt>pthread_cond_timedwait</tt>.
9989 </p>
9990
9991 <p>
9992 A noteworthy difference between <tt>pthread_cond_timedwait</tt> and
9993 <tt>condition_variable::wait_until</tt> is that the POSIX spec appears to
9994 say that <tt>ETIMEDOUT</tt> should be returned if <tt>pthread_cond_timedwait</tt>
9995 returns because of timeout signal, whether or not the system clock was discontinuously
9996 advanced during the wait.  In contrast <tt>condition_variable::wait_until</tt>
9997 always returns:
9998 </p>
9999
10000 <blockquote><pre><tt>Clock::now() &lt; abs_time</tt>
10001 </pre></blockquote>
10002
10003 <p>
10004 That is, the C++ spec requires that the clock be rechecked (detecting discontinuous
10005 adjustments during the wait) at the time of return.  <tt>condition_variable::wait_until</tt>
10006 may indeed return early or late.  But regardless it will return a value
10007 reflecting timeout status at the time of return (even if clocks have been adjusted).
10008 Of course the clock may be adjusted after the return value is computed but before the client has
10009 a chance to read the result of the return.  Thus there are no iron-clad guarantees
10010 here.
10011 </p>
10012
10013 <p>
10014 <tt>condition_variable::wait_until</tt> (and <tt>pthread_cond_timedwait</tt>)
10015 is little more than a convenience function for making sure
10016 <tt>condition_variable::wait</tt> doesn't hang for an unreasonable amount of
10017 time (where the client gets to define "unreasonable").  I do not think it
10018 is in anyone's interest to try to make it into anything more than that.
10019 </p>
10020
10021 <p>
10022 I maintain that this is a useful and flexible specification in the spirit of
10023 C++, and is implementable on POSIX.  The implementation technique described above
10024 is a reasonable approach.  There may also be higher quality approaches.  This
10025 specification, like the POSIX specification, gives a wide latitude for QOI.
10026 </p>
10027
10028 <p>
10029 I continue to recommend NAD, but would not object to a clarifying note regarding
10030 the behavior of <tt>condition_variable::wait_until</tt>.  At the moment, I do
10031 not have good wording for such a note, but welcome suggestions.
10032 </p>
10033
10034 </blockquote>
10035
10036 <p><i>[
10037 2009-09-30: See <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2969.html">N2969</a>.
10038 ]</i></p>
10039
10040
10041 <p><i>[
10042 2009-10 Santa Cruz:
10043 ]</i></p>
10044
10045
10046 <blockquote>
10047 The LWG is in favor of Detlef to supply revision which adopts Option 2 from
10048 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2969.html">N2969</a>
10049 but is modified by saying that <tt>system_clock</tt> must be available for <tt>wait_until</tt>.
10050 </blockquote>
10051
10052
10053
10054 <p><b>Proposed resolution:</b></p>
10055 <p>
10056 </p>
10057
10058
10059
10060
10061
10062 <hr>
10063 <h3><a name="889"></a>889. thread::id comparisons</h3>
10064 <p><b>Section:</b> 30.3.1.1 [thread.thread.id] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
10065  <b>Submitter:</b> Lawrence Crowl <b>Opened:</b> 2008-09-15  <b>Last modified:</b> 2009-10-22</p>
10066 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.id">issues</a> in [thread.thread.id].</p>
10067 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
10068 <p><b>Discussion:</b></p>
10069
10070 <p><b>Addresses UK 324</b></p>
10071
10072 <p>
10073 The <tt>thread::id</tt> type supports the full set of comparison operators.  This
10074 is substantially more than is required for the associative containers that
10075 justified them.  Please place an issue against the threads library.
10076 </p>
10077
10078 <p><i>[
10079 San Francisco:
10080 ]</i></p>
10081
10082
10083 <blockquote>
10084 <p>
10085 Would depend on proposed extension to POSIX, or non-standard extension.
10086 What about hash? POSIX discussing op. POSIX not known to be considering
10087 support needed for hash, op.
10088 </p>
10089 <p>
10090 Group expresses support for putting ids in both unordered and ordered containers.
10091 </p>
10092 </blockquote>
10093
10094 <p><i>[
10095 post San Francisco:
10096 ]</i></p>
10097
10098
10099 <blockquote>
10100 <p>
10101 Howard:  It turns out the current working paper
10102 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>
10103 <i>already has</i> <tt>hash&lt;thread::id&gt;</tt>
10104 (20.7 [function.objects], 20.7.16 [unord.hash]).  We simply
10105 overlooked it in the meeting.  It is a good thing we voted in favor of it
10106 (again). :-)
10107 </p>
10108 <p>
10109 Recommend NAD.
10110 </p>
10111
10112 </blockquote>
10113
10114 <p><i>[
10115 Post Summit:
10116 ]</i></p>
10117
10118
10119 <blockquote>
10120 Recommend to close as NAD. For POSIX, see if we need to add a function to
10121 convert <tt>pthread_t</tt> to integer.
10122 </blockquote>
10123
10124 <p><i>[
10125 Post Summit, Alisdair adds:
10126 ]</i></p>
10127
10128
10129 <blockquote>
10130 <p>
10131 The recommendation for LWG-889/UK-324 is NAD, already specified.
10132 </p>
10133 <p>
10134 It is not clear to me that the specification is complete.
10135 </p>
10136 <p>
10137 In particular, the synopsis of <tt>&lt;functional&gt;</tt> in 20.7 [function.objects] does not mention <tt>hash&lt; thread::id
10138 &gt;</tt> nor <tt>hash&lt; error_code &gt;</tt>, although their
10139 existence is implied by 20.7.16 [unord.hash], p1.
10140 </p>
10141 <p>
10142 I am fairly uncomfortable putting the declaration for the
10143 <tt>thread_id</tt> specialization into <tt>&lt;functional&gt;</tt> as
10144 <tt>id</tt> is a nested class inside <tt>std::thread</tt>, so it implies
10145 that <tt>&lt;functional&gt;</tt> would require the definition of the
10146 <tt>thread</tt> class template in order to forward declared
10147 <tt>thread::id</tt> and form this specialization.
10148 </p>
10149 <p>
10150 It seems better to me that the dependency goes the other way around
10151 (<tt>&lt;thread&gt;</tt> will more typically make use of
10152 <tt>&lt;functional&gt;</tt> than vice-versa) and the
10153 <tt>hash&lt;thread::id&gt;</tt> specialization be declared in the
10154 <tt>&lt;thread&gt;</tt> header.
10155 </p>
10156 <p>
10157 I think <tt>hash&lt;error_code&gt;</tt> could go into either
10158 <tt>&lt;system_error&gt;</tt> or <tt>&lt;functional&gt;</tt> and have no
10159 immediate preference either way.  However, it should clearly appear in
10160 the synopsis of one of these two.
10161 </p>
10162 <p>
10163 Recommend moving 889 back to open, and tying in a reference to UK-324.
10164 </p>
10165 </blockquote>
10166
10167 <p><i>[
10168 Batavia (2009-05):
10169 ]</i></p>
10170
10171 <blockquote>
10172 Howard observes that <tt>thread::id</tt> need not be a nested class;
10173 it could be a <tt>typedef</tt> for a more visible type.
10174 </blockquote>
10175
10176 <p><i>[
10177 2009-05-24 Alisdair adds:
10178 ]</i></p>
10179
10180 <blockquote>
10181 I do not believe this is correct.  <tt>thread::id</tt> is explicitly documents as a
10182 nested class, rather than as an unspecified typedef analogous to an
10183 iterator.  If the intent is that this is not implemented as a nested class
10184 (under the as-if freedoms) then this is a novel form of standardese.
10185 </blockquote>
10186
10187 <p><i>[
10188 2009-07 Frankfurt
10189 ]</i></p>
10190
10191
10192 <blockquote>
10193 Decided we want to move hash specialization for thread_id to the thread
10194 header. Alisdair to provide wording.
10195 </blockquote>
10196
10197 <p><i>[
10198 2009-07-28 Alisdair provided wording, moved to Review.
10199 ]</i></p>
10200
10201
10202 <p><i>[
10203 2009-10 Santa Cruz:
10204 ]</i></p>
10205
10206
10207 <blockquote>
10208 Add a strike for <tt>hash&lt;thread::id&gt;</tt>. Move to Ready
10209 </blockquote>
10210
10211
10212
10213 <p><b>Proposed resolution:</b></p>
10214
10215 <p>
10216 Remove the following prototype from the synopsis in
10217 20.7 [function.objects]:
10218 </p>
10219
10220 <blockquote><pre><del>
10221 template &lt;&gt; struct hash&lt;std::thread::id&gt;;
10222 </del></pre></blockquote>
10223
10224 <p>
10225 Add to 30.3 [thread.threads], p1 Header <tt>&lt;thread&gt;</tt> synopsis:
10226 </p>
10227
10228 <blockquote><pre><ins>template &lt;class T&gt; struct hash;
10229 template &lt;&gt; struct hash&lt;thread::id&gt;;</ins>
10230 </pre></blockquote>
10231
10232 <p>
10233 Add template specialization below class definition in 30.3.1.1 [thread.thread.id]
10234 </p>
10235
10236 <blockquote><pre><ins>template &lt;&gt;
10237 struct hash&lt;thread::id&gt; : public unary_function&lt;thread::id, size_t&gt; {
10238    size_t operator()(thread::id val) const;
10239 };</ins>
10240 </pre></blockquote>
10241
10242 <p>
10243 Extend note in p2 30.3.1.1 [thread.thread.id] with second sentence:
10244 </p>
10245
10246 <blockquote>
10247 [<i>Note:</i> Relational operators allow <tt>thread::id</tt> objects to be used
10248 as keys in associative containers.
10249 <ins><tt>hash</tt> template specialization allow <tt>thread::id</tt> objects to be used as keys
10250 in unordered containers.</ins>
10251 &#8212; <i>end note</i>]
10252 </blockquote>
10253
10254 <p>
10255 Add new paragraph to end of 30.3.1.1 [thread.thread.id]
10256 </p>
10257
10258 <blockquote><pre><ins>template &lt;&gt; struct hash&lt;thread::id&gt;;</ins>
10259 </pre>
10260 <blockquote><ins>
10261 An explicit specializations of the class template hash (20.7.16 [unord.hash])
10262 shall be provided for the values of type <tt>thread::id</tt>
10263 suitable for use as keys in unordered associative containers (23.5 [unord]).
10264 </ins></blockquote>
10265 </blockquote>
10266
10267
10268
10269
10270
10271 <hr>
10272 <h3><a name="891"></a>891. std::thread, std::call_once issue</h3>
10273 <p><b>Section:</b> 30.3.1.2 [thread.thread.constr], 30.4.5.2 [thread.once.callonce] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
10274  <b>Submitter:</b> Peter Dimov <b>Opened:</b> 2008-09-15  <b>Last modified:</b> 2009-10-24</p>
10275 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.constr">active issues</a> in [thread.thread.constr].</p>
10276 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.constr">issues</a> in [thread.thread.constr].</p>
10277 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
10278 <p><b>Discussion:</b></p>
10279 <p>
10280 I notice that the vararg overloads of <tt>std::thread</tt> and <tt>std::call_once</tt>
10281 (N2723 30.3.1.2 [thread.thread.constr] and 30.4.5.2 [thread.once.callonce]) are no longer specified in terms of
10282 <tt>std::bind</tt>; instead, some of the <tt>std::bind</tt> wording has been inlined into
10283 the specification.
10284 </p>
10285 <p>
10286 There are two problems with this.
10287 </p>
10288 <p>
10289 First, the specification (and implementation) in terms of <tt>std::bind</tt> allows, for example:
10290 </p>
10291
10292 <blockquote><pre>std::thread th( f, 1, std::bind( g ) );
10293 </pre></blockquote>
10294
10295 <p>
10296 which executes <tt>f( 1, g() )</tt> in a thread. This can be useful. The
10297 "inlined" formulation changes it to execute <tt>f( 1, bind(g) )</tt> in a thread.
10298 </p>
10299 <p>
10300 Second, assuming that we don't want the above, the specification has copied the wording
10301 </p>
10302
10303 <blockquote>
10304 <tt>INVOKE(func, w1, w2, ..., wN)</tt> (20.6.2) shall be a valid
10305 expression for some values <tt>w1, w2, ..., wN</tt>
10306 </blockquote>
10307
10308 <p>
10309 but this is not needed since we know that our argument list is args; it should simply be
10310 </p>
10311
10312 <blockquote>
10313 <tt>INVOKE(func, args...)</tt> (20.6.2) shall be a valid expression
10314 </blockquote>
10315
10316 <p><i>[
10317 Summit:
10318 ]</i></p>
10319
10320
10321 <blockquote>
10322 Move to open.
10323 </blockquote>
10324
10325 <p><i>[
10326 Post Summit Anthony provided proposed wording.
10327 ]</i></p>
10328
10329
10330 <p><i>[
10331 2009-07 Frankfurt
10332 ]</i></p>
10333
10334
10335 <blockquote>
10336 Leave Open. Await decision for thread variadic constructor.
10337 </blockquote>
10338
10339 <p><i>[
10340 2009-10 Santa Cruz:
10341 ]</i></p>
10342
10343
10344 <blockquote>
10345 See proposed wording for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a> for this, for the formulation
10346 on how to solve this.  <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a> modifies the thread constructor to
10347 have "pass by value" behavior with pass by reference efficiency through the use
10348 of the <tt>decay</tt> trait.  This same formula would be useful for <tt>call_once</tt>.
10349 </blockquote>
10350
10351
10352
10353 <p><b>Proposed resolution:</b></p>
10354 <p>
10355 Change paragraph 4 of 30.3.1.2 [thread.thread.constr] to:
10356 </p>
10357
10358 <blockquote>
10359 <pre>template &lt;class F&gt; explicit thread(F f);
10360 template &lt;class F, class ...Args&gt; thread(F&amp;&amp; f, Args&amp;&amp;... args);
10361 </pre>
10362 <blockquote>
10363 -4- <i>Requires:</i> <tt>F</tt> and each <tt>Ti</tt> in <tt>Args</tt>
10364 shall be <tt>CopyConstructible</tt> if an lvalue and otherwise
10365 <tt>MoveConstructible</tt>. <tt><i>INVOKE</i>(f, <del>w1, w2, ..., wN</del> <ins>args...</ins>)</tt>
10366 (20.6.2) shall be a valid expression<del> for some values <i>w1, w2, ...,
10367 wN</i>, where <tt>N == sizeof...(Args)</tt></del>.
10368 </blockquote>
10369 </blockquote>
10370
10371 <p>
10372 Change paragraph 1 of 30.4.5.2 [thread.once.callonce] to:
10373 </p>
10374
10375 <blockquote><pre>template&lt;class Callable, class ...Args&gt;
10376   void call_once(once_flag&amp; flag, Callable func, Args&amp;&amp;... args);
10377 </pre>
10378 <blockquote>
10379 -1- <i>Requires:</i> The template parameters <tt>Callable&gt;</tt> and each
10380 <tt>Ti</tt> in <tt>Args</tt> shall be <tt>CopyConstructible</tt> if an
10381 lvalue and otherwise <tt>MoveConstructible</tt>. <tt><i>INVOKE</i>(func,
10382 <del>w1, w2, ..., wN</del> <ins>args...</ins>)</tt> (20.6.2) shall be a
10383 valid expression<del> for some values <i>w1, w2, ..., wN</i>, where
10384 <tt>N == sizeof...(Args)</tt></del>.
10385 </blockquote>
10386 </blockquote>
10387
10388
10389
10390
10391
10392 <hr>
10393 <h3><a name="893"></a>893. std::mutex issue</h3>
10394 <p><b>Section:</b> 30.4.1.1 [thread.mutex.class] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
10395  <b>Submitter:</b> Peter Dimov <b>Opened:</b> 2008-09-15  <b>Last modified:</b> 2009-10-22</p>
10396 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.mutex.class">issues</a> in [thread.mutex.class].</p>
10397 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
10398 <p><b>Duplicate of:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#905">905</a></p>
10399 <p><b>Discussion:</b></p>
10400 <p>
10401 30.4.1.1 [thread.mutex.class]/27 (in
10402 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>)
10403 says that the behavior is undefined if:
10404 </p>
10405 <ul>
10406 <li>a thread that owns a <tt>mutex</tt> object calls <tt>lock()</tt> or
10407 <tt>try_lock()</tt> on that object</li>
10408 </ul>
10409 <p>
10410 I don't believe that this is right. Calling <tt>lock()</tt> or <tt>try_lock()</tt> on a
10411 locked <tt>mutex</tt> is well defined in the general case. <tt>try_lock()</tt> is required
10412 to fail and return <tt>false</tt>. <tt>lock()</tt> is required to either throw an
10413 exception (and is allowed to do so if it detects deadlock) or to block
10414 until the <tt>mutex</tt> is free. These general requirements apply regardless of
10415 the current owner of the <tt>mutex</tt>; they should apply even if it's owned by
10416 the current thread.
10417 </p>
10418 <p>
10419 Making double <tt>lock()</tt> undefined behavior probably can be justified (even
10420 though I'd still disagree with the justification), but <tt>try_lock()</tt> on a
10421 locked <tt>mutex</tt> must fail.
10422 </p>
10423
10424 <p><i>[
10425 Summit:
10426 ]</i></p>
10427
10428 <blockquote>
10429 <p>
10430 Move to open. Proposed resolution:
10431 </p>
10432 <ul>
10433 <li>
10434 In 30.4.1 [thread.mutex.requirements] paragraph 12, change the error
10435 condition for <tt>resource_deadlock_would_occur</tt> to: "if the implementation
10436 detects that a deadlock would occur"
10437 </li>
10438 <li>
10439 Strike 30.4.1.1 [thread.mutex.class] paragraph 3 bullet 2 "a thread that owns a mutex object
10440 calls <tt>lock()</tt> or <tt>try_lock()</tt> on that object, or"
10441 </li>
10442 </ul>
10443 </blockquote>
10444
10445 <p><i>[
10446 2009-07 Frankfurt
10447 ]</i></p>
10448
10449
10450 <blockquote>
10451 Move to Review. Alisdair to provide note.
10452 </blockquote>
10453
10454 <p><i>[
10455 2009-07-31 Alisdair provided note.
10456 ]</i></p>
10457
10458
10459 <p><i>[
10460 2009-10 Santa Cruz:
10461 ]</i></p>
10462
10463
10464 <blockquote>
10465 Moved to Ready.
10466 </blockquote>
10467
10468
10469
10470 <p><b>Proposed resolution:</b></p>
10471 <p>
10472 In 30.4.1 [thread.mutex.requirements] paragraph 12 change:
10473 </p>
10474
10475 <blockquote>
10476 <ul>
10477 <li>...</li>
10478 <li>
10479 <tt>resource_deadlock_would_occur</tt> -- if the <del>current thread already owns the mutex and is able
10480 to detect it</del> <ins>implementation detects that a deadlock would occur</ins>.
10481 </li>
10482 <li>...</li>
10483 </ul>
10484 </blockquote>
10485
10486 <p>
10487 Strike 30.4.1.1 [thread.mutex.class] paragraph 3 bullet 2:
10488 </p>
10489 <blockquote>
10490 <p>
10491 -3- The behavior of a program is undefined if:
10492 </p>
10493 <ul>
10494 <li>...</li>
10495 <li>
10496 <del>a thread that owns a <tt>mutex</tt> object calls <tt>lock()</tt> or <tt>try_lock()</tt> on that object, or</del>
10497 </li>
10498 <li>...</li>
10499 </ul>
10500 </blockquote>
10501
10502 <p>
10503 Add the following note after p3 30.4.1.1 [thread.mutex.class]
10504 </p>
10505
10506 <blockquote>
10507 [<i>Note:</i> a program may deadlock if the thread that owns a <tt>mutex</tt>
10508 object calls <tt>lock()</tt> or <tt>try_lock()</tt> on that object. If the program can
10509 detect the deadlock, a <tt>resource_deadlock_would_occur</tt> error condition may
10510 be observed. &#8212; <i>end note</i>]
10511 </blockquote>
10512
10513
10514
10515
10516
10517
10518 <hr>
10519 <h3><a name="896"></a>896. Library thread safety issue</h3>
10520 <p><b>Section:</b> 20.8.15.2 [util.smartptr.shared] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
10521  <b>Submitter:</b> Hans Boehm <b>Opened:</b> 2008-09-16  <b>Last modified:</b> 2009-10-25</p>
10522 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#util.smartptr.shared">issues</a> in [util.smartptr.shared].</p>
10523 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
10524 <p><b>Discussion:</b></p>
10525 <p>
10526 It is unclear whether <tt>shared_ptr</tt> is thread-safe in the sense that
10527 multiple threads may simultaneously copy a <tt>shared_ptr</tt>.  However this
10528 is a critical piece of information for the client, and it has significant
10529 impact on usability for many applications.  (Detlef Vollman thinks it
10530 is currently clear that it is not thread-safe.  Hans Boehm thinks
10531 it currently requires thread safety, since the <tt>use_count</tt> is not an
10532 explicit field, and constructors and assignment take a const reference
10533 to an existing <tt>shared_ptr</tt>.)
10534 </p>
10535
10536 <p>
10537 Pro thread-safety:
10538 </p>
10539 <p>
10540 Many multi-threaded usages are impossible.  A thread-safe version can
10541 be used to destroy an object when the last thread drops it, something
10542 that is often required, and for which we have no other easy mechanism.
10543 </p>
10544 <p>
10545 Against thread-safety:
10546 </p>
10547 <p>
10548 The thread-safe version is well-known to be far more expensive, even
10549 if used by a single thread.  Many applications, including all single-threaded
10550 ones, do not care.
10551 </p>
10552
10553 <p><i>[
10554 San Francisco:
10555 ]</i></p>
10556
10557
10558 <blockquote>
10559 <p>
10560 Beman: this is a complicated issue, and would like to move this to Open
10561 and await comment from Peter Dimov; we need very careful and complete
10562 rationale for any decision we make; let's go slow
10563 </p>
10564 <p>
10565 Detlef: I think that <tt>shared_ptr</tt> should not be thread-safe.
10566 </p>
10567 <p>
10568 Hans: When you create a thread with a lambda, it in some cases makes it
10569 very difficult for the lambda to reference anything in the heap. It's
10570 currently ambiguous as to whether you can use a <tt>shared_ptr</tt> to get at an
10571 object.
10572 </p>
10573 <p>
10574 Leave in Open. Detlef will submit an alternative proposed resolution
10575 that makes <tt>shared_ptr</tt> explicitly unsafe.
10576 </p>
10577 <p>
10578 A third option is to support both threadsafe and non-safe share_ptrs,
10579 and to let the programmer decide which behavior they want.
10580 </p>
10581
10582 <p>
10583 Beman:  Peter, do you support the PR?
10584 </p>
10585
10586 <p>
10587 Peter:
10588 </p>
10589 <blockquote>
10590 <p>
10591 Yes, I support the proposed resolution, and I certainly oppose any
10592 attempts to <tt>make shared_ptr</tt> thread-unsafe.
10593 </p>
10594 <p>
10595 I'd mildly prefer if
10596 </p>
10597 <blockquote>
10598 [<i>Note:</i> This is true in spite of that fact that such functions often
10599 modify <tt>use_count()</tt> <i>--end note</i>]
10600 </blockquote>
10601 <p>
10602 is changed to
10603 </p>
10604 <blockquote>
10605 [<i>Note:</i> This is true in spite of that fact that such functions often
10606 cause a change in <tt>use_count()</tt> <i>--end note</i>]
10607 </blockquote>
10608 <p>
10609 (or something along these lines) to emphasise that <tt>use_count()</tt> is not,
10610 conceptually, a variable, but a return value.
10611 </p>
10612 </blockquote>
10613
10614 </blockquote>
10615
10616 <p><i>[
10617 2009-07 Frankfurt
10618 ]</i></p>
10619
10620
10621 <blockquote>
10622 <p>
10623 Vote: Do we want one thread-safe shared pointer or two? If two, one
10624 would allow concurrent construction and destruction of shared pointers,
10625 and one would not be thread-safe. If one, then it would be thread-safe.
10626 </p>
10627 <p>
10628 No concensus on that vote.
10629 </p>
10630 <p>
10631 Hans to improve wording in consultation with Pete. Leave Open.
10632 </p>
10633 </blockquote>
10634
10635 <p><i>[
10636 2009-10 Santa Cruz:
10637 ]</i></p>
10638
10639
10640 <blockquote>
10641 Move to Ready. Ask Editor to clear up wording a little when integrating to
10642 make it clear that the portion after the first comma only applies for
10643 the presence of data races.
10644 </blockquote>
10645
10646 <p><i>[
10647 2009-10-24 Hans adds:
10648 ]</i></p>
10649
10650
10651 <blockquote>
10652 <p>
10653 I think we need to pull 896 back from ready, unfortunately.  My wording
10654 doesn't say the right thing.
10655 </p>
10656
10657 <p>
10658 I suspect we really want to say something along the lines of:
10659 </p>
10660
10661 <blockquote>
10662 For purposes of determining the presence of a data race, member
10663 functions access and modify only the <tt>shared_ptr</tt> and
10664 <tt>weak_ptr</tt> objects themselves and not objects they refer to.
10665 Changes in <tt>use_count()</tt> do not reflect modifications that can
10666 introduce data races.
10667 </blockquote>
10668
10669 <p>
10670 But I think this needs further discussion by experts to make sure this
10671 is right.
10672 </p>
10673
10674 <p>
10675 Detlef and I agree continue to disagree on the resolution, but I think
10676 we agree that it would be good to try to expedite this so that it can be
10677 in CD2, since it's likely to generate NB comments no matter what we do.
10678 And lack of clarity of intent is probably the worst option.  I think it
10679 would be good to look at this between meetings.
10680 </p>
10681 </blockquote>
10682
10683
10684
10685 <p><b>Proposed resolution:</b></p>
10686 <p>
10687 Make it explicitly thread-safe, in this weak sense, as I believe was intended:
10688 </p>
10689 <p>
10690 Insert in 20.8.15.2 [util.smartptr.shared], before p5:
10691 </p>
10692 <blockquote>
10693 <p>
10694 For purposes of determining the presence of a data race,
10695 member functions do not modify <tt>const shared_ptr</tt> and
10696 const <tt>weak_ptr</tt> arguments, nor any objects they
10697 refer to.  [<i>Note:</i> This is true in spite of that fact that such functions often
10698 cause a change in <tt>use_count()</tt> <i>--end note</i>]
10699 </p>
10700 </blockquote>
10701 <p>
10702 On looking at the text, I'm not sure we need a similar disclaimer
10703 anywhere else, since nothing else has the problem with the modified
10704 <tt>use_count()</tt>.  I think Howard arrived at a similar conclusion.
10705 </p>
10706
10707
10708
10709
10710
10711 <hr>
10712 <h3><a name="900"></a>900. stream move-assignment</h3>
10713 <p><b>Section:</b> 27.9.1.8 [ifstream.assign] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
10714  <b>Submitter:</b> Niels Dekker <b>Opened:</b> 2008-09-20  <b>Last modified:</b> 2009-10-20</p>
10715 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
10716 <p><b>Discussion:</b></p>
10717 <p>
10718 It
10719 appears that we have an issue similar to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a> regarding the move-assignment of
10720 stream types. For example, when assigning to an <tt>std::ifstream</tt>,
10721 <tt>ifstream1</tt>, it seems preferable to close the file originally held by
10722 <tt>ifstream1</tt>:
10723 </p>
10724
10725 <blockquote><pre>ifstream1 = std::move(ifstream2);
10726 </pre></blockquote>
10727
10728 <p>
10729 The current Draft
10730 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>)
10731 specifies that the move-assignment of
10732 stream types like <tt>ifstream</tt> has the same effect as a swap:
10733 </p>
10734
10735 <blockquote>
10736 <p>
10737 Assign and swap 27.9.1.8 [ifstream.assign]
10738 </p>
10739 <pre>basic_ifstream&amp; operator=(basic_ifstream&amp;&amp; rhs);
10740 </pre>
10741 <blockquote>
10742 <i>Effects:</i> <tt>swap(rhs)</tt>.
10743 </blockquote>
10744 </blockquote>
10745
10746 <p><i>[
10747 Batavia (2009-05):
10748 ]</i></p>
10749
10750 <blockquote>
10751 <p>
10752 Howard agrees with the analysis and the direction proposed.
10753 </p>
10754 <p>
10755 Move to Open pending specific wording to be supplied by Howard.
10756 </p>
10757 </blockquote>
10758
10759 <p><i>[
10760 2009-07 Frankfurt:
10761 ]</i></p>
10762
10763
10764 <blockquote>
10765 Howard is going to write wording.
10766 </blockquote>
10767
10768 <p><i>[
10769 2009-07-26 Howard provided wording.
10770 ]</i></p>
10771
10772
10773 <p><i>[
10774 2009-09-13 Niels adds:
10775 ]</i></p>
10776
10777
10778 <blockquote>
10779 Note: The proposed change of 27.9.1.3 [filebuf.assign]/1 depends on the
10780 resolution of LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a>, which allows implementations to assume that
10781 <tt>*this</tt> and <tt>rhs</tt> refer to different objects.
10782 </blockquote>
10783
10784 <p><i>[
10785 2009 Santa Cruz:
10786 ]</i></p>
10787
10788
10789 <blockquote>
10790 Leave as Open.  Too closely related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#911">911</a> to move on at this time.
10791 </blockquote>
10792
10793
10794
10795 <p><b>Proposed resolution:</b></p>
10796
10797 <p>
10798 Change 27.8.1.2 [stringbuf.assign]/1:
10799 </p>
10800
10801 <blockquote><pre>basic_stringbuf&amp; operator=(basic_stringbuf&amp;&amp; rhs);
10802 </pre>
10803 <blockquote>
10804 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10805 <ins>After the move assignment <tt>*this</tt> reflects the same observable
10806 state it would have if it had been move constructed from <tt>rhs</tt>
10807 (27.8.1.1 [stringbuf.cons]).
10808 </ins>
10809 </blockquote>
10810 </blockquote>
10811
10812 <p>
10813 Change 27.8.2.2 [istringstream.assign]/1:
10814 </p>
10815
10816 <blockquote><pre>basic_istringstream&amp; operator=(basic_istringstream&amp;&amp; rhs);
10817 </pre>
10818 <blockquote>
10819 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10820 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10821 base and members of <tt>rhs</tt>.
10822 </ins>
10823 </blockquote>
10824 </blockquote>
10825
10826 <p>
10827 Change 27.8.3.2 [ostringstream.assign]/1:
10828 </p>
10829
10830 <blockquote><pre>basic_ostringstream&amp; operator=(basic_ostringstream&amp;&amp; rhs);
10831 </pre>
10832 <blockquote>
10833 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10834 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10835 base and members of <tt>rhs</tt>.
10836 </ins>
10837 </blockquote>
10838 </blockquote>
10839
10840 <p>
10841 Change 27.8.5.1 [stringstream.assign]/1:
10842 </p>
10843
10844 <blockquote><pre>basic_stringstream&amp; operator=(basic_stringstream&amp;&amp; rhs);
10845 </pre>
10846 <blockquote>
10847 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10848 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10849 base and members of <tt>rhs</tt>.
10850 </ins>
10851 </blockquote>
10852 </blockquote>
10853
10854 <p>
10855 Change 27.9.1.3 [filebuf.assign]/1:
10856 </p>
10857
10858 <blockquote><pre>basic_filebuf&amp; operator=(basic_filebuf&amp;&amp; rhs);
10859 </pre>
10860 <blockquote>
10861 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10862 <ins>Begins by calling <tt>this-&gt;close()</tt>.
10863 After the move assignment <tt>*this</tt> reflects the same observable
10864 state it would have if it had been move constructed from <tt>rhs</tt>
10865 (27.9.1.2 [filebuf.cons]).
10866 </ins>
10867 </blockquote>
10868 </blockquote>
10869
10870 <p>
10871 Change 27.9.1.8 [ifstream.assign]/1:
10872 </p>
10873
10874 <blockquote><pre>basic_ifstream&amp; operator=(basic_ifstream&amp;&amp; rhs);
10875 </pre>
10876 <blockquote>
10877 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10878 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10879 base and members of <tt>rhs</tt>.</ins>
10880 </blockquote>
10881 </blockquote>
10882
10883 <p>
10884 Change 27.9.1.12 [ofstream.assign]/1:
10885 </p>
10886
10887 <blockquote><pre>basic_ofstream&amp; operator=(basic_ofstream&amp;&amp; rhs);
10888 </pre>
10889 <blockquote>
10890 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10891 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10892 base and members of <tt>rhs</tt>.</ins>
10893 </blockquote>
10894 </blockquote>
10895
10896 <p>
10897 Change 27.9.1.16 [fstream.assign]/1:
10898 </p>
10899
10900 <blockquote><pre>basic_fstream&amp; operator=(basic_fstream&amp;&amp; rhs);
10901 </pre>
10902 <blockquote>
10903 -1- <i>Effects:</i> <del><tt>swap(rhs)</tt>.</del>
10904 <ins>Move assigns the base and members of <tt>*this</tt> with the respective
10905 base and members of <tt>rhs</tt>.</ins>
10906 </blockquote>
10907 </blockquote>
10908
10909
10910
10911
10912
10913
10914 <hr>
10915 <h3><a name="910"></a>910. Effects of MoveAssignable</h3>
10916 <p><b>Section:</b> 20.2.1 [utility.arg.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
10917  <b>Submitter:</b> Alberto Ganesh Barbati <b>Opened:</b> 2008-09-29  <b>Last modified:</b> 2009-11-03</p>
10918 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#utility.arg.requirements">active issues</a> in [utility.arg.requirements].</p>
10919 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#utility.arg.requirements">issues</a> in [utility.arg.requirements].</p>
10920 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
10921 <p><b>Discussion:</b></p>
10922 <p><b>Addresses UK 150</b></p>
10923
10924 <p>
10925 The description of the effect of <tt>operator=</tt> in the <tt>MoveAssignable</tt>
10926 concept, given in paragraph 7 is:
10927 </p>
10928
10929 <blockquote><pre>result_type  T::operator=(T&amp;&amp;  rv);  // inherited from HasAssign&lt;T, T&amp;&amp;&gt;
10930 </pre>
10931
10932 <blockquote>
10933 <i>Postconditions:</i> the constructed <tt>T</tt> object is equivalent to the value of
10934 <tt>rv</tt> before the assignment. [<i>Note:</i> there is no
10935 requirement on the value of <tt>rv</tt> after the assignment.  <i>--end note</i>]
10936 </blockquote>
10937 </blockquote>
10938
10939 <p>
10940 The sentence contains a typo (what is the "constructed <tt>T</tt> object"?)
10941 probably due to a cut&amp;paste from <tt>MoveConstructible</tt>. Moreover, the
10942 discussion of LWG issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#675">675</a> shows that the postcondition is too generic
10943 and might not reflect the user expectations. An implementation of the
10944 move assignment that just calls <tt>swap()</tt> would always fulfill the
10945 postcondition as stated, but might have surprising side-effects in case
10946 the source rvalue refers to an object that is not going to be
10947 immediately destroyed. See LWG issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#900">900</a> for another example. Due to
10948 the sometimes intangible nature of the "user expectation", it seems
10949 difficult to have precise normative wording that could cover all cases
10950 without introducing unnecessary restrictions. However a non-normative
10951 clarification could be a very helpful warning sign that swapping is not
10952 always the correct thing to do.
10953 </p>
10954
10955 <p><i>[
10956 2009-05-09 Alisdair adds:
10957 ]</i></p>
10958
10959
10960 <blockquote>
10961 <p>
10962 Issue 910 is exactly the reason BSI advanced the Editorial comment UK-150.
10963 </p>
10964 <p>
10965 The post-conditions after assignment are at a minimum that the object
10966 referenced by rv must be safely destructible, and the transaction should not
10967 leak resources.  Ideally it should be possible to simply assign rv a new
10968 valid state after the call without invoking undefined behaviour, but any
10969 other use of the referenced object would depend upon additional guarantees
10970 made by that type.
10971 </p>
10972 </blockquote>
10973
10974 <p><i>[
10975 2009-05-09 Howard adds:
10976 ]</i></p>
10977
10978
10979 <blockquote>
10980 <p>
10981 The intent of the rvalue reference work is that the moved from <tt>rv</tt> is
10982 a valid object.  Not one in a singular state.  If, for example, the moved from
10983 object is a <tt>vector</tt>, one should be able to do anything on that moved-from
10984 <tt>vector</tt> that you can do with any other <tt>vector</tt>.  However you would
10985 first have to query it to find out what its current state is.  E.g. it might have <tt>capacity</tt>,
10986 it might not.  It might have a non-zero <tt>size</tt>, it might not.  But regardless,
10987 you can <tt>push_back</tt> on to it if you want.
10988 </p>
10989
10990 <p>
10991 That being said, most standard code is now conceptized.  That is, the concepts
10992 list the only operations that can be done with templated types - whether or not
10993 the values have been moved from.
10994 </p>
10995
10996 <p>
10997 Here is user-written code which must be allowed to be legal:
10998 </p>
10999 <blockquote><pre>#include &lt;vector&gt;
11000 #include &lt;cstdio&gt;
11001
11002 template &lt;class Allocator&gt;
11003 void
11004 inspect(std::vector&lt;double, Allocator&gt;&amp;&amp; v)
11005 {
11006     std::vector&lt;double, Allocator&gt; result(move(v));
11007     std::printf("moved from vector has %u size and %u capacity\n", v.size(), v.capacity());
11008     std::printf("The contents of the vector are:\n");
11009     typedef typename std::vector&lt;double, Allocator&gt;::iterator I;
11010     for (I i = v.begin(), e = v.end(); i != e; ++i)
11011         printf("%f\n", *i);
11012 }
11013
11014 int main()
11015 {
11016     std::vector&lt;double&gt; v1(100, 5.5);
11017     inspect(move(v1));
11018 }
11019 </pre></blockquote>
11020
11021 <p>
11022 The above program does not treat the moved-from <tt>vector</tt> as singular.  It
11023 only treats it as a <tt>vector</tt> with an unknown value.
11024 </p>
11025 <p>
11026 I believe the current proposed wording is consistent with my view on this.
11027 </p>
11028 </blockquote>
11029
11030 <p><i>[
11031 Batavia (2009-05):
11032 ]</i></p>
11033
11034 <blockquote>
11035 We agree that the proposed resolution
11036 is an improvement over the current wording.
11037 </blockquote>
11038
11039 <p><i>[
11040 2009-07 Frankfurt:
11041 ]</i></p>
11042
11043
11044 <blockquote>
11045 Need to look at again without concepts.
11046 </blockquote>
11047
11048 <p><i>[
11049 2009-07 Frankfurt:
11050 ]</i></p>
11051
11052
11053 <blockquote>
11054 Walter will consult with Dave and Doug.
11055 </blockquote>
11056
11057 <p><i>[
11058 2009-10 Santa Cruz:
11059 ]</i></p>
11060
11061
11062 <blockquote>
11063 We believe this is handled by the resolution to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a>,
11064 but there is to much going on in this area to be sure.  Defer for now.
11065 </blockquote>
11066
11067
11068
11069 <p><b>Proposed resolution:</b></p>
11070 <p>
11071 In  [concept.copymove], replace the postcondition in paragraph 7 with:
11072 </p>
11073
11074 <blockquote>
11075 <i>Postconditions:</i> <tt>*this</tt> is equivalent to the value of <tt>rv</tt> before the
11076 assignment. [<i>Note:</i> there is no requirement on the value of <tt>rv</tt> after the
11077 assignment, but the
11078 effect should be unsurprising to the user even in case <tt>rv</tt> is not
11079 immediately destroyed. This may require that resources previously owned
11080 by <tt>*this</tt> are released instead of transferred to <tt>rv</tt>. <i>-- end note</i>]
11081 </blockquote>
11082
11083
11084
11085
11086
11087 <hr>
11088 <h3><a name="911"></a>911. I/O streams and <tt>move/swap</tt> semantic</h3>
11089 <p><b>Section:</b> 27.7.1 [input.streams], 27.7.2 [output.streams] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
11090  <b>Submitter:</b> Alberto Ganesh Barbati <b>Opened:</b> 2008-09-29  <b>Last modified:</b> 2009-10-20</p>
11091 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
11092 <p><b>Discussion:</b></p>
11093 <p>
11094 Class template <tt>basic_istream</tt>, <tt>basic_ostream</tt> and <tt>basic_iostream</tt>
11095 implements public move constructors, move assignment operators and <tt>swap</tt>
11096 method and free functions. This might induce both the user and the
11097 compiler to think that those types are <tt>MoveConstructible</tt>, <tt>MoveAssignable</tt>
11098 and <tt>Swappable</tt>. However, those class templates fail to fulfill the user
11099 expectations. For example:
11100 </p>
11101
11102 <blockquote><pre>std::ostream os(std::ofstream("file.txt"));
11103 assert(os.rdbuf() == 0); // buffer object is not moved to os, file.txt has been closed
11104
11105 std::vector&lt;std::ostream&gt; v;
11106 v.push_back(std::ofstream("file.txt"));
11107 v.reserve(100); // causes reallocation
11108 assert(v[0].rdbuf() == 0); // file.txt has been closed!
11109
11110 std::ostream&amp;&amp; os1 = std::ofstream("file1.txt");
11111 os1 = std::ofstream("file2.txt");
11112 os1 &lt;&lt; "hello, world"; // still writes to file1.txt, not to file2.txt!
11113
11114 std::ostream&amp;&amp; os1 = std::ofstream("file1.txt");
11115 std::ostream&amp;&amp; os2 = std::ofstream("file2.txt");
11116 std::swap(os1, os2);
11117 os1 &lt;&lt; "hello, world"; // writes to file1.txt, not to file2.txt!
11118 </pre></blockquote>
11119
11120 <p>
11121 This is because the move constructor, the move assignment operator and
11122 <tt>swap</tt> are all implemented through calls to <tt>std::basic_ios</tt> member
11123 functions <tt>move()</tt> and <tt>swap()</tt> that do not move nor swap the controlled
11124 stream buffers. That can't happen because the stream buffers may have
11125 different types.
11126 </p>
11127
11128 <p>
11129 Notice that for <tt>basic_streambuf</tt>, the member function <tt>swap()</tt> is
11130 protected. I believe that is correct and all of <tt>basic_istream</tt>,
11131 <tt>basic_ostream</tt>, <tt>basic_iostream</tt> should do the same as the move ctor, move
11132 assignment operator and swap member function are needed by the derived
11133 <tt>fstream</tt>s and <tt>stringstream</tt>s template. The free swap functions for
11134 <tt>basic_(i|o|io)stream</tt> templates should be removed for the same reason.
11135 </p>
11136
11137 <p><i>[
11138 Batavia (2009-05):
11139 ]</i></p>
11140
11141 <blockquote>
11142 <p>
11143 We note that the rvalue swap functions have already been removed.
11144 </p>
11145 <p>
11146 Bill is unsure about making the affected functions protected;
11147 he believes they may need to be public.
11148 </p>
11149 <p>
11150 We are also unsure about removing the lvalue swap functions as proposed.
11151 </p>
11152 <p>
11153 Move to Open.
11154 </p>
11155 </blockquote>
11156
11157 <p><i>[
11158 2009-07 Frankfurt:
11159 ]</i></p>
11160
11161
11162 <blockquote>
11163 <p>
11164 It's not clear that the use case is compelling.
11165 </p>
11166 <p>
11167 Howard: This needs to be implemented and tested.
11168 </p>
11169 </blockquote>
11170
11171 <p><i>[
11172 2009-07-26 Howard adds:
11173 ]</i></p>
11174
11175
11176 <blockquote>
11177 <p>
11178 I started out thinking I would recommend NAD for this one.  I've turned around
11179 to agree with the proposed resolution (which I've updated to the current draft).
11180 I did not fully understand Ganesh's rationale, and attempt to describe my
11181 improved understanding below.
11182 </p>
11183
11184 <p>
11185 The move constructor, move assignment operator, and swap function are different
11186 for <tt>basic_istream</tt>, <tt>basic_ostream</tt> and <tt>basic_iostream</tt>
11187 than other classes.  A timely conversation with Daniel reminded me of this long
11188 forgotten fact.  These members are sufficiently different that they would be
11189 extremely confusing to use in general, but they are very much needed for derived
11190 clients.
11191 </p>
11192
11193 <ul>
11194 <li>
11195 The move constructor moves everything but the <tt>rdbuf</tt> pointer.
11196 </li>
11197 <li>
11198 The move assignment operator moves everything but the <tt>rdbuf</tt> pointer.
11199 </li>
11200 <li>
11201 The swap function swaps everything but the <tt>rdbuf</tt> pointer.
11202 </li>
11203 </ul>
11204
11205 <p>
11206 The reason for this behavior is that for the std-derived classes (stringstreams,
11207 filestreams), the <tt>rdbuf</tt> pointer points back into the class itself
11208 (self referencing).  It can't be swapped or moved.  But this fact isn't born out
11209 at the <tt>stream</tt> level.  Rather it is born out at the <tt>fstream</tt>/<tt>sstream</tt>
11210 level.  And the lower levels just need to deal with that fact by not messing around
11211 with the <tt>rdbuf</tt> pointer which is stored down at the lower levels.
11212 </p>
11213
11214 <p>
11215 In a nutshell, it is very confusing for all of those who are not so intimately
11216 related with streams that they've implemented them.  And it is even fairly
11217 confusing for some of those who have (including myself).  I do not think it is
11218 safe to swap or move <tt>istreams</tt> or <tt>ostreams</tt> because this will
11219 (by necessary design) separate stream state from streambuffer state.  Derived
11220 classes (such as <tt>fstream</tt> and <tt>stringstream</tt> must be used to
11221 keep the stream state and stream buffer consistently packaged as one unit during
11222 a move or swap.
11223 </p>
11224
11225 <p>
11226 I've implemented this proposal and am living with it day to day.
11227 </p>
11228
11229 </blockquote>
11230
11231 <p><i>[
11232 2009 Santa Cruz:
11233 ]</i></p>
11234
11235
11236 <blockquote>
11237 Leave Open.  Pablo expected to propose alternative wording which would rename
11238 move construction, move assignment and swap, and may or may not make them
11239 protected.  This will impact issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#900">900</a>.
11240 </blockquote>
11241
11242
11243
11244 <p><b>Proposed resolution:</b></p>
11245 <p>
11246 27.7.1.1 [istream]: make the following member functions protected:
11247 </p>
11248
11249 <blockquote><pre>basic_istream(basic_istream&amp;&amp;  rhs);
11250 basic_istream&amp;  operator=(basic_istream&amp;&amp;  rhs);
11251 void  swap(basic_istream&amp;  rhs);
11252 </pre></blockquote>
11253
11254 <p>
11255 Ditto: remove the swap free function signature
11256 </p>
11257
11258 <blockquote><pre><del>// swap:
11259 template &lt;class charT, class traits&gt;
11260   void swap(basic_istream&lt;charT, traits&gt;&amp; x, basic_istream&lt;charT, traits&gt;&amp; y);</del>
11261 </pre></blockquote>
11262
11263 <p>
11264 27.7.1.1.2 [istream.assign]: remove paragraph 4
11265 </p>
11266
11267 <blockquote><pre><del>template &lt;class charT, class traits&gt;
11268   void swap(basic_istream&lt;charT, traits&gt;&amp; x, basic_istream&lt;charT, traits&gt;&amp; y);</del>
11269 </pre>
11270 <blockquote>
11271 <del><i>Effects:</i> <tt>x.swap(y)</tt>.</del>
11272 </blockquote>
11273 </blockquote>
11274
11275 <p>
11276 27.7.1.5 [iostreamclass]: make the following member function protected:
11277 </p>
11278
11279 <blockquote><pre>basic_iostream(basic_iostream&amp;&amp;  rhs);
11280 basic_iostream&amp;  operator=(basic_iostream&amp;&amp;  rhs);
11281 void  swap(basic_iostream&amp;  rhs);
11282 </pre></blockquote>
11283
11284 <p>
11285 Ditto: remove the swap free function signature
11286 </p>
11287
11288 <blockquote><pre><del>template &lt;class charT, class traits&gt;
11289   void swap(basic_iostream&lt;charT, traits&gt;&amp; x, basic_iostream&lt;charT, traits&gt;&amp; y);</del>
11290 </pre></blockquote>
11291
11292 <p>
11293 27.7.1.5.3 [iostream.assign]: remove paragraph 3
11294 </p>
11295
11296 <blockquote><pre><del>template &lt;class charT, class traits&gt;
11297   void swap(basic_iostream&lt;charT, traits&gt;&amp; x, basic_iostream&lt;charT, traits&gt;&amp; y);</del>
11298 </pre>
11299 <blockquote>
11300 <del><i>Effects:</i> <tt>x.swap(y)</tt>.</del>
11301 </blockquote>
11302 </blockquote>
11303
11304 <p>
11305 27.7.2.1 [ostream]: make the following member function protected:
11306 </p>
11307
11308 <blockquote><pre>basic_ostream(basic_ostream&amp;&amp;  rhs);
11309 basic_ostream&amp;  operator=(basic_ostream&amp;&amp;  rhs);
11310 void  swap(basic_ostream&amp;  rhs);
11311 </pre></blockquote>
11312
11313 <p>
11314 Ditto: remove the swap free function signature
11315 </p>
11316
11317 <blockquote><pre><del>// swap:
11318 template &lt;class charT, class traits&gt;
11319   void swap(basic_ostream&lt;charT, traits&gt;&amp; x, basic_ostream&lt;charT, traits&gt;&amp; y);</del>
11320 </pre></blockquote>
11321
11322 <p>
11323 27.7.2.3 [ostream.assign]: remove paragraph 4
11324 </p>
11325
11326 <blockquote><pre><del>template &lt;class charT, class traits&gt;
11327   void swap(basic_ostream&lt;charT, traits&gt;&amp; x, basic_ostream&lt;charT, traits&gt;&amp; y);</del>
11328 </pre>
11329 <blockquote>
11330 <del><i>Effects:</i> <tt>x.swap(y)</tt>.</del>
11331 </blockquote>
11332 </blockquote>
11333
11334
11335
11336
11337
11338
11339 <hr>
11340 <h3><a name="915"></a>915. <tt>minmax</tt> with <tt>initializer_list</tt> should return
11341 <tt>pair</tt> of <tt>T</tt>, not <tt>pair</tt> of <tt>const T&amp;</tt></h3>
11342 <p><b>Section:</b> 25.4.7 [alg.min.max] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
11343  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2008-10-04  <b>Last modified:</b> 2009-10-23</p>
11344 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#alg.min.max">issues</a> in [alg.min.max].</p>
11345 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
11346 <p><b>Discussion:</b></p>
11347 <p>
11348 It seems that the proposed changes for
11349 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2772.pdf">N2772</a>
11350 were not clear enough in
11351 this point:
11352 </p>
11353
11354 <blockquote>
11355 25.4.7 [alg.min.max], before p.23 + p.24 + before p. 27 + p. 28 say that the return
11356 type of the <tt>minmax</tt> overloads with an <tt>initializer_list</tt> is
11357 <tt>pair&lt;const T&amp;, const T&amp;&gt;</tt>,
11358 which is inconsistent with the decision for the other <tt>min/max</tt> overloads which take
11359 a <tt>initializer_list</tt> as argument and return a <tt>T</tt>, not a <tt>const T&amp;</tt>.
11360 Doing otherwise for <tt>minmax</tt> would easily lead to unexpected life-time
11361 problems by using <tt>minmax</tt> instead of <tt>min</tt> and <tt>max</tt> separately.
11362 </blockquote>
11363
11364 <p><i>[
11365 Batavia (2009-05):
11366 ]</i></p>
11367
11368 <blockquote>
11369 We agree with the proposed resolution.
11370 Move to Tentatively Ready.
11371 </blockquote>
11372
11373 <p><i>[
11374 2009-07 Frankfurt
11375 ]</i></p>
11376
11377
11378 <blockquote>
11379 Moved from Tentatively Ready to Open only because the wording needs to be
11380 tweaked for concepts removal.
11381 </blockquote>
11382
11383 <p><i>[
11384 2009-08-18 Daniel adds:
11385 ]</i></p>
11386
11387
11388 <blockquote>
11389 Recommend NAD since the proposed changes have already been performed
11390 as part of editorial work of
11391 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>.
11392 </blockquote>
11393
11394 <p><i>[
11395 2009-10 Santa Cruz:
11396 ]</i></p>
11397
11398
11399 <blockquote>
11400 Can't find initializer_list form of minmax anymore, only variadic
11401 version. Seems like we had an editing clash with concepts. Leave Open,
11402 at least until editorial issues resolved. Bring this to Editor's
11403 attention.
11404 </blockquote>
11405
11406
11407
11408 <p><b>Proposed resolution:</b></p>
11409 <ol>
11410 <li>
11411 <p>
11412 In 25 [algorithms]/2, header <tt>&lt;algorithm&gt;</tt> synopsis change as indicated:
11413 </p>
11414
11415 <blockquote><pre>template&lt;<del>class</del><ins>LessThanComparable</ins> T&gt;
11416 <ins>requires CopyConstructible&lt;T&gt;</ins>
11417 pair&lt;<del>const </del>T<del>&amp;</del>, <del>const </del>T<del>&amp;</del>&gt;
11418 minmax(initializer_list&lt;T&gt; t);
11419
11420 template&lt;class T, <del>class</del><ins>StrictWeakOrder&lt;auto, T&gt;</ins> Compare&gt;
11421 <ins>requires CopyConstructible&lt;T&gt;</ins>
11422 pair&lt;<del>const </del>T<del>&amp;</del>, <del>const </del>T<del>&amp;</del>&gt;
11423 minmax(initializer_list&lt;T&gt; t, Compare comp);
11424 </pre></blockquote>
11425 </li>
11426 <li>
11427 <p>
11428 In 25.4.7 [alg.min.max] change as indicated (Begin: Just before p.20):
11429 </p>
11430 <blockquote><pre>template&lt;<del>class</del><ins>LessThanComparable</ins> T&gt;
11431   <ins>requires CopyConstructible&lt;T&gt;</ins>
11432   pair&lt;<del>const </del>T<del>&amp;</del>, <del>const </del>T<del>&amp;</del>&gt;
11433   minmax(initializer_list&lt;T&gt; t);
11434 </pre>
11435 <blockquote>
11436 <p>
11437 <del>-20- <i>Requires:</i> <tt>T</tt> is <tt>LessThanComparable</tt> and
11438 <tt>CopyConstructible</tt>.</del>
11439 </p>
11440 <p>
11441 -21- <i>Returns:</i> <tt>pair&lt;<del>const </del>T<del>&amp;</del>, <del>const
11442 </del>T<del>&amp;</del>&gt;(x, y)</tt> where <tt>x</tt> is the
11443 smallest value and <tt>y</tt> the largest value in the <tt>initializer_list</tt>.
11444 </p>
11445 </blockquote>
11446
11447 <p>[..]</p>
11448 <pre>template&lt;class T, <del>class</del><ins>StrictWeakOrder&lt;auto, T&gt;</ins> Compare&gt;
11449   <ins>requires CopyConstructible&lt;T&gt;</ins>
11450   pair&lt;<del>const </del>T<del>&amp;</del>, <del>const </del>T<del>&amp;</del>&gt;
11451   minmax(initializer_list&lt;T&gt; t, Compare comp);
11452 </pre>
11453
11454 <blockquote>
11455 <p>
11456 <del>-24- <i>Requires:</i> type <tt>T</tt> is <tt>LessThanComparable</tt> and <tt>CopyConstructible</tt>.</del>
11457 </p>
11458 <p>
11459 -25- <i>Returns:</i> <tt>pair&lt;<del>const </del>T<del>&amp;</del>, <del>const
11460 </del>T<del>&amp;</del>&gt;(x, y)</tt> where <tt>x</tt> is the
11461 smallest value and <tt>y</tt> largest value in the <tt>initializer_list</tt>.
11462 </p>
11463 </blockquote>
11464 </blockquote>
11465 </li>
11466 </ol>
11467
11468
11469
11470
11471
11472
11473 <hr>
11474 <h3><a name="920"></a>920. Ref-qualification support in the library</h3>
11475 <p><b>Section:</b> 20.7.14 [func.memfn] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
11476  <b>Submitter:</b> Bronek Kozicki <b>Opened:</b> 2008-10-06  <b>Last modified:</b> 2009-10-23</p>
11477 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#func.memfn">issues</a> in [func.memfn].</p>
11478 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
11479 <p><b>Duplicate of:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1230">1230</a></p>
11480 <p><b>Discussion:</b></p>
11481 <p>
11482 Daniel Krügler wrote:
11483 </p>
11484
11485 <blockquote>
11486 <p>
11487 Shouldn't above list be completed for &amp;- and &amp;&amp;-qualified
11488 member functions This would cause to add:
11489 </p>
11490 <blockquote><pre>template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11491 unspecified mem_fn(R (T::* pm)(Args...) &amp;);
11492 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11493 unspecified mem_fn(R (T::* pm)(Args...) const &amp;);
11494 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11495 unspecified mem_fn(R (T::* pm)(Args...) volatile &amp;);
11496 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11497 unspecified mem_fn(R (T::* pm)(Args...) const volatile &amp;);
11498 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11499 unspecified mem_fn(R (T::* pm)(Args...) &amp;&amp;);
11500 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11501 unspecified mem_fn(R (T::* pm)(Args...) const &amp;&amp;);
11502 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11503 unspecified mem_fn(R (T::* pm)(Args...) volatile &amp;&amp;);
11504 template&lt;Returnable R, class T, CopyConstructible... Args&gt;
11505 unspecified mem_fn(R (T::* pm)(Args...) const volatile &amp;&amp;);
11506 </pre></blockquote>
11507
11508 </blockquote>
11509
11510 <p>
11511 yes, absolutely. Thanks for spotting this. Without this change <tt>mem_fn</tt>
11512 cannot be initialized from pointer to ref-qualified member function. I
11513 believe semantics of such function pointer is well defined.
11514 </p>
11515
11516 <p><i>[
11517 Post Summit Daniel provided wording.
11518 ]</i></p>
11519
11520
11521 <p><i>[
11522 Batavia (2009-05):
11523 ]</i></p>
11524
11525 <blockquote>
11526 <p>
11527 We need to think about whether we really want to go down the proposed path
11528 of combinatorial explosion.
11529 Perhaps a Note would suffice.
11530 </p>
11531 <p>
11532 We would really like to have an implementation before proceeding.
11533 </p>
11534 <p>
11535 Move to Open, and recommend this be deferred until after the next
11536 Committee Draft has been issued.
11537 </p>
11538 </blockquote>
11539
11540 <p><i>[
11541 2009-10-10 Daniel updated wording to post-concepts.
11542 ]</i></p>
11543
11544
11545 <blockquote>
11546 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1230">1230</a> has a similar proposed resolution
11547 </blockquote>
11548
11549 <p><i>[
11550 2009-10 Santa Cruz:
11551 ]</i></p>
11552
11553
11554 <blockquote>
11555 Move to Ready.
11556 </blockquote>
11557
11558
11559
11560 <p><b>Proposed resolution:</b></p>
11561 <ol>
11562 <li>
11563 <p>
11564 Change 20.7 [function.objects]/2, header
11565 <tt>&lt;functional&gt;</tt> synopsis as follows:
11566 </p>
11567
11568 <blockquote><pre>// 20.7.14, member function adaptors:
11569 template&lt;class R, class T&gt; <i>unspecified</i> mem_fn(R T::*);
11570
11571 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...));</ins>
11572 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const);</ins>
11573 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) volatile);</ins>
11574 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const volatile);</ins>
11575
11576 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) &amp;);</ins>
11577 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const &amp;);</ins>
11578 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) volatile &amp;);</ins>
11579 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const volatile &amp;);</ins>
11580
11581 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) &amp;&amp;);</ins>
11582 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const &amp;&amp;);</ins>
11583 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) volatile &amp;&amp;);</ins>
11584 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::*)(Args...) const volatile &amp;&amp;);</ins>
11585 </pre></blockquote>
11586 </li>
11587
11588 <li>
11589 <p>
11590 Change the prototype list of 20.7.14 [func.memfn] as follows [NB: The
11591 following text, most notably p.2 and p.3 which
11592 discuss influence of the cv-qualification on the definition of the
11593 base class's first template parameter remains
11594 unchanged. ]:
11595 </p>
11596
11597 <blockquote><pre>template&lt;class R, class T&gt; <i>unspecified</i> mem_fn(R T::* pm);
11598
11599 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...));</ins>
11600 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const);</ins>
11601 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) volatile);</ins>
11602 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const volatile);</ins>
11603
11604 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) &amp;);</ins>
11605 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const &amp;);</ins>
11606 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) volatile &amp;);</ins>
11607 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const volatile &amp;);</ins>
11608
11609 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) &amp;&amp;);</ins>
11610 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const &amp;&amp;);</ins>
11611 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) volatile &amp;&amp;);</ins>
11612 <ins>template&lt;class R, class T, class ...Args&gt; <i>unspecified</i> mem_fn(R (T::* pm)(Args...) const volatile &amp;&amp;);</ins>
11613 </pre></blockquote>
11614 </li>
11615
11616 <li>
11617 <p>
11618 Remove 20.7.14 [func.memfn]/5:
11619 </p>
11620
11621 <blockquote>
11622 <del><i>Remarks:</i> Implementations may implement <tt>mem_fn</tt> as a set of
11623 overloaded function templates.</del>
11624 </blockquote>
11625 </li>
11626 </ol>
11627
11628
11629
11630
11631
11632
11633 <hr>
11634 <h3><a name="921"></a>921. Rational Arithmetic should use template aliases</h3>
11635 <p><b>Section:</b> 20.4.1 [ratio.ratio] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
11636  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2008-10-07  <b>Last modified:</b> 2009-10-21</p>
11637 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#ratio.ratio">issues</a> in [ratio.ratio].</p>
11638 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
11639 <p><b>Discussion:</b></p>
11640 <p>
11641 The compile-time functions that operate on <tt>ratio&lt;N,D&gt;</tt> require the
11642 cumbersome and error-prone "evaluation" of a <tt>type</tt> member using a
11643 meta-programming style that predates the invention of template aliases.
11644 Thus, multiplying three ratios <tt>a</tt>, <tt>b</tt>, and <tt>c</tt> requires the expression:
11645 </p>
11646
11647 <blockquote><pre>ratio_multiply&lt;a, ratio_multiply&lt;b, c&gt;::type&gt;::type
11648 </pre></blockquote>
11649
11650 <p>
11651 The simpler expression:
11652 </p>
11653
11654 <blockquote><pre>ratio_multiply&lt;a, ratio_multiply&lt;b, c&gt;&gt;
11655 </pre></blockquote>
11656
11657 <p>
11658 Could be used by if template aliases were employed in the definitions.
11659 </p>
11660
11661 <p><i>[
11662 Post Summit:
11663 ]</i></p>
11664
11665
11666 <blockquote>
11667 <p>
11668 Jens: not a complete proposed resolution: "would need to make similar change"
11669 </p>
11670 <p>
11671 Consensus: We agree with the direction of the issue.
11672 </p>
11673 <p>
11674 Recommend Open.
11675 </p>
11676 </blockquote>
11677
11678 <p><i>[
11679 2009-05-11 Daniel adds:
11680 ]</i></p>
11681
11682
11683 <blockquote>
11684 <p>
11685 Personally I'm <em>not</em> in favor for the addition of:
11686 </p>
11687 <blockquote><pre>typedef ratio type;
11688 </pre></blockquote>
11689 <p>
11690 For a reader of the
11691 standard it's usage or purpose is unclear. I haven't seen similar examples
11692 of attempts to satisfy non-feature complete compilers.
11693 </p>
11694 </blockquote>
11695
11696 <p><i>[
11697 2009-05-11 Pablo adds:
11698 ]</i></p>
11699
11700
11701 <blockquote>
11702 <p>
11703 The addition of type to the <tt>ratio</tt> template allows the previous style
11704 (i.e., in the prototype implementations) to remain valid and permits the
11705 use of transitional library implementations for C++03 compilers.  I do
11706 not feel strongly about its inclusion, however, and leave it up to the
11707 reviewers to decide.
11708 </p>
11709 </blockquote>
11710
11711 <p><i>[
11712 Batavia (2009-05):
11713 ]</i></p>
11714
11715 <blockquote>
11716 Bill asks for additional discussion in the issue
11717 that spells out more details of the implementation.
11718 Howard points us to issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#948">948</a>
11719 which has at least most of the requested details.
11720 Tom is strongly in favor of overflow-checking at compile time.
11721 Pete points out that there is no change of functionality implied.
11722 We agree with the proposed resolution,
11723 but recommend moving the issue to Review
11724 to allow time to improve the discussion if needed.
11725 </blockquote>
11726
11727 <p><i>[
11728 2009-07-21 Alisdair adds:
11729 ]</i></p>
11730
11731
11732 <blockquote>
11733 See <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1121">1121</a> for a potentially incompatible proposal.
11734 </blockquote>
11735
11736 <p><i>[
11737 2009-10 Santa Cruz:
11738 ]</i></p>
11739
11740
11741 <blockquote>
11742 Move to Ready.
11743 </blockquote>
11744
11745
11746
11747 <p><b>Proposed resolution:</b></p>
11748
11749
11750  <ol start="0">
11751 <li>
11752 <p>
11753 In 20.4 [ratio]/3 change as indicated:
11754 </p>
11755
11756 <blockquote><pre>// ratio arithmetic
11757 template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_add<ins> = <em>see below</em></ins>;
11758 template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_subtract<ins> = <em>see below</em></ins>;
11759 template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_multiply<ins> = <em>see below</em></ins>;
11760 template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_divide<ins> = <em>see below</em></ins>;
11761 </pre></blockquote>
11762 </li>
11763 <li>
11764 <p>
11765 In 20.4.1 [ratio.ratio], change as indicated:
11766 </p>
11767 <blockquote><pre>namespace std {
11768   template &lt;intmax_t N, intmax_t D = 1&gt;
11769   class ratio {
11770   public:
11771     <ins>typedef ratio type;</ins>
11772     static const intmax_t num;
11773     static const intmax_t den;
11774   };
11775 }
11776 </pre></blockquote>
11777 </li>
11778 <li>
11779 <p>
11780 In 20.4.2 [ratio.arithmetic] change as indicated:
11781 </p>
11782
11783 <blockquote><pre>template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_add<ins> = <em>see below</em></ins><del>{
11784   typedef <em>see below</em> type;
11785 }</del>;
11786 </pre>
11787
11788 <blockquote>
11789 <p>
11790 1 The <del>nested typedef</del> type <tt><ins>ratio_add&lt;R1, R2&gt;</ins></tt>
11791 shall be a synonym for <tt>ratio&lt;T1, T2&gt;</tt>
11792 where <tt>T1</tt> has the value <tt>R1::num * R2::den + R2::num * R1::den</tt> and <tt>T2</tt>
11793 has the value <tt>R1::den * R2::den</tt>.
11794 </p>
11795 </blockquote>
11796 </blockquote>
11797 <blockquote><pre>template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_subtract<ins> = <em>see below</em></ins><del>{
11798   typedef <em>see below</em> type;
11799 }</del>;
11800 </pre>
11801 <blockquote>
11802 <p>
11803 2 The <del>nested typedef</del> type <tt><ins>ratio_subtract&lt;R1, R2&gt;</ins></tt>
11804 shall be a synonym for <tt>ratio&lt;T1, T2&gt;</tt>
11805 where <tt>T1</tt> has the value <tt>R1::num * R2::den - R2::num * R1::den</tt> and <tt>T2</tt>
11806 has the value <tt>R1::den * R2::den</tt>.
11807 </p>
11808 </blockquote>
11809 </blockquote>
11810 <blockquote><pre>template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_multiply<ins> = <em>see below</em></ins><del>{
11811   typedef <em>see below</em> type;
11812 }</del>;
11813 </pre>
11814 <blockquote>
11815 <p>
11816 3 The <del>nested typedef</del> type <tt><ins>ratio_multiply&lt;R1, R2&gt;</ins></tt>
11817 shall be a synonym for <tt>ratio&lt;T1, T2&gt;</tt>
11818 where <tt>T1</tt> has the value <tt>R1::num * R2::num</tt> and <tt>T2</tt> has the value <tt>R1::den * R2::den</tt>.
11819 </p>
11820 </blockquote>
11821 </blockquote>
11822 <blockquote><pre>template &lt;class R1, class R2&gt; <del>struct</del><ins>using</ins> ratio_divide<ins> = <em>see below</em></ins><del>{
11823   typedef <em>see below</em> type;
11824 }</del>;
11825 </pre>
11826 <blockquote>
11827 <p>
11828 4 The <del>nested typedef</del> type <tt><ins>ratio_divide&lt;R1, R2&gt;</ins></tt>
11829 shall be a synonym for <tt>ratio&lt;T1, T2&gt;</tt>
11830 where <tt>T1</tt> has the value <tt>R1::num * R2::den</tt> and <tt>T2</tt> has the value <tt>R1::den * R2::num</tt>.
11831 </p>
11832 </blockquote>
11833 </blockquote>
11834 </li>
11835 <li>
11836 <p>
11837 In 20.9.3.1 [time.duration.cons]/4 change as indicated:
11838 </p>
11839 <blockquote>
11840 <p>
11841 <i>Requires:</i> <tt>treat_as_floating_point&lt;rep&gt;::value</tt> shall be true or
11842 <tt>ratio_divide&lt;Period2, period&gt;::<del>type::</del>den</tt> shall be 1.[..]
11843 </p>
11844 </blockquote>
11845 </li>
11846 <li>
11847 <p>
11848 In 20.9.3.7 [time.duration.cast]/2 change as indicated:
11849 </p>
11850 <blockquote>
11851 <p>
11852 <i>Returns:</i> Let CF be <tt>ratio_divide&lt;Period, typename
11853 ToDuration::period&gt;<del>::type</del></tt>, and [..]
11854 </p>
11855 </blockquote>
11856 </li>
11857 </ol>
11858
11859
11860
11861
11862
11863 <hr>
11864 <h3><a name="929"></a>929. Thread constructor</h3>
11865 <p><b>Section:</b> 30.3.1.2 [thread.thread.constr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
11866  <b>Submitter:</b> Anthony Williams <b>Opened:</b> 2008-10-23  <b>Last modified:</b> 2009-10-25</p>
11867 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.constr">active issues</a> in [thread.thread.constr].</p>
11868 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.constr">issues</a> in [thread.thread.constr].</p>
11869 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
11870 <p><b>Discussion:</b></p>
11871
11872 <p><b>Addresses UK 323</b></p>
11873
11874 <p>
11875 The <tt>thread</tt> constructor for starting a new thread with a function and
11876 arguments is overly constrained by the signature requiring rvalue
11877 references for <tt>func</tt> and <tt>args</tt> and the <tt>CopyConstructible</tt> requirements
11878 for the elements of <tt>args</tt>. The use of an rvalue reference for the
11879 function restricts the potential use of a plain function name, since
11880 the type of the bound parameter will be deduced to be a function
11881 reference and decay to pointer-to-function will not happen. This
11882 therefore complicates the implementation in order to handle a simple
11883 case. Furthermore, the use of rvalue references for args prevents the
11884 array to pointer decay. Since arrays are not <tt>CopyConstructible</tt> or even
11885 <tt>MoveConstructible</tt>, this essentially prevents the passing of arrays as
11886 parameters. In particular it prevents the passing of string literals.
11887 Consequently a simple case such as
11888 </p>
11889
11890 <blockquote><pre>void f(const char*);
11891 std::thread t(f,"hello");
11892 </pre></blockquote>
11893
11894 <p>
11895 is ill-formed since the type of the string literal is <tt>const char[6]</tt>.
11896 </p>
11897
11898 <p>
11899 By changing the signature to take all parameters by value we can
11900 eliminate the <tt>CopyConstructible</tt> requirement and permit the use of
11901 arrays, as the parameter passing semantics will cause the necessary
11902 array-to-pointer decay. They will also cause the function name to
11903 decay to a pointer to function and allow the implementation to handle
11904 functions and function objects identically.
11905 </p>
11906
11907 <p>
11908 The new signature of the <tt>thread</tt> constructor for a function and
11909 arguments is thus:
11910 </p>
11911
11912 <blockquote><pre>template&lt;typename F,typename... Args&gt;
11913 thread(F,Args... args);
11914 </pre></blockquote>
11915
11916 <p>
11917 Since the parameter pack <tt>Args</tt> can be empty, the single-parameter
11918 constructor that takes just a function by value is now redundant.
11919 </p>
11920
11921 <p><i>[
11922 Howard adds:
11923 ]</i></p>
11924
11925
11926 <blockquote>
11927 <p>
11928 I agree with everything Anthony says in this issue.  However I believe we
11929 can optimize in such a way as to get the pass-by-value behavior with the
11930 pass-by-rvalue-ref performance.  The performance difference is that the latter
11931 removes a <tt>move</tt> when passing in an lvalue.
11932 </p>
11933
11934 <p>
11935 This circumstance is very analogous to <tt>make_pair</tt> (20.3.4 [pairs])
11936 where we started with passing by const reference, changed to pass by value to
11937 get pointer decay, and then changed to pass by rvalue reference, but modified with
11938 <tt>decay&lt;T&gt;</tt> to retain the pass-by-value behavior.  If we were to
11939 apply the same solution here it would look like:
11940 </p>
11941
11942 <blockquote><pre><del>template &lt;class F&gt; explicit thread(F f);</del>
11943 template &lt;class F, class ...Args&gt; thread(F&amp;&amp; f, Args&amp;&amp;... args);
11944 </pre>
11945 <blockquote>
11946 <p>
11947 -4- <i>Requires:</i> <tt>F</tt> and each <tt>Ti</tt> in <tt>Args</tt> shall be <del><tt>CopyConstructible</tt>
11948 if an lvalue and otherwise</del> <tt>MoveConstructible</tt>.
11949 <tt><i>INVOKE</i>(f, w1, w2, ..., wN)</tt> (20.7.2 [func.require]) shall be a valid expression for
11950 some values <tt>w1, w2, ... , wN,</tt> where <tt>N == sizeof...(Args)</tt>.
11951 </p>
11952 <p>
11953 -5- <i>Effects:</i> Constructs an object of type <tt>thread</tt>
11954 <del>and executes <tt><i>INVOKE</i>(f, t1, t2, ..., tN)</tt> in a new
11955 thread of execution, where <tt>t1, t2, ..., tN</tt> are the values in <tt>args...</tt></del>.
11956 <ins>Constructs
11957 the following objects in memory which is accessible to a new thread of execution
11958 as if:</ins>
11959 </p>
11960 <blockquote><pre><ins>typename decay&lt;F&gt;::type g(std::forward&lt;F&gt;(f));</ins>
11961 <ins>tuple&lt;typename decay&lt;Args&gt;::type...&gt; w(std::forward&lt;Args&gt;(args)...);</ins>
11962 </pre></blockquote>
11963 <p>
11964 <ins>The new thread of
11965 execution executes <tt><i>INVOKE</i>(g, wi...)</tt> where the <tt>wi...</tt> refers
11966 to the elements stored in the <tt>tuple w</tt>.</ins>
11967 Any return value from <tt>g</tt> is ignored.
11968 <del>If <tt>f</tt> terminates with an uncaught exception, <tt>std::terminate()</tt> shall be called.</del>
11969 <ins>If the evaluation of <tt><i>INVOKE</i>(g,  wi...)</tt> terminates
11970 with an uncaught exception, <tt>std::terminate()</tt> shall be called [<i>Note:</i>
11971 <tt>std::terminate()</tt> could be called before entering <tt>g</tt>. -- <i>end note</i>]. Any
11972 exception thrown before the evaluation of <tt><i>INVOKE</i></tt> has started shall be
11973 catchable in the calling thread.</ins>
11974 </p>
11975 </blockquote>
11976 </blockquote>
11977
11978 <p>
11979 Text referring to when <tt>terminate()</tt> is called was contributed by Ganesh.
11980 </p>
11981
11982 </blockquote>
11983
11984 <p><i>[
11985 Batavia (2009-05):
11986 ]</i></p>
11987
11988 <blockquote>
11989 We agree with the proposed resolution,
11990 but would like the final sentence to be reworded
11991 since "catchable" is not a term of art (and is used nowhere else).
11992 </blockquote>
11993
11994 <p><i>[
11995 2009-07 Frankfurt:
11996 ]</i></p>
11997
11998
11999 <blockquote>
12000 <p>
12001 This is linked to
12002 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2901.pdf">N2901</a>.
12003 </p>
12004 <p>
12005 Howard to open a separate issue to remove (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1176">1176</a>).
12006 </p>
12007 <p>
12008 In Frankfurt there is no consensus for removing the variadic constructor.
12009 </p>
12010 </blockquote>
12011
12012 <p><i>[
12013 2009-10 Santa Cruz:
12014 ]</i></p>
12015
12016
12017 <blockquote>
12018 We want to move forward with this issue.  If we later take it out via <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1176">1176</a>
12019 then that's ok too.  Needs small group to improve wording.
12020 </blockquote>
12021
12022 <p><i>[
12023 2009-10 Santa Cruz:
12024 ]</i></p>
12025
12026
12027 <blockquote>
12028 <p>
12029 Stefanus provided revised wording.  Moved to Review  Here is the original wording:
12030 </p>
12031 <blockquote>
12032 <p>
12033 Modify the class definition of <tt>std::thread</tt> in 30.3.1 [thread.thread.class] to remove the
12034 following signature:
12035 </p>
12036
12037 <blockquote><pre><del>template&lt;class F&gt; explicit thread(F f);</del>
12038 template&lt;class F, class ... Args&gt; <ins>explicit</ins> thread(F&amp;&amp; f, Args&amp;&amp; ... args);
12039 </pre></blockquote>
12040
12041 <p>
12042 Modify 30.3.1.2 [thread.thread.constr] to replace the constructors prior to paragraph 4 with
12043 the single constructor as above. Replace paragraph 4 - 6 with the
12044 following:
12045 </p>
12046
12047 <blockquote>
12048 <p>
12049 -4- <i>Requires:</i> <tt>F</tt> and each <tt>Ti</tt> in <tt>Args</tt> shall be <del><tt>CopyConstructible</tt>
12050 if an lvalue and otherwise</del> <tt>MoveConstructible</tt>.
12051 <tt><i>INVOKE</i>(f, w1, w2, ..., wN)</tt> (20.7.2 [func.require]) shall be a valid expression for
12052 some values <tt>w1, w2, ... , wN,</tt> where <tt>N == sizeof...(Args)</tt>.
12053 </p>
12054 <p>
12055 -5- <i>Effects:</i> Constructs an object of type <tt>thread</tt>
12056 <del>and executes <tt><i>INVOKE</i>(f, t1, t2, ..., tN)</tt> in a new
12057 thread of execution, where <tt>t1, t2, ..., tN</tt> are the values in <tt>args...</tt></del>.
12058 <ins>Constructs
12059 the following objects:</ins>
12060 </p>
12061 <blockquote><pre><ins>typename decay&lt;F&gt;::type g(std::forward&lt;F&gt;(f));</ins>
12062 <ins>tuple&lt;typename decay&lt;Args&gt;::type...&gt; w(std::forward&lt;Args&gt;(args)...);</ins>
12063 </pre></blockquote>
12064 <p>
12065 <ins>and executes <tt><i>INVOKE</i>(g, wi...)</tt> in a new thread of execution.
12066 These objects shall be destroyed when the new thread of execution completes.</ins>
12067 Any return value from <tt>g</tt> is ignored.
12068 <del>If <tt>f</tt> terminates with an uncaught exception, <tt>std::terminate()</tt> shall be called.</del>
12069 <ins>If the evaluation of <tt><i>INVOKE</i>(g,  wi...)</tt> terminates
12070 with an uncaught exception, <tt>std::terminate()</tt> shall be called [<i>Note:</i>
12071 <tt>std::terminate()</tt> could be called before entering <tt>g</tt>. -- <i>end note</i>]. Any
12072 exception thrown before the evaluation of <tt><i>INVOKE</i></tt> has started shall be
12073 catchable in the calling thread.</ins>
12074 </p>
12075 <p>
12076 -6- <i>Synchronization:</i> The invocation of the constructor <i>happens before</i> the
12077 invocation of <del><tt>f</tt></del> <ins><tt>g</tt></ins>.
12078 </p>
12079 </blockquote>
12080
12081 </blockquote>
12082 </blockquote>
12083
12084
12085 <p><b>Proposed resolution:</b></p>
12086 <p>
12087 Modify the class definition of <tt>std::thread</tt> in 30.3.1 [thread.thread.class] to remove the
12088 following signature:
12089 </p>
12090
12091 <blockquote><pre><del>template&lt;class F&gt; explicit thread(F f);</del>
12092 template&lt;class F, class ... Args&gt; <ins>explicit</ins> thread(F&amp;&amp; f, Args&amp;&amp; ... args);
12093 </pre></blockquote>
12094
12095 <p>
12096 Modify 30.3.1.2 [thread.thread.constr] to replace the constructors prior to paragraph 4 with
12097 the single constructor as above. Replace paragraph 4 - 6 with the
12098 following:
12099 </p>
12100
12101 <blockquote>
12102 <p>
12103 -4- <i>Requires:</i> <tt>F</tt> and each <tt>Ti</tt> in <tt>Args</tt>
12104 shall be <del><tt>CopyConstructible</tt> if an lvalue and
12105 otherwise</del> <tt>MoveConstructible</tt>. <tt><i>INVOKE</i>(f, w1, w2,
12106 ..., wN)</tt> (20.7.2 [func.require]) shall be a valid expression
12107 for some values <tt>w1, w2, ... , wN,</tt> where <tt>N ==
12108 sizeof...(Args)</tt>.
12109 </p>
12110
12111 <p>
12112 -5- <i>Effects:</i> Constructs an object of type <tt>thread</tt> <del>and executes
12113 <tt>INVOKE(f, t1, t2, ..., tN)</tt> in a new thread of execution, where
12114 <tt>t1, t2, ..., tN</tt> are the values in <tt>args...</tt></del>.
12115 <ins>Given a function as follows:</ins>
12116 <del>Any return
12117 value from <tt>f</tt> is ignored. If <tt>f</tt> terminates with an
12118 uncaught exception, <tt>std::terminate()</tt> shall be called.</del>
12119 </p>
12120
12121 <blockquote><pre><ins>
12122 template&lt;typename T&gt; typename decay&lt;T&gt;::type decay_copy(T&amp;&amp; v)
12123     { return std::forward&lt;T&gt;(v); }
12124 </ins></pre></blockquote>
12125
12126 <p><ins>
12127 The new thread of execution executes <tt>INVOKE(decay_copy(f),
12128 decay_copy(args)...)</tt> with the calls to <tt>decay_copy()</tt> being evaluated in
12129 the constructing thread. Any return value from this invocation is
12130 ignored. [<i>Note:</i> this implies any exceptions not thrown from the
12131 invocation of the copy of <tt>f</tt> will be thrown in the constructing thread,
12132 not the new thread. &#8212; <i>end note</i>].
12133 </ins></p>
12134
12135 <p>
12136 -6- <i>Synchronization:</i> The invocation of the constructor <i>happens before</i> the
12137 invocation of <ins>copy of</ins> <tt>f</tt>.
12138 </p>
12139 </blockquote>
12140
12141
12142
12143
12144
12145
12146 <hr>
12147 <h3><a name="932"></a>932. <tt>unique_ptr(pointer p)</tt> for pointer deleter types</h3>
12148 <p><b>Section:</b> 20.8.14.2.1 [unique.ptr.single.ctor] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
12149  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2008-11-26  <b>Last modified:</b> 2009-10-22</p>
12150 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unique.ptr.single.ctor">active issues</a> in [unique.ptr.single.ctor].</p>
12151 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unique.ptr.single.ctor">issues</a> in [unique.ptr.single.ctor].</p>
12152 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
12153 <p><b>Discussion:</b></p>
12154
12155 <p><b>Addresses US 79</b></p>
12156
12157 <p>
12158 20.8.14.2.1 [unique.ptr.single.ctor]/5 no longer requires for <tt>D</tt>
12159 not to be a pointer type.  I believe this restriction was accidently removed
12160 when we relaxed the completeness reuqirements on <tt>T</tt>. The restriction
12161 needs to be put back in.  Otherwise we have a run time failure that could
12162 have been caught at compile time:
12163 </p>
12164
12165 <blockquote><pre>{
12166 unique_ptr&lt;int, void(*)(void*)&gt; p1(malloc(sizeof(int)));  <font color="#c80000">// should not compile</font>
12167 }  <font color="#c80000">// p1.~unique_ptr() dereferences a null function pointer</font>
12168 unique_ptr&lt;int, void(*)(void*)&gt; p2(malloc(sizeof(int)), free);  <font color="#c80000">// ok</font>
12169 </pre></blockquote>
12170
12171 <p><i>[
12172 Post Summit:
12173 ]</i></p>
12174
12175
12176 <blockquote>
12177 Recommend Tentatively Ready.
12178 </blockquote>
12179
12180 <p><i>[
12181 2009-07 Frankfurt
12182 ]</i></p>
12183
12184
12185 <blockquote>
12186 Moved from Tentatively Ready to Open only because the wording needs to be
12187 improved for enable_if type constraining, possibly following Robert's
12188 formula.
12189 </blockquote>
12190
12191 <p><i>[
12192 2009-07 Frankfurt:
12193 ]</i></p>
12194
12195
12196 <blockquote>
12197 <p>
12198 We need to consider whether some requirements in the Requires paragraphs
12199 of [unique.ptr] should instead be Remarks.
12200 </p>
12201 <p>
12202 Leave Open. Howard to provide wording, and possibly demonstrate how this
12203 can be implemented using enable_if.
12204 </p>
12205 </blockquote>
12206
12207 <p><i>[
12208 2009-07-27 Howard adds:
12209 ]</i></p>
12210
12211
12212 <blockquote>
12213 <p>
12214 The two constructors to which this issue applies are not easily constrained
12215 with <tt>enable_if</tt> as they are not templated:
12216 </p>
12217
12218 <blockquote><pre>unique_ptr();
12219 explicit unique_ptr(pointer p);
12220 </pre></blockquote>
12221
12222 <p>
12223 To "SFINAE" these constructors away would take heroic effort such as specializing
12224 the entire <tt>unique_ptr</tt> class template on pointer deleter types.  There
12225 is insufficient motivation for such heroics.  Here is the expected and
12226 reasonable implementation for these constructors:
12227 </p>
12228
12229 <blockquote><pre>unique_ptr()
12230     : ptr_(pointer())
12231     {
12232         static_assert(!is_pointer&lt;deleter_type&gt;::value,
12233             "unique_ptr constructed with null function pointer deleter");
12234     }
12235 explicit unique_ptr(pointer p)
12236     : ptr_(p)
12237     {
12238         static_assert(!is_pointer&lt;deleter_type&gt;::value,
12239             "unique_ptr constructed with null function pointer deleter");
12240     }
12241 </pre></blockquote>
12242
12243 <p>
12244 I.e. just use <tt>static_assert</tt> to verify that the constructor is not
12245 instantiated with a function pointer for a deleter.  The compiler will automatically
12246 take care of issuing a diagnostic if the deleter is a reference type (uninitialized
12247 reference error).
12248 </p>
12249
12250 <p>
12251 In keeping with our discussions in Frankfurt, I'm moving this requirement on
12252 the implementation from the Requires paragraph to a Remarks paragraph.
12253 </p>
12254
12255 </blockquote>
12256
12257 <p><i>[
12258 2009-08-17 Daniel adds:
12259 ]</i></p>
12260
12261
12262 <blockquote>
12263 <p>
12264 It is insufficient to require a diagnostic. This doesn't imply an
12265 ill-formed program
12266 as of 1.3.3 [defns.diagnostic] (a typical alternative would be a compiler
12267 warning), but
12268 exactly that seems to be the intend. I suggest to use the following
12269 remark instead:
12270 </p>
12271
12272 <blockquote>
12273 <i>Remarks:</i> The program shall be ill-formed if this constructor is
12274 instantiated when <tt>D</tt> is a pointer type or reference type.
12275 </blockquote>
12276
12277 <p>
12278 Via the general standard rules of 1.4 [intro.compliance] the "diagnostic
12279 required" is implied.
12280 </p>
12281
12282 </blockquote>
12283
12284 <p><i>[
12285 2009-10 Santa Cruz:
12286 ]</i></p>
12287
12288
12289 <blockquote>
12290 Moved to Ready.
12291 </blockquote>
12292
12293
12294
12295 <p><b>Proposed resolution:</b></p>
12296 <p>
12297 Change the description of the default constructor in 20.8.14.2.1 [unique.ptr.single.ctor]:
12298 </p>
12299
12300 <blockquote><pre>unique_ptr();
12301 </pre>
12302 <blockquote>
12303 <p>
12304 -1- <i>Requires:</i> <tt>D</tt> shall be default constructible, and that construction
12305 shall not throw an exception. <del><tt>D</tt> shall
12306 not be a reference type or pointer type (diagnostic required).</del>
12307 </p>
12308 <p>...</p>
12309 <p><ins>
12310 <i>Remarks:</i> The program shall be ill-formed if this constructor is
12311 instantiated when <tt>D</tt> is a pointer type or reference type.
12312
12313 </ins></p>
12314 </blockquote>
12315 </blockquote>
12316
12317 <p>
12318 Add  after 20.8.14.2.1 [unique.ptr.single.ctor]/8:
12319 </p>
12320
12321 <blockquote><pre>unique_ptr(pointer p);
12322 </pre>
12323 <blockquote>
12324 <p>...</p>
12325 <p><ins>
12326 <i>Remarks:</i> The program shall be ill-formed if this constructor is
12327 instantiated when <tt>D</tt> is a pointer type or reference type.
12328
12329 </ins></p>
12330 </blockquote>
12331 </blockquote>
12332
12333
12334
12335
12336
12337 <hr>
12338 <h3><a name="939"></a>939. Problem with <tt>std::identity</tt> and reference-to-temporaries</h3>
12339 <p><b>Section:</b> 20.3.3 [forward] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
12340  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2008-12-11  <b>Last modified:</b> 2009-10-29</p>
12341 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#forward">issues</a> in [forward].</p>
12342 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
12343 <p><b>Discussion:</b></p>
12344 <p>
12345 <tt>std::identity</tt> takes an argument of type <tt>T const &amp;</tt>
12346 and returns a result of <tt>T const &amp;</tt>.
12347 </p>
12348 <p>
12349 Unfortunately, this signature will accept a value of type other than <tt>T</tt> that
12350 is convertible-to-<tt>T</tt>, and then return a reference to the dead temporary.  The
12351 constraint in the concepts version simply protects against returning
12352 reference-to-<tt>void</tt>.
12353 </p>
12354 <p>
12355 Solutions:
12356 </p>
12357 <blockquote>
12358 <p>
12359 i/  Return-by-value, potentially slicing bases and rejecting non-copyable
12360 types
12361 </p>
12362 <p>
12363 ii/ Provide an additional overload:
12364 </p>
12365 <blockquote><pre>template&lt; typename T &gt;
12366 template operator( U &amp; ) = delete;
12367 </pre></blockquote>
12368 <p>
12369 This seems closer on intent, but moves beyond the original motivation for
12370 the operator, which is compatibility with existing (non-standard)
12371 implementations.
12372 </p>
12373 <p>
12374 iii/ Remove the <tt>operator()</tt> overload.  This restores the original definition
12375 of the <tt>identity</tt>, although now effectively a type_trait rather than part of
12376 the perfect forwarding protocol.
12377 </p>
12378 <p>
12379 iv/ Remove <tt>std::identity</tt> completely; its original reason to exist is
12380 replaced with the <tt>IdentityOf</tt> concept.
12381 </p>
12382 </blockquote>
12383 <p>
12384 My own preference is somewhere between (ii) and (iii) - although I stumbled
12385 over the issue with a specific application hoping for resolution (i)!
12386 </p>
12387
12388 <p><i>[
12389 Batavia (2009-05):
12390 ]</i></p>
12391
12392 <blockquote>
12393 <p>
12394 We dislike options i and iii, and option ii seems like overkill.
12395 If we remove it (option iv), implementers can still provide it under a
12396 different name.
12397 </p>
12398 <p>
12399 Move to Open pending wording (from Alisdair) for option iv.
12400 </p>
12401 </blockquote>
12402
12403 <p><i>[
12404 2009-05-23 Alisdair provided wording for option iv.
12405 ]</i></p>
12406
12407
12408 <p><i>[
12409 2009-07-20 Alisdair adds:
12410 ]</i></p>
12411
12412
12413 <blockquote>
12414 <p>
12415 I'm not sure why this issue was not discussed at Frankfurt (or I missed
12416 the discussion) but the rationale is now fundamentally flawed.  With the
12417 removal of concepts, <tt>std::identity</tt> again becomes an important library
12418 type so we cannot simply remove it.
12419 </p>
12420 <p>
12421 At that point, we need to pick one of the other suggested resolutions,
12422 but have no guidance at the moment.
12423 </p>
12424 </blockquote>
12425
12426 <p><i>[
12427 2009-07-20 Howard adds:
12428 ]</i></p>
12429
12430
12431 <blockquote>
12432 <p>
12433 I believe the rationale for not addressing this issue in Frankfurt was that it did
12434 not address a national body comment.
12435 </p>
12436 <p>
12437 I also believe that removal of <tt>identity</tt> is still a practical option as
12438 my latest reformulation of <tt>forward</tt>, which is due to comments suggested
12439 at Summit, no longer uses <tt>identity</tt>. :-)
12440 </p>
12441
12442 <blockquote><pre>template &lt;class T, class U,
12443     class = typename enable_if
12444             &lt;
12445                 !is_lvalue_reference&lt;T&gt;::value ||
12446                  is_lvalue_reference&lt;T&gt;::value &amp;&amp;
12447                  is_lvalue_reference&lt;U&gt;::value
12448             &gt;::type,
12449     class = typename enable_if
12450             &lt;
12451                 is_same&lt;typename remove_all&lt;T&gt;::type,
12452                         typename remove_all&lt;U&gt;::type&gt;::value
12453             &gt;::type&gt;
12454 inline
12455 T&amp;&amp;
12456 forward(U&amp;&amp; t)
12457 {
12458     return static_cast&lt;T&amp;&amp;&gt;(t);
12459
12460 }
12461 </pre>
12462
12463 <p><i>[
12464 The above code assumes acceptance of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1120">1120</a> for the definition of
12465 <tt>remove_all</tt>.  This is just to make the syntax a little more palatable.
12466 Without this trait the above is still very implementable.
12467 ]</i></p>
12468
12469
12470 </blockquote>
12471
12472 <p>
12473 Paper with rationale is on the way ... <i>really</i>, I promise this time! ;-)
12474 </p>
12475 </blockquote>
12476
12477 <p><i>[
12478 2009-07-30 Daniel adds:  See <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#823">823</a> for an alternative resolution.
12479 ]</i></p>
12480
12481
12482 <p><i>[
12483 2009-10 Santa Cruz:
12484 ]</i></p>
12485
12486
12487 <blockquote>
12488 Move to Ready. Howard will update proposed wording to reflect current draft.
12489 </blockquote>
12490
12491
12492
12493 <p><b>Proposed resolution:</b></p>
12494 <p>
12495 Strike from 20.3 [utility]:
12496 </p>
12497
12498 <blockquote><pre><del>template &lt;class T&gt; struct identity;</del>
12499 </pre></blockquote>
12500
12501 <p>
12502 Remove from 20.3.3 [forward]:
12503 </p>
12504
12505 <blockquote>
12506 <pre><del>template &lt;class T&gt; struct identity {
12507   typedef T type;
12508
12509   const T&amp; operator()(const T&amp; x) const;
12510 };</del>
12511
12512 <del>const T&amp; operator()(const T&amp; x) const;</del>
12513 </pre>
12514 <blockquote>
12515 <del>-2-  <i>Returns:</i> <tt>x</tt></del>
12516 </blockquote>
12517 </blockquote>
12518
12519
12520
12521
12522
12523
12524 <hr>
12525 <h3><a name="940"></a>940. <tt>std::distance</tt></h3>
12526 <p><b>Section:</b> 24.4.4 [iterator.operations] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
12527  <b>Submitter:</b> Thomas <b>Opened:</b> 2008-12-14  <b>Last modified:</b> 2009-10-22</p>
12528 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.operations">active issues</a> in [iterator.operations].</p>
12529 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.operations">issues</a> in [iterator.operations].</p>
12530 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
12531 <p><b>Discussion:</b></p>
12532
12533 <p><b>Addresses UK 270</b></p>
12534
12535 <p>
12536 Regarding the <tt>std::distance</tt> - function, 24.4.4 [iterator.operations]
12537 / 4 says:
12538 </p>
12539 <blockquote>
12540 Returns the
12541 number of increments or decrements needed to get from first to last.
12542 </blockquote>
12543 <p>
12544 This sentence is completely silent about the sign of the return value.
12545 24.4.4 [iterator.operations] / 1 gives more information about the
12546 underlying operations, but
12547 again no inferences about the sign can be made.
12548 Strictly speaking, that is taking that sentence literally, I think this
12549 sentence even implies a positive return value in all cases, as the
12550 number of increments or decrements is clearly a ratio scale variable,
12551 with a natural zero bound.
12552 </p>
12553 <p>
12554 Practically speaking, my implementations did what common sense and
12555 knowledge based on pointer arithmetic forecasts, namely a positive sign
12556 for increments (that is, going from <tt>first</tt> to <tt>last</tt> by <tt>operator++</tt>), and a
12557 negative sign for decrements (going from <tt>first</tt> to <tt>last</tt> by <tt>operator--</tt>).
12558 </p>
12559 <p>
12560 Here are my two questions:
12561 </p>
12562 <p>
12563 First, is that paragraph supposed to be interpreted in the way what I
12564 called 'common sense', that is negative sign for decrements ? I am
12565 fairly sure that's the supposed behavior, but a double-check here in
12566 this group can't hurt.
12567 </p>
12568 <p>
12569 Second, is the present wording (2003 standard version - no idea about
12570 the draft for the upcoming standard) worth an edit to make it a bit more
12571 sensible, to mention the sign of the return value explicitly ?
12572 </p>
12573
12574 <p><i>[
12575 Daniel adds:
12576 ]</i></p>
12577
12578
12579 <blockquote>
12580 <p>
12581 My first thought was that resolution <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#204">204</a> would already cover the
12582 issue report, but it seems that current normative wording is in
12583 contradiction to that resolution:
12584 </p>
12585
12586 <p>
12587 Referring to
12588 <a href="" ref="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2798.pdf">N2798</a>,
12589 24.4.4 [iterator.operations]/ p.4 says:
12590 </p>
12591
12592 <blockquote>
12593 <i>Effects:</i> Returns the number of increments or decrements needed to get
12594 from <tt>first</tt> to <tt>last</tt>.
12595 </blockquote>
12596
12597 <p>
12598 IMO the part " or decrements" is in contradiction to p. 5 which says
12599 </p>
12600
12601 <blockquote>
12602 <i>Requires:</i> <tt>last</tt> shall be reachable from <tt>first</tt>.
12603 </blockquote>
12604
12605 <p>
12606 because "reachable" is defined in X [iterator.concepts]/7 as
12607 </p>
12608
12609 <blockquote>
12610 An iterator <tt>j</tt> is called reachable from an iterator <tt>i</tt> if and only if
12611 there is a finite
12612 sequence of applications of the expression <tt>++i</tt> that makes <tt>i == j</tt>.[..]
12613 </blockquote>
12614
12615 <p>
12616 Here is wording that would be consistent with this definition of "reachable":
12617 </p>
12618
12619 <p>
12620 Change 24.4.4 [iterator.operations] p4 as follows:
12621 </p>
12622
12623 <blockquote>
12624 <i>Effects:</i> Returns the number of increments <del>or decrements</del>
12625 needed to get from <tt>first</tt> to <tt>last</tt>.
12626 </blockquote>
12627
12628 </blockquote>
12629
12630 <p>
12631 Thomas adds more discussion and an alternative view point
12632 <a href="http://groups.google.com/group/comp.std.c++/browse_thread/thread/e8e46dcda0a5d797#">here</a>.
12633 </p>
12634
12635 <p><i>[
12636 Summit:
12637 ]</i></p>
12638
12639
12640 <blockquote>
12641 The proposed wording below was verbally agreed to.  Howard provided.
12642 </blockquote>
12643
12644 <p><i>[
12645 Batavia (2009-05):
12646 ]</i></p>
12647
12648 <blockquote>
12649 <p>
12650 Pete reports that a recent similar change has been made
12651 for the <tt>advance()</tt> function.
12652 </p>
12653 <p>
12654 We agree with the proposed resolution.
12655 Move to Tentatively Ready.
12656 </p>
12657 </blockquote>
12658
12659 <p><i>[
12660 2009-07 Frankfurt
12661 ]</i></p>
12662
12663
12664 <blockquote>
12665 Moved from Tentatively Ready to Open only because the wording needs to be
12666 tweaked for concepts removal.
12667 </blockquote>
12668
12669 <p><i>[
12670 2009-07 Frankfurt:
12671 ]</i></p>
12672
12673
12674 <blockquote>
12675 Leave Open pending arrival of a post-Concepts WD.
12676 </blockquote>
12677
12678 <p><i>[
12679 2009-10-14 Daniel provided de-conceptified wording.
12680 ]</i></p>
12681
12682
12683 <p><i>[
12684 2009-10 Santa Cruz:
12685 ]</i></p>
12686
12687
12688 <blockquote>
12689 Move to Ready, replacing the Effects clause in the proposed wording with
12690 "If InputIterator meets the requirements of random access iterator then
12691 returns (last - first), otherwise returns the number of increments
12692 needed to get from first to list.".
12693 </blockquote>
12694
12695
12696
12697 <p><b>Proposed resolution:</b></p>
12698 <ol>
12699 <li>
12700 <p>
12701 Change 24.2.5 [random.access.iterators], Table 105 as indicated [This change is not
12702 essential but it simplifies the specification] for the row with
12703 expression "<tt>b - a</tt>"
12704 and the column Operational semantics:
12705 </p>
12706
12707 <blockquote><pre><del>(a &lt; b) ? </del>distance(a,b)
12708 <del>: -distance(b,a)</del>
12709 </pre></blockquote>
12710 </li>
12711
12712 <li>
12713 <p>
12714 Change 24.4.4 [iterator.operations]/4+5 as indicated:
12715 </p>
12716
12717 <blockquote><pre>template&lt;class InputIterator&gt;
12718   typename iterator_traits&lt;InputIterator&gt;::difference_type
12719     distance(InputIterator first, InputIterator last);
12720 </pre>
12721 <blockquote>
12722 <p>
12723 4 <i>Effects:</i> <ins>If <tt>InputIterator</tt> meets the requirements
12724 of random access iterator then returns <tt>(last - first)</tt>,
12725 otherwise</ins> <del>R</del><ins>r</ins>eturns the number of increments
12726 <del>or decrements</del> needed to get from <tt>first</tt> to
12727 <tt>last</tt>.
12728 </p>
12729
12730 <p>
12731 5 <i>Requires:</i> <ins>If <tt>InputIterator</tt> meets the requirements
12732 of random access iterator then <tt>last</tt> shall be reachable from
12733 <tt>first</tt> or <tt>first</tt> shall be reachable from <tt>last</tt>,
12734 otherwise</ins> <tt>last</tt> shall be reachable from <tt>first</tt>.
12735 </p>
12736 </blockquote>
12737 </blockquote>
12738 </li>
12739 </ol>
12740
12741
12742
12743
12744
12745
12746
12747
12748
12749 <hr>
12750 <h3><a name="950"></a>950. unique_ptr converting ctor shouldn't accept array form</h3>
12751 <p><b>Section:</b> 20.8.14.2.1 [unique.ptr.single.ctor] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
12752  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-21</p>
12753 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unique.ptr.single.ctor">active issues</a> in [unique.ptr.single.ctor].</p>
12754 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unique.ptr.single.ctor">issues</a> in [unique.ptr.single.ctor].</p>
12755 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
12756 <p><b>Discussion:</b></p>
12757 <p>
12758 <tt>unique_ptr</tt>'s of array type should not convert to
12759 <tt>unique_ptr</tt>'s which do not have an array type.
12760 </p>
12761
12762 <blockquote><pre>struct Deleter
12763 {
12764    void operator()(void*) {}
12765 };
12766
12767 int main()
12768 {
12769    unique_ptr&lt;int[], Deleter&gt; s;
12770    unique_ptr&lt;int, Deleter&gt; s2(std::move(s));  <font color="#c80000">// should not compile</font>
12771 }
12772 </pre></blockquote>
12773
12774 <p><i>[
12775 Post Summit:
12776 ]</i></p>
12777
12778
12779 <blockquote>
12780 <p>
12781 Walter: Does the "diagnostic required" apply to both arms of the "and"?
12782 </p>
12783 <p>
12784 Tom Plum: suggest to break into several sentences
12785 </p>
12786 <p>
12787 Walter: suggest "comma" before the "and" in both places
12788 </p>
12789 <p>
12790 Recommend Review.
12791 </p>
12792 </blockquote>
12793
12794 <p><i>[
12795 Batavia (2009-05):
12796 ]</i></p>
12797
12798 <blockquote>
12799 The post-Summit comments have been applied to the proposed resolution.
12800 We now agree with the proposed resolution.
12801 Move to Tentatively Ready.
12802 </blockquote>
12803
12804 <p><i>[
12805 2009-07 Frankfurt
12806 ]</i></p>
12807
12808
12809 <blockquote>
12810 Moved from Tentatively Ready to Open only because the wording needs to be
12811 improved for enable_if type constraining, possibly following Robert's
12812 formula.
12813 </blockquote>
12814
12815 <p><i>[
12816 2009-08-01 Howard updates wording and sets to Review.
12817 ]</i></p>
12818
12819
12820 <p><i>[
12821 2009-10 Santa Cruz:
12822 ]</i></p>
12823
12824
12825 <blockquote>
12826 Move to Ready.
12827 </blockquote>
12828
12829
12830
12831 <p><b>Proposed resolution:</b></p>
12832 <p>
12833 Change 20.8.14.2.1 [unique.ptr.single.ctor]:
12834 </p>
12835
12836 <blockquote>
12837 <pre>template &lt;class U, class E&gt; unique_ptr(unique_ptr&lt;U, E&gt;&amp;&amp; u);
12838 </pre>
12839 <blockquote>
12840 <p>
12841 -20- <i>Requires:</i> If <tt>D</tt> is not a reference type,
12842 construction of the deleter <tt>D</tt> from an rvalue of type <tt>E</tt>
12843 shall be well formed and shall not throw an exception. <del>If <tt>D</tt> is
12844 a reference type, then <tt>E</tt> shall be the same type as <tt>D</tt>
12845 (diagnostic required). <tt>unique_ptr&lt;U, E&gt;::pointer</tt> shall be
12846 implicitly convertible to <tt>pointer</tt>. [<i>Note:</i> These requirements
12847 imply that <tt>T</tt> and <tt>U</tt> are complete types. &#8212; <i>end note</i>]</del>
12848 </p>
12849
12850 <p><ins>
12851 <i>Remarks:</i> If <tt>D</tt> is
12852 a reference type, then <tt>E</tt> shall be the same type as <tt>D</tt>, else this
12853 constructor shall not participate in overload resolution. <tt>unique_ptr&lt;U, E&gt;::pointer</tt> shall be
12854 implicitly convertible to <tt>pointer</tt>, else this
12855 constructor shall not participate in overload resolution. <tt>U</tt> shall not be
12856 an array type, else this
12857 constructor shall not participate in overload resolution. [<i>Note:</i> These requirements
12858 imply that <tt>T</tt> and <tt>U</tt> are complete types. &#8212; <i>end note</i>]
12859 </ins></p>
12860
12861 </blockquote>
12862 </blockquote>
12863
12864 <p>
12865 Change 20.8.14.2.3 [unique.ptr.single.asgn]:
12866 </p>
12867
12868 <blockquote>
12869 <pre>template &lt;class U, class E&gt; unique_ptr&amp; operator=(unique_ptr&lt;U, E&gt;&amp;&amp; u);
12870 </pre>
12871 <blockquote>
12872 <p>
12873 -6- <i>Requires:</i> Assignment of the deleter <tt>D</tt> from an rvalue
12874 <tt>D</tt> shall not throw an exception. <del><tt>unique_ptr&lt;U,
12875 E&gt;::pointer</tt> shall be implicitly convertible to <tt>pointer</tt>.
12876 [<i>Note:</i> These requirements imply that <tt>T</tt> and <tt>U</tt>
12877 are complete types. &#8212; <i>end note</i>]</del>
12878 </p>
12879
12880 <p><ins>
12881 <i>Remarks:</i> <tt>unique_ptr&lt;U,
12882 E&gt;::pointer</tt> shall be implicitly convertible to <tt>pointer</tt>, else this
12883 operator shall not participate in overload resolution.
12884 <tt>U</tt> shall not be an array type, else this
12885 operator shall not participate in overload resolution.
12886 [<i>Note:</i> These requirements imply that <tt>T</tt> and <tt>U</tt>
12887 are complete types. &#8212; <i>end note</i>]
12888 </ins></p>
12889
12890 </blockquote>
12891 </blockquote>
12892
12893
12894
12895
12896
12897
12898 <hr>
12899 <h3><a name="951"></a>951. Various threading bugs #1</h3>
12900 <p><b>Section:</b> 20.9.2.1 [time.traits.is_fp] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
12901  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-23</p>
12902 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
12903 <p><b>Discussion:</b></p>
12904
12905 <p>
12906 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#953">953</a>.
12907 </p>
12908
12909 <p>
12910 20.9.2.1 [time.traits.is_fp] says that the type <tt>Rep</tt> "is
12911 assumed to be ... a class emulating an integral type." What are the
12912 requirements for such a type?
12913 </p>
12914 <p><i>[
12915 2009-05-10 Howard adds:
12916 ]</i></p>
12917
12918
12919 <blockquote>
12920 <tt>IntegralLike</tt>.
12921 </blockquote>
12922
12923 <p><i>[
12924 Batavia (2009-05):
12925 ]</i></p>
12926
12927 <blockquote>
12928 <p>
12929 As with issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#953">953</a>,
12930 we recommend this issue be addressed in the context of providing concepts for the entire <tt>thread</tt> header.
12931 </p>
12932 <p>
12933 We look forward to proposed wording.
12934 </p>
12935 <p>
12936 Move to Open.
12937 </p>
12938 </blockquote>
12939
12940 <p><i>[
12941 2009-08-01 Howard adds:
12942 ]</i></p>
12943
12944
12945 <blockquote>
12946 <p>
12947 I have surveyed all clauses of 20.9.2.2 [time.traits.duration_values],
12948 20.9.2.3 [time.traits.specializations] and 20.9.3 [time.duration].
12949 I can not find any clause which involves the use of a <tt>duration::rep</tt> type
12950 where the requirements on the <tt>rep</tt> type are not clearly spelled out.
12951 These requirements were carefully crafted to allow any arithmetic type, or
12952 any user-defined type emulating an arithmetic type.
12953 </p>
12954
12955 <p>
12956 Indeed, <tt>treat_as_floating_point</tt>
12957 becomes completely superfluous if <tt>duration::rep</tt> can never be a class type.
12958 </p>
12959
12960 <p>
12961 There will be some <tt>Rep</tt> types which will not meet the requirements of
12962 <em>every</em> <tt>duration</tt> operation.  This is no different than the fact
12963 that <tt>vector&lt;T&gt;</tt> can easily be used for types <tt>T</tt> which are
12964 not <tt>DefaultConstructible</tt>, even though some members of <tt>vector&lt;T&gt;</tt>
12965 require <tt>T</tt> to be <tt>DefaultConstructible</tt>.  This is why the requirements
12966 on <tt>Rep</tt> are specified for each operation individually.
12967 </p>
12968
12969 <p>
12970 In 20.9.2.1 [time.traits.is_fp] p1:
12971 </p>
12972
12973 <blockquote><pre>template &lt;class Rep&gt; struct treat_as_floating_point
12974   : is_floating_point&lt;Rep&gt; { };
12975 </pre>
12976
12977 <blockquote>
12978 The <tt>duration</tt> template uses the <tt>treat_as_floating_point</tt> trait to help
12979 determine if a <tt>duration</tt> object can be converted to another <tt>duration</tt>
12980 with a different tick period. If <tt>treat_as_floating_point&lt;Rep&gt;::value</tt> is
12981 <tt>true</tt>, then <tt>Rep</tt> is a floating-point type and implicit conversions are
12982 allowed among <tt>duration</tt>s. Otherwise, the implicit convertibility depends
12983 on the tick periods of the <tt>duration</tt>s. If <tt>Rep</tt> is <u>a class type which
12984 emulates a floating-point type</u>, the author of <tt>Rep</tt> can specialize
12985 <tt>treat_as_floating_point</tt> so that <tt>duration</tt> will treat this <tt>Rep</tt> as if it
12986 were a floating-point type. Otherwise <tt>Rep</tt> is assumed to be an integral
12987 type or <u>a class emulating an integral type</u>.
12988 </blockquote>
12989 </blockquote>
12990
12991 <p>
12992 The phrases "a class type which emulates a floating-point type" and
12993 "a class emulating an integral type" are clarifying phrases which refer to
12994 the summation of all the requirements on the <tt>Rep</tt> type specified in
12995 detail elsewhere (and <em>should not</em> be repeated here).
12996 </p>
12997
12998 <p>
12999 This specification has been implemented, now multiple times, and the experience
13000 has been favorable.  The current specification clearly specifies the requirements
13001 at each point of use (though I'd be happy to fix any place I may have missed,
13002 but none has been pointed out).
13003 </p>
13004
13005 <p>
13006 I am amenable to improved wording of this paragraph (and any others), but do not have any
13007 suggestions for improved wording at this time.  I am <em>strongly</em> opposed to
13008 changes which would significantly alter the semantics of the
13009 specification under 20.9 [time] without firmly grounded and
13010 documented rationale, example implementation, testing, and user
13011 experience which relates a positive experience.
13012 </p>
13013
13014 <p>
13015 I recommend NAD unless someone wants to produce some clarifying wording.
13016 </p>
13017 </blockquote>
13018
13019 <p><i>[
13020 2009-10 Santa Cruz:
13021 ]</i></p>
13022
13023
13024 <blockquote>
13025 Stefanus to provide wording to turn this into a note.
13026 </blockquote>
13027
13028
13029
13030 <p><b>Proposed resolution:</b></p>
13031 <p>
13032 </p>
13033
13034
13035
13036
13037
13038 <hr>
13039 <h3><a name="953"></a>953. Various threading bugs #3</h3>
13040 <p><b>Section:</b> 20.9.1 [time.clock.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
13041  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-23</p>
13042 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#time.clock.req">active issues</a> in [time.clock.req].</p>
13043 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.clock.req">issues</a> in [time.clock.req].</p>
13044 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
13045 <p><b>Discussion:</b></p>
13046
13047 <p>
13048 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#951">951</a>.
13049 </p>
13050
13051 <p>
13052 20.9.1 [time.clock.req] says that a clock's <tt>rep</tt> member is "an
13053 arithmetic type or a class emulating an arithmetic type." What are the
13054 requirements for such a type?
13055 </p>
13056
13057 <p><i>[
13058 2009-05-10 Howard adds:
13059 ]</i></p>
13060
13061
13062 <blockquote>
13063 This wording was aimed directly at the <tt>ArithmeticLike</tt> concept.
13064 </blockquote>
13065
13066 <p><i>[
13067 Batavia (2009-05):
13068 ]</i></p>
13069
13070 <blockquote>
13071 <p>
13072 We recommend this issue be addressed in the context of providing concepts
13073 for the entire <tt>thread</tt> header.
13074 </p>
13075 <p>
13076 May resolve for now by specifying arithmetic types,
13077 and in future change to <tt>ArithmeticLike</tt>.
13078 However, Alisdair believes this is not feasible.
13079 </p>
13080 <p>
13081 Bill disagrees.
13082 </p>
13083 <p>
13084 We look forward to proposed wording.  Move to Open.
13085 </p>
13086 </blockquote>
13087
13088 <p><i>[
13089 2009-08-01 Howard adds:
13090 ]</i></p>
13091
13092
13093 <blockquote>
13094 See commented dated 2009-08-01 in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#951">951</a>.
13095 </blockquote>
13096
13097 <p><i>[
13098 2009-10 Santa Cruz:
13099 ]</i></p>
13100
13101
13102 <blockquote>
13103 Stefanus to provide wording to turn this into a note.
13104 </blockquote>
13105
13106
13107
13108 <p><b>Proposed resolution:</b></p>
13109 <p>
13110 </p>
13111
13112
13113
13114
13115
13116 <hr>
13117 <h3><a name="954"></a>954. Various threading bugs #4</h3>
13118 <p><b>Section:</b> 20.9.1 [time.clock.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
13119  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-21</p>
13120 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#time.clock.req">active issues</a> in [time.clock.req].</p>
13121 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.clock.req">issues</a> in [time.clock.req].</p>
13122 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
13123 <p><b>Discussion:</b></p>
13124 <p>
13125 Table 55 -- Clock Requirements (in 20.9.1 [time.clock.req])
13126 </p>
13127
13128 <ol type="a">
13129 <li>
13130 the requirements for <tt>C1::time_point</tt> require <tt>C1</tt> and <tt>C2</tt>
13131 to "refer to the same epoch", but "epoch" is not defined.
13132 </li>
13133 <li>
13134 "Different clocks may share a <tt>time_point</tt> definition if it is
13135 valid to compare their <tt>time_point</tt>s by comparing their
13136 respective <tt>duration</tt>s." What does "valid" mean here? And, since
13137 <tt>C1::rep</tt> is "**THE** representation type of the native
13138 <tt>duration</tt> and <tt>time_point</tt>" (emphasis added), there
13139 doesn't seem to be much room for some other representation.
13140 </li>
13141 <li>
13142 <tt>C1::is_monotonic</tt> has type "<tt>const bool</tt>". The
13143 "<tt>const</tt>" should be removed.
13144 </li>
13145 <li>
13146 <tt>C1::period</tt> has type <tt>ratio</tt>. <tt>ratio</tt> isn't a type,
13147 it's a template. What is the required type?
13148 </li>
13149 </ol>
13150
13151 <p><i>[
13152 2009-05-10 Howard adds:
13153 ]</i></p>
13154
13155
13156 <ol type="a">
13157 <li>
13158 <p>
13159 "epoch" is purposefully not defined beyond the common English
13160 <a href="http://www.dictionary.net/epoch">definition</a>.  The C standard
13161 also chose not to define epoch, though POSIX did.  I believe it is a strength
13162 of the C standard that epoch is not defined.  When it is known that two <tt>time_point</tt>s
13163 refer to the same epoch, then a definition of the epoch is not needed to compare
13164 the two <tt>time_point</tt>s, or subtract them.
13165 </p>
13166 <p>
13167 A <tt>time_point</tt> and a <tt>Clock</tt> implicitly refer to an (unspecified) epoch.
13168 The <tt>time_point</tt> represents an offset (<tt>duration</tt>) from an epoch.
13169 </p>
13170 </li>
13171 <li>
13172 <p>
13173 The sentence:
13174 </p>
13175 <blockquote>
13176 Different clocks
13177 may share a <tt>time_point</tt>
13178 definition if it is valid to
13179 compare their <tt>time_point</tt>s by
13180 comparing their respective
13181 <tt>duration</tt>s.
13182 </blockquote>
13183
13184 <p>
13185 is redundant and could be removed.  I believe the sentence which follows the above:
13186 </p>
13187
13188 <blockquote>
13189 <tt>C1</tt> and <tt>C2</tt> shall refer to the same epoch.
13190 </blockquote>
13191
13192 <p>
13193 is sufficient.  If two clocks share the same epoch, then by definition, comparing
13194 their <tt>time_point</tt>s is valid.
13195 </p>
13196 </li>
13197 <li>
13198 <tt>is_monotonic</tt> is meant to never change (be <tt>const</tt>).  It is also
13199 desired that this value be usable in compile-time computation and branching.
13200 </li>
13201 <li>
13202 <p>
13203 This should probably instead be worded:
13204 </p>
13205 <blockquote>
13206 An instantiation of <tt>ratio</tt>.
13207 </blockquote>
13208 </li>
13209 </ol>
13210
13211 <p><i>[
13212 Batavia (2009-05):
13213 ]</i></p>
13214
13215 <blockquote>
13216 <p>
13217 Re (a): It is not clear to us whether "epoch" is a term of art.
13218 </p>
13219 <p>
13220 Re (b), (c), and (d):  We agree with Howard's comments,
13221 and would consider adding to (c) a <tt>static constexpr</tt> requirement.
13222 </p>
13223 <p>
13224 Move to Open pending proposed wording.
13225 </p>
13226 </blockquote>
13227
13228 <p><i>[
13229 2009-05-25 Daniel adds:
13230 ]</i></p>
13231
13232
13233 <blockquote>
13234 In regards to (d) I suggest to say "a specialization of ratio" instead of
13235 "An instantiation of ratio". This seems to be the better matching standard
13236 core language term for this kind of entity.
13237 </blockquote>
13238
13239 <p><i>[
13240 2009-05-25 Ganesh adds:
13241 ]</i></p>
13242
13243
13244 <blockquote>
13245 <p>
13246 Regarding (a), I found this paper on the ISO website using the term "epoch" consistently with the current wording:
13247 </p>
13248
13249 <p>
13250 <a href="http://standards.iso.org/ittf/PubliclyAvailableStandards/C030811e_FILES/MAIN_C030811e/text/ISOIEC_18026E_TEMPORAL_CS.HTM">http://standards.iso.org/ittf/PubliclyAvailableStandards/C030811e_FILES/MAIN_C030811e/text/ISOIEC_18026E_TEMPORAL_CS.HTM</a>
13251 </p>
13252 <p>
13253 which is part of ISO/IEC 18026 "Information technology -- Spatial Reference Model (SRM)".
13254 </p>
13255 </blockquote>
13256
13257 <p><i>[
13258 2009-08-01 Howard: Moved to Reivew as the wording requested in Batavia has been provided.
13259 ]</i></p>
13260
13261
13262 <p><i>[
13263 2009-10 Santa Cruz:
13264 ]</i></p>
13265
13266
13267 <blockquote>
13268 Move to Ready.
13269 </blockquote>
13270
13271
13272
13273 <p><b>Proposed resolution:</b></p>
13274 <ol type="a">
13275 <li>
13276 <p>
13277 Change 20.9.1 [time.clock.req] p1:
13278 </p>
13279 <blockquote>
13280 -1- A clock is a bundle consisting of a native <tt>duration</tt>, a native <tt>time_point</tt>, and a function <tt>now()</tt> to get the
13281 current <tt>time_point</tt>.  <ins>The origin of the clock's <tt>time_point</tt> is referred to as the clock's <i>epoch</i> as defined in
13282 section 6.3 of ISO/IEC 18026.</ins>
13283 A clock shall meet the requirements in Table 45.
13284 </blockquote>
13285 </li>
13286 <li>
13287 <p>
13288 Remove the sentence from the <tt>time_point</tt> row of the table "Clock Requirements":
13289 </p>
13290 <table border="1">
13291 <caption>Clock requirements</caption>
13292 <tbody><tr>
13293 <td>
13294 <tt>C1::time_point</tt>
13295 </td>
13296 <td>
13297 <tt>chrono::time_point&lt;C1&gt;</tt> or <tt>chrono::time_point&lt;C2, C1::duration&gt;</tt>
13298 </td>
13299 <td>
13300 The native <tt>time_point</tt> type of the clock.
13301 <del>Different clocks may share a <tt>time_point</tt> definition if it is valid to compare their <tt>time_point</tt>s by comparing their respective <tt>duration</tt>s.</del>
13302 <tt>C1</tt> and <tt>C2</tt> shall refer to the same epoch.
13303 </td>
13304 </tr>
13305 </tbody></table>
13306 </li>
13307 </ol>
13308 <ol start="4" type="a">
13309 <li>
13310 <p>
13311 Change the row starting with <tt>C1::period</tt> of the table "Clock Requirements":
13312 </p>
13313 <table border="1">
13314 <caption>Clock requirements</caption>
13315 <tbody><tr>
13316 <td>
13317 <tt>C1::period</tt>
13318 </td>
13319 <td>
13320 <ins>a specialization of</ins> <tt>ratio</tt>
13321 </td>
13322 <td>
13323 The tick period of the clock in seconds.
13324 </td>
13325 </tr>
13326 </tbody></table>
13327
13328 </li>
13329 </ol>
13330
13331
13332
13333
13334
13335 <hr>
13336 <h3><a name="956"></a>956. Various threading bugs #6</h3>
13337 <p><b>Section:</b> 20.9.1 [time.clock.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
13338  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-24</p>
13339 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#time.clock.req">active issues</a> in [time.clock.req].</p>
13340 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.clock.req">issues</a> in [time.clock.req].</p>
13341 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
13342 <p><b>Discussion:</b></p>
13343 <p>
13344 20.9.1 [time.clock.req] uses the word "native" in several places,
13345 but doesn't define it. What is a "native <tt>duration</tt>"?
13346 </p>
13347
13348 <p><i>[
13349 2009-05-10 Howard adds:
13350 ]</i></p>
13351
13352
13353 <blockquote>
13354 The standard uses "native" in several places without defining it (e.g.
13355 2.14.3 [lex.ccon]).  It is meant to mean "that which is defined
13356 by the facility", or something along those lines.  In this case it refers
13357 to the nested <tt>time_point</tt> and <tt>duration</tt> types of the clock.
13358 Better wording is welcome.
13359 </blockquote>
13360
13361 <p><i>[
13362 Batavia (2009-05):
13363 ]</i></p>
13364
13365 <blockquote>
13366 Move to Open pending proposed wording from Pete.
13367 </blockquote>
13368
13369 <p><i>[
13370 2009-10-23 Pete provides wording:
13371 ]</i></p>
13372
13373
13374
13375 <p><b>Proposed resolution:</b></p>
13376 <p>
13377 Remove every occurrence of "native" in 20.9.1 [time.clock.req].
13378 </p>
13379
13380 <p>
13381 Add the following sentence at the end of 20.9.1 [time.clock.req]/1:
13382 </p>
13383
13384 <blockquote>
13385 A clock is a bundle consisting of a <del>native</del> <tt>duration</tt>, a <del>native</del>
13386 <tt>time_point</tt>, and a function <tt>now()</tt> to get the current <tt>time_point</tt>. A clock
13387 shall meet the requirements in Table 55.
13388 <ins>The <tt>duration</tt> and <tt>time_point</tt> types have the natural size and resolution
13389 suggested by the architecture of the execution environment.</ins>
13390 </blockquote>
13391
13392
13393
13394
13395
13396 <hr>
13397 <h3><a name="957"></a>957. Various threading bugs #7</h3>
13398 <p><b>Section:</b> 20.9.5.1 [time.clock.system] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
13399  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-21</p>
13400 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.clock.system">issues</a> in [time.clock.system].</p>
13401 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
13402 <p><b>Discussion:</b></p>
13403 <p>
13404 20.9.5.1 [time.clock.system]: <tt>to_time_t</tt> is overspecified. It
13405 requires truncation, but should allow rounding. For example, suppose a
13406 system has a clock that gives times in milliseconds, but <tt>time()</tt> rounds
13407 those times to the nearest second. Then <tt>system_clock</tt> can't use any
13408 resolution finer than one second, because if it did, truncating times
13409 between half a second and a full second would produce the wrong <tt>time_t</tt>
13410 value.
13411 </p>
13412
13413 <p><i>[
13414 Post Summit Anthony Williams provided proposed wording.
13415 ]</i></p>
13416
13417
13418 <p><i>[
13419 Batavia (2009-05):
13420 ]</i></p>
13421
13422 <blockquote>
13423 Move to Review pending input from Howard. and other stakeholders.
13424 </blockquote>
13425
13426 <p><i>[
13427 2009-05-23 Howard adds:
13428 ]</i></p>
13429
13430
13431 <blockquote>
13432 I am in favor of the wording provided by Anthony.
13433 </blockquote>
13434
13435 <p><i>[
13436 2009-10 Santa Cruz:
13437 ]</i></p>
13438
13439
13440 <blockquote>
13441 Move to Ready.
13442 </blockquote>
13443
13444
13445
13446 <p><b>Proposed resolution:</b></p>
13447 <p>
13448 In 20.9.5.1 [time.clock.system] replace paragraphs 3 and 4 with:
13449 </p>
13450
13451 <blockquote>
13452 <pre>time_t to_time_t(const time_point&amp; t);
13453 </pre>
13454 <blockquote>
13455 -3- <i>Returns:</i> A <tt>time_t</tt> object that represents the same
13456 point in time as <tt>t</tt> when both values are <del>truncated</del>
13457 <ins>restricted</ins> to the coarser of the precisions of
13458 <tt>time_t</tt> and <tt>time_point</tt>. <ins> It is implementation
13459 defined whether values are rounded or truncated to the required
13460 precision.</ins>
13461 </blockquote>
13462
13463 <pre>time_point from_time_t(time_t t);
13464 </pre>
13465 <blockquote>
13466 -4- <i>Returns:</i> A <tt>time_point</tt> object that represents the
13467 same point in time as <tt>t</tt> when both values are <del>truncated</del>
13468 <ins>restricted</ins> to the
13469 coarser of the precisions of <tt>time_t</tt> and <tt>time_point</tt>.
13470 <ins>It is implementation defined whether values are
13471 rounded or truncated to the required precision.</ins>
13472 </blockquote>
13473 </blockquote>
13474
13475
13476
13477
13478
13479 <hr>
13480 <h3><a name="959"></a>959. Various threading bugs #9</h3>
13481 <p><b>Section:</b> 30.5.1 [thread.condition.condvar] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
13482  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-26</p>
13483 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvar">active issues</a> in [thread.condition.condvar].</p>
13484 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvar">issues</a> in [thread.condition.condvar].</p>
13485 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
13486 <p><b>Discussion:</b></p>
13487 <p>
13488 30.5.1 [thread.condition.condvar]: <tt>condition_variable::wait_for</tt>
13489 is required to compute the absolute time by adding the duration value to
13490 <tt>chrono::monotonic_clock::now()</tt>, but <tt>monotonic_clock</tt> is not required to
13491 exist.
13492 </p>
13493
13494 <p><i>[
13495 Summit:
13496 ]</i></p>
13497
13498
13499 <blockquote>
13500 Move to open. Associate with LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a> and any other monotonic-clock
13501 related issues.
13502 </blockquote>
13503
13504 <p><i>[
13505 2009-08-01 Howard adds:
13506 ]</i></p>
13507
13508
13509 <blockquote>
13510 I believe that <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a> (currently Ready) addresses this issue, and
13511 that this issue should be marked NAD, solved by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a> (assuming
13512 it moves to WP).
13513 </blockquote>
13514
13515 <p><i>[
13516 2009-10 Santa Cruz:
13517 ]</i></p>
13518
13519
13520 <blockquote>
13521 Leave open, but expect to be fixed by N2969 revision that Detlef is writing.
13522 </blockquote>
13523
13524
13525
13526 <p><b>Proposed resolution:</b></p>
13527 <p>
13528 </p>
13529
13530
13531
13532
13533
13534 <hr>
13535 <h3><a name="960"></a>960. Various threading bugs #10</h3>
13536 <p><b>Section:</b> 30.4.1 [thread.mutex.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
13537  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-26</p>
13538 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.mutex.requirements">active issues</a> in [thread.mutex.requirements].</p>
13539 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.mutex.requirements">issues</a> in [thread.mutex.requirements].</p>
13540 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
13541 <p><b>Discussion:</b></p>
13542 <p>
13543 30.4.1 [thread.mutex.requirements]: paragraph 4 is entitled
13544 "Error conditions", but according to 17.5.1.4 [structure.specifications], "Error
13545 conditions:" specifies "the error conditions for error codes reported by
13546 the function." It's not clear what this should mean when there is no
13547 function in sight.
13548 </p>
13549
13550 <p><i>[
13551 Summit:
13552 ]</i></p>
13553
13554
13555 <blockquote>
13556 Move to open.
13557 </blockquote>
13558
13559 <p><i>[
13560 Beman provided proposed wording.
13561 ]</i></p>
13562
13563
13564 <p><i>[
13565 2009-10 Santa Cruz:
13566 ]</i></p>
13567
13568
13569 <blockquote>
13570 Move to Ready. Fix the proposed wording with "functions of type Mutex"
13571 -&gt; "functions of Mutex type"
13572 </blockquote>
13573
13574
13575
13576 <p><b>Proposed resolution:</b></p>
13577 <p>
13578 Change 30.4.1 [thread.mutex.requirements] Mutex requirements,
13579 paragraph 4 as indicated:
13580 </p>
13581
13582 <blockquote>
13583 <p>
13584 -4- <del><i>Error conditions:</i></del>
13585 <ins>The error conditions for error codes, if any, reported by member
13586 functions of Mutex type shall be:</ins>
13587 </p>
13588 <ul>
13589 <li>
13590 <tt>not_enough_memory</tt> -- if there is not enough memory to construct
13591 the mutex object.
13592 </li>
13593 <li>
13594 <tt>resource_unavailable_try_again</tt> -- if any native handle type
13595 manipulated is not available.
13596 </li>
13597 <li>
13598 <tt>operation_not_permitted</tt> -- if the thread does not have the
13599 necessary permission to change the state of the mutex object.
13600 </li>
13601 <li>
13602 <tt>device_or_resource_busy</tt> -- if any native handle type
13603 manipulated is already locked.
13604 </li>
13605 <li>
13606 <tt>invalid_argument</tt> -- if any native handle type manipulated as
13607 part of mutex construction is incorrect.
13608 </li>
13609 </ul>
13610 </blockquote>
13611
13612
13613
13614
13615
13616 <hr>
13617 <h3><a name="962"></a>962. Various threading bugs #12</h3>
13618 <p><b>Section:</b> 30.4.3.2.2 [thread.lock.unique.locking] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
13619  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-22</p>
13620 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.lock.unique.locking">active issues</a> in [thread.lock.unique.locking].</p>
13621 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.lock.unique.locking">issues</a> in [thread.lock.unique.locking].</p>
13622 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
13623 <p><b>Discussion:</b></p>
13624 <p>
13625 30.4.3.2.2 [thread.lock.unique.locking]:  <tt>unique_lock::lock</tt> is
13626 required to throw an object of type <tt>std::system_error</tt> "when the
13627 postcondition cannot be achieved." The postcondition is <tt>owns == true</tt>,
13628 and this is trivial to achieve. Presumably, the requirement is intended
13629 to mean something more than that.
13630 </p>
13631
13632 <p><i>[
13633 Summit:
13634 ]</i></p>
13635
13636 <blockquote>
13637 Move to open.
13638 </blockquote>
13639
13640 <p><i>[
13641 Beman has volunteered to provide proposed wording.
13642 ]</i></p>
13643
13644
13645 <p><i>[
13646 2009-07-21 Beman added wording to address 30.2.2 [thread.req.exception]
13647 in response to the Frankfurt notes in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>.
13648 ]</i></p>
13649
13650
13651 <p><i>[
13652 2009-09-25 Beman: minor update to wording.
13653 ]</i></p>
13654
13655
13656 <p><i>[
13657 2009-10 Santa Cruz:
13658 ]</i></p>
13659
13660
13661 <blockquote>
13662 Move to Ready.
13663 </blockquote>
13664
13665
13666
13667 <p><b>Proposed resolution:</b></p>
13668
13669 <p><i>Change Exceptions 30.2.2 [thread.req.exception] as indicated:</i></p>
13670 <blockquote>
13671 <p>Some functions described in this Clause are specified to throw exceptions of
13672 type <code>system_error</code> (19.5.5). Such exceptions shall be thrown if <ins>
13673 any of the <i>Error conditions</i> are detected or</ins> a call to an operating
13674 system or other underlying API results in an error that prevents the library
13675 function from <del>satisfying its postconditions or from returning a meaningful
13676 value</del> <ins>meeting its specifications</ins>. <ins>Failure to
13677 allocate storage shall be reported as described in
13678 17.6.4.11 [res.on.exception.handling].</ins></p>
13679 </blockquote>
13680
13681 <p><i>Change thread assignment 30.3.1.5 [thread.thread.member], join(),
13682 paragraph 8 as indicated:</i></p>
13683 <blockquote>
13684 <p><i>Throws:</i> <code>std::system_error</code> when <del>the postconditions cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13685
13686 </blockquote>
13687
13688 <p><i>Change thread assignment 30.3.1.5 [thread.thread.member], detach(), paragraph
13689 13 as indicated:</i></p>
13690 <blockquote>
13691
13692 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13693 postconditions cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13694
13695 </blockquote>
13696
13697 <p><i>Change Mutex requirements 30.4.1 [thread.mutex.requirements], paragraph
13698 11, as indicated:</i></p>
13699 <blockquote>
13700
13701 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13702 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13703 </blockquote>
13704 <p><i>Change unique_lock locking 30.4.3.2.2 [thread.lock.unique.locking],
13705 paragraph 3, as indicated:</i></p>
13706 <blockquote>
13707
13708 <p><i>Throws:</i> <code>std::system_error</code> when <del>the  postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13709 </blockquote>
13710 <p><i>Change unique_lock locking 30.4.3.2.2 [thread.lock.unique.locking],
13711 paragraph 8, as indicated:</i></p>
13712 <blockquote>
13713
13714 <p><i>Throws:</i> <code>std::system_error</code> when <del>the  postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13715 </blockquote>
13716 <p><i>Change unique_lock locking 30.4.3.2.2 [thread.lock.unique.locking],
13717 paragraph 13, as indicated:</i></p>
13718 <blockquote>
13719
13720 <p><i>Throws:</i> <code>std::system_error</code> when <del>the  postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13721 </blockquote>
13722 <p><i>Change unique_lock locking 30.4.3.2.2 [thread.lock.unique.locking],
13723 paragraph 18, as indicated:</i></p>
13724 <blockquote>
13725
13726 <p><i>Throws:</i> <code>std::system_error</code> when <del>the  postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13727 </blockquote>
13728 <p><i>Change unique_lock locking 30.4.3.2.2 [thread.lock.unique.locking],
13729 paragraph 22, as indicated:</i></p>
13730 <blockquote>
13731
13732 <p><i>Throws:</i> <code>std::system_error</code> when <del>the  postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13733 </blockquote>
13734 <p><i>Change Function call_once 30.4.5.2 [thread.once.callonce], paragraph 4, as
13735 indicated</i></p>
13736 <blockquote>
13737   <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>,
13738   or any exception thrown by <code>func</code>.</p>
13739 </blockquote>
13740 <p><i>Change Class condition_variable 30.5.1 [thread.condition.condvar],
13741 paragraph 12, as indicated:</i></p>
13742 <blockquote>
13743
13744 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13745 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13746 </blockquote>
13747 <p><i>Change Class condition_variable 30.5.1 [thread.condition.condvar],
13748 paragraph 19, as indicated:</i></p>
13749 <blockquote>
13750
13751 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13752 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13753 </blockquote>
13754 <p><i>Change Class condition_variable_any 30.5.2 [thread.condition.condvarany],
13755 paragraph 10, as indicated:</i></p>
13756 <blockquote>
13757
13758 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13759 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13760 </blockquote>
13761 <p><i>Change Class condition_variable_any 30.5.2 [thread.condition.condvarany],
13762 paragraph 16, as indicated:</i></p>
13763 <blockquote>
13764
13765 <p><i>Throws:</i> <code>std::system_error</code> when <del>the returned value, effects, or
13766 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13767 </blockquote>
13768
13769 <p><i>Assuming issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, Monotonic Clock is Conditionally Supported?, has been
13770 applied to the working paper, change Change 30.5.1 [thread.condition.condvar] as
13771 indicated:</i></p>
13772 <blockquote>
13773 <pre>template &lt;class Rep, class Period&gt;
13774 bool wait_for(unique_lock&lt;mutex&gt;&amp; lock,
13775               const chrono::duration&lt;Rep, Period&gt;&amp; rel_time);</pre>
13776 <pre>...</pre>
13777
13778 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13779 postcondition cannot be achieved</del> <ins>an exception is required ([thread.req.exception])</ins>.</p>
13780 </blockquote>
13781
13782 <p><i>Assuming issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, Monotonic Clock is Conditionally Supported?, has been
13783 applied to the working paper, change Change 30.5.1 [thread.condition.condvar] as
13784 indicated:</i></p>
13785 <blockquote>
13786 <pre>template &lt;class Rep, class Period, class Predicate&gt;
13787   bool wait_for(unique_lock&lt;mutex&gt;&amp; lock,
13788                 const chrono::duration&lt;Rep, Period&gt;&amp; rel_time,
13789                 Predicate pred);</pre>
13790   <pre>...</pre>
13791
13792 <p><i>Throws:</i> <code>std::system_error</code> when <del>the effects or
13793 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13794 </blockquote>
13795
13796 <p><i>Assuming issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, Monotonic Clock is Conditionally Supported?, has been
13797 applied to the working paper, change 30.5.2 [thread.condition.condvarany] as
13798 indicated:</i></p>
13799 <blockquote>
13800 <pre>template &lt;class Lock, class Rep, class Period&gt;
13801   bool wait_for(Lock&amp; lock, const chrono::duration&lt;Rep, Period&gt;&amp; rel_time);</pre>
13802   <pre>...</pre>
13803
13804 <p><i>Throws:</i> <code>std::system_error</code> when <del>the returned value, effects or
13805 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13806 </blockquote>
13807
13808 <p><i>Assuming issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#859">859</a>, Monotonic Clock is Conditionally Supported?, has been
13809 applied to the working paper, change 30.5.2 [thread.condition.condvarany] as
13810 indicated:</i></p>
13811 <blockquote>
13812 <pre>template &lt;class Lock, class Rep, class Period, class Predicate&gt;
13813   bool wait_for(Lock&amp; lock, const chrono::duration&lt;Rep, Period&gt;&amp; rel_time, Predicate pred);</pre>
13814   <pre>...</pre>
13815
13816 <p><i>Throws:</i> <code>std::system_error</code> when <del>the returned value, effects or
13817 postcondition cannot be achieved</del> <ins>an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
13818 </blockquote>
13819
13820
13821
13822
13823
13824
13825 <hr>
13826 <h3><a name="963"></a>963. Various threading bugs #13</h3>
13827 <p><b>Section:</b> 30.3.1.5 [thread.thread.member] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
13828  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-26</p>
13829 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.member">active issues</a> in [thread.thread.member].</p>
13830 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.member">issues</a> in [thread.thread.member].</p>
13831 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
13832 <p><b>Discussion:</b></p>
13833 <p>
13834 30.3.1.5 [thread.thread.member]:  <tt>thread::detach</tt> is required to
13835 throw an exception if the thread is "not a detachable thread".
13836 "Detachable" is never defined.
13837 </p>
13838
13839 <p><i>[
13840 Howard adds:
13841 ]</i></p>
13842
13843
13844 <blockquote>
13845 Due to a mistake on my part, 3 proposed resolutions appeared at approximately
13846 the same time.  They are all three noted below in the discussion.
13847 </blockquote>
13848
13849 <p><i>[
13850 Summit, proposed resolution:
13851 ]</i></p>
13852
13853
13854 <blockquote>
13855 <p>
13856 In 30.3.1.5 [thread.thread.member] change:
13857 </p>
13858
13859 <blockquote><pre>void detach();
13860 </pre>
13861 <blockquote>
13862 <p>...</p>
13863 <p>-14- <i>Error conditions:</i></p>
13864 <ul>
13865 <li><tt>no_such_process</tt> --  <ins>if the thread is</ins> not <del>a</del> valid<del> thread</del>.</li>
13866 <li><tt>invalid_argument</tt> -- <ins>if the thread is</ins> not <del>a detachable</del> <ins>joinable</ins><del> thread</del>.</li>
13867 </ul>
13868 </blockquote>
13869
13870 </blockquote>
13871
13872 </blockquote>
13873
13874 <p><i>[
13875 Post Summit, Jonathan Wakely adds:
13876 ]</i></p>
13877
13878
13879 <blockquote>
13880 <p>
13881 A <tt>thread</tt> is detachable if it is joinable. As we've defined joinable,
13882 we can just use that.
13883 </p>
13884 <p>
13885 This corresponds to the pthreads specification, where pthread_detach
13886 fails if the thread is not joinable:
13887 </p>
13888 <blockquote>
13889 EINVAL: The  implementation  has  detected  that  the value specified by
13890 thread does not refer to a joinable thread.
13891 </blockquote>
13892 <p>
13893 Jonathan recommends this proposed wording:
13894 </p>
13895 <blockquote>
13896 <p>
13897 In 30.3.1.5 [thread.thread.member] change:
13898 </p>
13899
13900 <blockquote><pre>void detach();
13901 </pre>
13902 <blockquote>
13903 <p>...</p>
13904 <p>-14- <i>Error conditions:</i></p>
13905 <ul>
13906 <li>...</li>
13907 <li><tt>invalid_argument</tt> -- not a <del>detachable</del> <ins>joinable</ins> thread.</li>
13908 </ul>
13909 </blockquote>
13910
13911 </blockquote>
13912 </blockquote>
13913 </blockquote>
13914
13915 <p><i>[
13916 Post Summit, Anthony Williams adds:
13917 ]</i></p>
13918
13919
13920 <blockquote>
13921 <p>
13922 This is covered by the precondition that <tt>joinable()</tt> be <tt>true</tt>.
13923 </p>
13924 <p>
13925 Anthony recommends this proposed wording:
13926 </p>
13927
13928 <blockquote>
13929 <p>
13930 In 30.3.1.5 [thread.thread.member] change:
13931 </p>
13932
13933 <blockquote><pre>void detach();
13934 </pre>
13935 <blockquote>
13936 <p>...</p>
13937 <p>-14- <i>Error conditions:</i></p>
13938 <ul>
13939 <li>...</li>
13940 <li><del><tt>invalid_argument</tt> -- not a detachable thread.</del></li>
13941 </ul>
13942 </blockquote>
13943
13944 </blockquote>
13945
13946 </blockquote>
13947
13948 </blockquote>
13949
13950 <p><i>[
13951 2009-10 Santa Cruz:
13952 ]</i></p>
13953
13954
13955 <blockquote>
13956 Mark as Ready with proposed resolution from Summit.
13957 </blockquote>
13958
13959
13960
13961 <p><b>Proposed resolution:</b></p>
13962
13963 <p>
13964 In 30.3.1.5 [thread.thread.member] change:
13965 </p>
13966
13967 <blockquote><pre>void detach();
13968 </pre>
13969 <blockquote>
13970 <p>...</p>
13971 <p>-14- <i>Error conditions:</i></p>
13972 <ul>
13973 <li><tt>no_such_process</tt> --  <ins>if the thread is</ins> not <del>a</del> valid<del> thread</del>.</li>
13974 <li><tt>invalid_argument</tt> -- <ins>if the thread is</ins> not <del>a detachable</del> <ins>joinable</ins><del> thread</del>.</li>
13975 </ul>
13976 </blockquote>
13977
13978 </blockquote>
13979
13980
13981
13982
13983
13984
13985 <hr>
13986 <h3><a name="964"></a>964. Various threading bugs #14</h3>
13987 <p><b>Section:</b> 30.5.2 [thread.condition.condvarany] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
13988  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-26</p>
13989 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvarany">active issues</a> in [thread.condition.condvarany].</p>
13990 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvarany">issues</a> in [thread.condition.condvarany].</p>
13991 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
13992 <p><b>Discussion:</b></p>
13993 <p>
13994 The requirements for the constructor for <tt>condition_variable</tt> has several
13995 error conditions, but the requirements for the constructor for
13996 <tt>condition_variable_any</tt> has none. Is this difference intentional?
13997 </p>
13998
13999 <p><i>[
14000 Summit:
14001 ]</i></p>
14002
14003
14004 <blockquote>
14005 Move to open, pass to Howard. If this is intentional, a note may be
14006 helpful. If the error conditions are to be copied from
14007 <tt>condition_variable</tt>, this depends on LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#965">965</a>.
14008 </blockquote>
14009
14010 <p><i>[
14011 Post Summit Howard adds:
14012 ]</i></p>
14013
14014
14015 <blockquote>
14016 The original intention
14017 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2447.htm#ConditionVariablesWording">N2447</a>)
14018 was to let the OS return whatever errors it was going to return, and for
14019 those to be translated into exceptions, for both
14020 <tt>condition_variable</tt> and <tt>condition_variable_any</tt>.  I have not
14021 received any complaints about specific error conditions from vendors on
14022 non-POSIX platforms, but such complaints would not surprise me if they surfaced.
14023 </blockquote>
14024
14025 <p><i>[
14026 2009-10 Santa Cruz:
14027 ]</i></p>
14028
14029
14030 <blockquote>
14031 Leave open. Benjamin to provide wording.
14032 </blockquote>
14033
14034
14035
14036 <p><b>Proposed resolution:</b></p>
14037
14038
14039
14040
14041
14042 <hr>
14043 <h3><a name="966"></a>966. Various threading bugs #16</h3>
14044 <p><b>Section:</b> 30.5.1 [thread.condition.condvar] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
14045  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-26</p>
14046 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvar">active issues</a> in [thread.condition.condvar].</p>
14047 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvar">issues</a> in [thread.condition.condvar].</p>
14048 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
14049 <p><b>Discussion:</b></p>
14050 <p>
14051 30.5.1 [thread.condition.condvar]:
14052 <tt>condition_variable::wait</tt> and
14053 <tt>condition_variable::wait_until</tt> both have a postcondition that
14054 <tt>lock</tt> is locked by the calling thread, and a throws clause that
14055 requires throwing an exception if this postcondition cannot be achieved.
14056 How can the implementation detect that this <tt>lock</tt> can never be
14057 obtained?
14058 </p>
14059
14060 <p><i>[
14061 Summit:
14062 ]</i></p>
14063
14064
14065 <blockquote>
14066 Move to open. Requires wording. Agreed this is an issue, and the
14067 specification should not require detecting deadlocks.
14068 </blockquote>
14069
14070 <p><i>[
14071 2009-08-01 Howard provides wording.
14072 ]</i></p>
14073
14074
14075 <blockquote>
14076 <p>
14077 The proposed wording is inspired by the POSIX spec which says:
14078 </p>
14079
14080 <blockquote>
14081 <dl>
14082 <dt>[EINVAL]</dt>
14083 <dd>The value specified by cond or mutex is invalid.</dd>
14084 <dt>[EPERM]</dt>
14085 <dd>The mutex was not owned by the current thread at the time of the call.</dd>
14086 </dl>
14087 </blockquote>
14088
14089 <p>
14090 I do not believe [EINVAL] is possible without memory corruption (which we don't
14091 specify).  [EPERM] is possible if this thread doesn't own the mutex, which is
14092 listed as a precondition.  "May" is used instead of "Shall" because not all
14093 OS's are POSIX.
14094 </p>
14095 </blockquote>
14096
14097 <p><i>[
14098 2009-10 Santa Cruz:
14099 ]</i></p>
14100
14101
14102 <blockquote>
14103 Leave open, Detlef to provide improved wording.
14104 </blockquote>
14105
14106 <p><i>[
14107 2009-10-23 Detlef Provided wording.
14108 ]</i></p>
14109
14110
14111 <blockquote>
14112 <p>
14113 Detlef's wording put in Proposed resolution.  Original wording here:
14114 </p>
14115 <blockquote>
14116 <p>
14117 Change 30.5.1 [thread.condition.condvar] p12, p19 and
14118 30.5.2 [thread.condition.condvarany] p10, p16:
14119 </p>
14120
14121 <blockquote>
14122 <i>Throws:</i> <ins>May throw</ins> <tt>std::system_error</tt>
14123 <ins>
14124 if a precondition is not met.
14125 </ins>
14126 <del>when the effects or postcondition
14127 cannot be achieved.</del>
14128 </blockquote>
14129 </blockquote>
14130 </blockquote>
14131
14132 <p><i>[
14133 2009-10 Santa Cruz:
14134 ]</i></p>
14135
14136
14137 <blockquote>
14138 Leave open, Detlef to provide improved wording.
14139 </blockquote>
14140
14141
14142
14143 <p><b>Proposed resolution:</b></p>
14144 <p>
14145 Replace 30.5.1 [thread.condition.condvar] p12, p19 and
14146 30.5.2 [thread.condition.condvarany] p10, p16:
14147 </p>
14148
14149 <blockquote>
14150 <p><del>
14151 <i>Throws:</i> <tt>std::system_error</tt> when the effects or
14152 postcondition cannot be achieved.
14153 </del></p>
14154 <p><del>
14155 Error conditions:
14156 </del></p>
14157 <ul>
14158 <li><del>
14159 equivalent error condition from <tt>lock.lock()</tt> or <tt>lock.unlock()</tt>.
14160 </del></li>
14161 </ul>
14162
14163 <p><ins>
14164 <i>Throws:</i> It's implementation-defined whether a <tt>std::system_error</tt>
14165 with implementation-defined error condition is thrown if the
14166 precondition is not met.
14167 </ins></p>
14168 </blockquote>
14169
14170
14171
14172
14173
14174
14175 <hr>
14176 <h3><a name="967"></a>967. Various threading bugs #17</h3>
14177 <p><b>Section:</b> 30.3.1.2 [thread.thread.constr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
14178  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-22</p>
14179 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.constr">active issues</a> in [thread.thread.constr].</p>
14180 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.constr">issues</a> in [thread.thread.constr].</p>
14181 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
14182 <p><b>Discussion:</b></p>
14183 <p>
14184 the error handling for the constructor for <tt>condition_variable</tt>
14185 distinguishes lack of memory from lack of other resources, but the error
14186 handling for the thread constructor does not. Is this difference
14187 intentional?
14188 </p>
14189
14190 <p><i>[
14191 Beman has volunteered to provide proposed wording.
14192 ]</i></p>
14193
14194
14195 <p><i>[
14196 2009-09-25 Beman provided proposed wording.
14197 ]</i></p>
14198
14199
14200 <blockquote>
14201 The proposed resolution assumes <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a> has been accepted and
14202 its proposed resolution applied to the working paper.
14203 </blockquote>
14204
14205 <p><i>[
14206 2009-10 Santa Cruz:
14207 ]</i></p>
14208
14209
14210 <blockquote>
14211 Move to Ready.
14212 </blockquote>
14213
14214
14215
14216 <p><b>Proposed resolution:</b></p>
14217
14218
14219 <p><span style="font-style: italic;">Change Mutex requirements 30.4.1 [thread.mutex.requirements],
14220 paragraph 4, as indicated:</span></p>
14221 <blockquote>
14222
14223 <p><i>Error conditions:</i></p>
14224   <blockquote>
14225
14226 <ul>
14227 <li><del> <code>not_enough_memory</code> &#8212; if there is not enough memory to construct
14228 the mutex object.</del></li>
14229
14230 <li><code>resource_unavailable_try_again</code> &#8212; if any native handle type
14231 manipulated is not available.</li>
14232
14233 <li><code>operation_not_permitted</code> &#8212; if the thread does not have the
14234 necessary permission to change the state of the mutex object.</li>
14235
14236 <li><code>device_or_resource_busy</code> &#8212; if any native handle type
14237 manipulated is already locked.</li>
14238
14239 <li><code>invalid_argument</code> &#8212; if any native handle type manipulated as
14240 part of mutex construction is incorrect.</li>
14241 </ul>
14242   </blockquote>
14243 </blockquote>
14244
14245 <p><span style="font-style: italic;">Change Class condition_variable 30.5.1 [thread.condition.condvar],
14246 default constructor, as indicated:</span></p>
14247 <blockquote>
14248   <p><code>condition_variable();</code></p>
14249   <blockquote>
14250     <p><i>Effects:</i> Constructs an object of type <code>condition_variable</code>.</p>
14251     <p><ins><i>Throws:</i> <code>std::system_error</code> when an exception is required (30.2.2 [thread.req.exception])</ins>.</p>
14252     <p><i>Error conditions:</i></p>
14253     <blockquote>
14254     <ul>
14255       <li><del><code>not_enough_memory</code> &#8212; if a memory limitation prevents
14256       initialization.</del></li>
14257       <li> <code>resource_unavailable_try_again</code> &#8212; if some non-memory
14258       resource limitation prevents initialization.</li>
14259       <li> <code>device_or_resource_busy</code> &#8212; if attempting to initialize a
14260       previously-initialized but as of yet undestroyed <code>condition_variable</code>.</li>
14261     </ul>
14262     </blockquote>
14263   </blockquote>
14264 </blockquote>
14265
14266
14267
14268
14269
14270 <hr>
14271 <h3><a name="968"></a>968. Various threading bugs #18</h3>
14272 <p><b>Section:</b> 30.4.1 [thread.mutex.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
14273  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-01-07  <b>Last modified:</b> 2009-10-22</p>
14274 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.mutex.requirements">active issues</a> in [thread.mutex.requirements].</p>
14275 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.mutex.requirements">issues</a> in [thread.mutex.requirements].</p>
14276 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
14277 <p><b>Discussion:</b></p>
14278 <p>
14279 30.4.1 [thread.mutex.requirements]: several functions are
14280 required to throw exceptions "if the thread does not have the necessary
14281 permission ...". "The necessary permission" is not defined.
14282 </p>
14283
14284 <p><i>[
14285 Summit:
14286 ]</i></p>
14287
14288 <blockquote>
14289 Move to open.
14290 </blockquote>
14291
14292
14293 <p><i>[
14294 Beman has volunteered to provide proposed wording.
14295 ]</i></p>
14296
14297
14298 <p><i>[
14299 2009-10 Santa Cruz:
14300 ]</i></p>
14301
14302
14303 <blockquote>
14304 Moved to Ready with minor word-smithing in the example.
14305 </blockquote>
14306
14307
14308
14309 <p><b>Proposed resolution:</b></p>
14310
14311
14312 <p><i>Change Exceptions 30.2.2 [thread.req.exception] as indicated:</i></p>
14313 <blockquote>
14314 <p>Some functions described in this Clause are
14315 specified to throw exceptions of type <code>system_error</code> (19.5.5). Such exceptions
14316 shall be thrown if any of the <i>Error conditions</i> are detected or a call to an operating system or other underlying API
14317 results in an error that prevents the library function from meeting its specifications.
14318 <i>[Note:</i> See 17.6.4.11 [res.on.exception.handling] for exceptions thrown to report
14319 storage allocation failures. <i>&#8212;end
14320 note]</i></p>
14321
14322 <p><ins><i>[Example:</i></ins></p>
14323
14324 <blockquote>
14325
14326 <p><ins>Consider a function in this clause that is specified to throw exceptions of type <code>
14327 system_error</code> and specifies <i>Error conditions</i> that include <code>
14328 operation_not_permitted</code> for a thread that does not have the privilege to
14329 perform the operation. Assume that, during the execution of this function, an <code>errno</code>
14330 of <code>EPERM</code> is reported by a POSIX API call used by the
14331 implementation. Since POSIX specifies an <code>errno</code> of <code>EPERM</code>
14332 when "the caller does not have the privilege to perform the operation",
14333 the implementation maps <code>EPERM</code>&nbsp; to an <code>error_condition</code>
14334 of <code>operation_not_permitted</code> (19.5 [syserr]) and an exception of type <code>
14335 system_error</code> is thrown. </ins></p>
14336
14337 </blockquote>
14338
14339 <p><ins><i>&#8212;end example]</i></ins></p>
14340
14341 <p><span style="font-style: italic;">Editorial note: For the sake of exposition,
14342 the existing text above is shown with the changes proposed in issues 962 and 967. The
14343 proposed additional example is independent of whether or not the 962 and 967
14344 proposed resolutions are accepted.</span></p>
14345
14346 </blockquote>
14347
14348 <p><span style="font-style: italic;">Change Mutex requirements 30.4.1 [thread.mutex.requirements],
14349 paragraph 4, as indicated:</span></p>
14350
14351 <blockquote>
14352
14353 <p>&#8212; <code>operation_not_permitted</code> &#8212; if the thread does not have the
14354 <del>necessary permission to change the state of the mutex object</del> <ins>privilege to perform the operation</ins>.</p>
14355
14356 </blockquote>
14357
14358 <p><span style="font-style: italic;">Change Mutex requirements 30.4.1 [thread.mutex.requirements],
14359 paragraph 12, as indicated:</span></p>
14360
14361 <blockquote>
14362
14363 <p>&#8212; <code>operation_not_permitted</code> &#8212; if the thread does not have the
14364 <del>necessary permission to change the state of the mutex</del> <ins>privilege to perform the operation</ins>.</p>
14365
14366 </blockquote>
14367
14368
14369
14370
14371
14372
14373 <hr>
14374 <h3><a name="974"></a>974. <tt>duration&lt;double&gt;</tt> should not implicitly convert to <tt>duration&lt;int&gt;</tt></h3>
14375 <p><b>Section:</b> 20.9.3.1 [time.duration.cons] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
14376  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-01-21  <b>Last modified:</b> 2009-10-26</p>
14377 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
14378 <p><b>Discussion:</b></p>
14379 <p>
14380 The following code should not compile because it involves implicit truncation
14381 errors (against the design philosophy of the <tt>duration</tt> library).
14382 </p>
14383
14384 <blockquote><pre>duration&lt;double&gt; d(3.5);
14385 duration&lt;int&gt; i = d;  <font color="#c80000">// implicit truncation, should not compile</font>
14386 </pre></blockquote>
14387
14388 <p>
14389 This intent was codified in the example implementation which drove this proposal
14390 but I failed to accurately translate the code into the specification in this
14391 regard.
14392 </p>
14393
14394 <p><i>[
14395 Batavia (2009-05):
14396 ]</i></p>
14397
14398 <blockquote>
14399 <p>
14400 We agree with the proposed resolution.
14401 </p>
14402 <p>
14403 Move to Tentatively Ready.
14404 </p>
14405 </blockquote>
14406
14407 <p><i>[
14408 2009-07 Frankfurt
14409 ]</i></p>
14410
14411
14412 <blockquote>
14413 Moved from Tentatively Ready to Open only because the wording needs to be
14414 improved for enable_if type constraining, possibly following Robert's
14415 formula.
14416 </blockquote>
14417
14418 <p><i>[
14419 2009-08-01 Howard adds:
14420 ]</i></p>
14421
14422
14423 <blockquote>
14424 Addressed by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1177">1177</a>.
14425 </blockquote>
14426
14427 <p><i>[
14428 2009-10 Santa Cruz:
14429 ]</i></p>
14430
14431
14432 <blockquote>
14433 Not completely addressed by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1177">1177</a>.  Move to Ready.
14434 </blockquote>
14435
14436
14437
14438 <p><b>Proposed resolution:</b></p>
14439 <p>
14440 Change 20.9.3.1 [time.duration.cons], p4:
14441 </p>
14442
14443 <blockquote>
14444 <pre>template &lt;class Rep2, class Period2&gt;
14445   duration(const duration&lt;Rep2, Period2&gt;&amp; d);
14446 </pre>
14447 <blockquote>
14448 -4- <i>Requires:</i> <tt>treat_as_floating_point&lt;rep&gt;::value</tt>
14449 shall be <tt>true</tt> or <ins>both</ins> <tt>ratio_divide&lt;Period2,
14450 period&gt;::type::den</tt> shall be 1
14451 <ins>and <tt>treat_as_floating_point&lt;Rep2&gt;::value</tt>
14452 shall be <tt>false</tt></ins>.
14453 Diagnostic required.
14454 [<i>Note:</i> This requirement prevents implicit truncation error when
14455 converting between integral-based <tt>duration</tt> types. Such a
14456 construction could easily lead to confusion about the value of the
14457 <tt>duration</tt>. -- <i>end note</i>]
14458 </blockquote>
14459 </blockquote>
14460
14461
14462
14463
14464
14465 <hr>
14466 <h3><a name="978"></a>978. Hashing smart pointers</h3>
14467 <p><b>Section:</b> 20.7.16 [unord.hash] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
14468  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-02-02  <b>Last modified:</b> 2009-10-27</p>
14469 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.hash">active issues</a> in [unord.hash].</p>
14470 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.hash">issues</a> in [unord.hash].</p>
14471 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
14472 <p><b>Discussion:</b></p>
14473 <p><b>Addresses UK 208</b></p>
14474 <p>
14475 I don't see an open issue on supporting <tt>std::hash</tt> for smart pointers
14476 (<tt>unique_ptr</tt> and <tt>shared_ptr</tt> at least).
14477 </p>
14478 <p>
14479 It seems reasonable to at least expect support for the smart
14480 pointers, especially as they support comparison for use in ordered
14481 associative containers.
14482 </p>
14483
14484 <p><i>[
14485 Batavia (2009-05):
14486 ]</i></p>
14487
14488 <blockquote>
14489 <p>
14490 Howard points out that the client can always supply a custom hash function.
14491 </p>
14492 <p>
14493 Alisdair replies that the smart pointer classes are highly likely
14494 to be frequently used as hash keys.
14495 </p>
14496 <p>
14497 Bill would prefer to be conservative.
14498 </p>
14499 <p>
14500 Alisdair mentions that this issue may also be viewed as a subissue or
14501 duplicate of issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1025">1025</a>.
14502 </p>
14503 <p>
14504 Move to Open, and recommend the issue be deferred until after the next
14505 Committee Draft is issued.
14506 </p>
14507 </blockquote>
14508
14509 <p><i>[
14510 2009-05-31 Peter adds:
14511 ]</i></p>
14512
14513
14514 <blockquote>
14515 <blockquote>
14516 Howard points out that the client can always supply a custom hash function.
14517 </blockquote>
14518 <p>
14519 Not entirely true. The client cannot supply the function that hashes the
14520 address of the control block (the equivalent of the old <tt>operator&lt;</tt>, now
14521 proudly carrying the awkward name of '<tt>owner_before</tt>'). Only the
14522 implementation can do that, not necessarily via specializing <tt>hash&lt;&gt;</tt>, of
14523 course.
14524 </p>
14525 <p>
14526 This hash function makes sense in certain situations for <tt>shared_ptr</tt>
14527 (when one needs to switch from <tt>set/map</tt> using ownership ordering to
14528 <tt>unordered_set/map</tt>) and is the only hash function that makes sense for
14529 <tt>weak_ptr</tt>.
14530 </p>
14531 </blockquote>
14532
14533 <p><i>[
14534 2009-07-28 Alisdair provides wording.
14535 ]</i></p>
14536
14537
14538 <p><i>[
14539 2009-10 Santa Cruz:
14540 ]</i></p>
14541
14542
14543 <blockquote>
14544 Move to Ready.
14545 </blockquote>
14546
14547
14548
14549 <p><b>Proposed resolution:</b></p>
14550 <p>
14551 Add the following declarations to the synopsis of <tt>&lt;memory&gt;</tt>
14552 in 20.8 [memory]
14553 </p>
14554
14555 <blockquote><pre>// 20.8.10.X hash support
14556 template &lt;class T&gt; struct hash;
14557 template &lt;class T, class D&gt; struct hash&lt;unique_ptr&lt;T,D&gt;&gt;;
14558 template &lt;class T&gt; struct hash&lt;shared_ptr&lt;T&gt;&gt;;
14559 </pre></blockquote>
14560
14561 <p>
14562 Add a new subclause 20.8.10.X hash support
14563 </p>
14564
14565 <blockquote>
14566 <p>
14567 20.8.10.X hash support [util.smartptr.hash]
14568 </p>
14569
14570 <pre>template &lt;class T, class D&gt; struct hash&lt;unique_ptr&lt;T,D&gt;&gt;;
14571 </pre>
14572
14573 <blockquote>
14574 A partial specialization of the class template <tt>hash</tt> (20.7.16 [unord.hash]) shall be provided for instances of the
14575 <tt>unique_ptr</tt> template suitable for use as a key in unordered
14576 associative containers (23.5 [unord]) if and only if there is a
14577 <tt>hash</tt> specialization available for the type <tt>D::pointer</tt>.
14578 For an object <tt>p</tt> of type <tt>unqiue_ptr&lt;T,D&gt;</tt> the
14579 <tt>hash</tt> shall evaluate to the same value as <tt>hash&lt;typename
14580 D::pointer&gt;{}(p.get())</tt>.
14581 </blockquote>
14582
14583 <pre>template &lt;class T&gt; struct hash&lt;shared_ptr&lt;T&gt;&gt;;
14584 </pre>
14585
14586 <blockquote>
14587 A partial specialization of the class template <tt>hash</tt> (20.7.16 [unord.hash])
14588 shall be provided for instances of the <tt>shared_ptr</tt> template
14589 suitable for use as a key in unordered associative containers
14590 (23.5 [unord]). For an object <tt>p</tt> of type <tt>shared_ptr&lt;T&gt;</tt>
14591 the <tt>hash</tt> shall evaluate
14592 to the same value as <tt>hash&lt;T*&gt;{}(p.get())</tt>.
14593 </blockquote>
14594 </blockquote>
14595
14596
14597
14598
14599
14600 <hr>
14601 <h3><a name="983"></a>983. <tt>unique_ptr</tt> reference deleters should not be moved from</h3>
14602 <p><b>Section:</b> 20.8.14.2 [unique.ptr.single] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
14603  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-02-10  <b>Last modified:</b> 2009-10-21</p>
14604 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unique.ptr.single">active issues</a> in [unique.ptr.single].</p>
14605 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unique.ptr.single">issues</a> in [unique.ptr.single].</p>
14606 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
14607 <p><b>Discussion:</b></p>
14608 <p>
14609 Dave brought to my attention that when a <tt>unique_ptr</tt> has a non-const reference
14610 type deleter, move constructing from it, even when the <tt>unique_ptr</tt> containing
14611 the reference is an rvalue, could have surprising results:
14612 </p>
14613
14614 <blockquote><pre>D d(some-state);
14615 unique_ptr&lt;A, D&amp;&gt; p(new A, d);
14616 unique_ptr&lt;A, D&gt; p2 = std::move(p);
14617 <font color="#c80000">// has d's state changed here?</font>
14618 </pre></blockquote>
14619
14620 <p>
14621 I agree with him.  It is the <tt>unique_ptr</tt> that is the rvalue, not the
14622 deleter.  When the deleter is a reference type, the <tt>unique_ptr</tt> should
14623 respect the "lvalueness" of the deleter.
14624 </p>
14625
14626 <p>
14627 Thanks Dave.
14628 </p>
14629
14630 <p><i>[
14631 Batavia (2009-05):
14632 ]</i></p>
14633
14634 <blockquote>
14635 Seems correct, but complicated enough that we recommend moving to Review.
14636 </blockquote>
14637
14638 <p><i>[
14639 2009-10 Santa Cruz:
14640 ]</i></p>
14641
14642
14643 <blockquote>
14644 Move to Ready.
14645 </blockquote>
14646
14647
14648
14649 <p><b>Proposed resolution:</b></p>
14650 <p>
14651 Change 20.8.14.2.1 [unique.ptr.single.ctor], p20-21
14652 </p>
14653
14654 <blockquote>
14655 <pre>template &lt;class U, class E&gt; unique_ptr(unique_ptr&lt;U, E&gt;&amp;&amp; u);
14656 </pre>
14657
14658 <blockquote>
14659
14660 <p>
14661 -20- <i>Requires:</i> If <tt><del>D</del> <ins>E</ins></tt> is not a reference type,
14662 construction of the deleter <tt>D</tt> from an rvalue of type <tt>E</tt>
14663 shall be well formed and shall not throw an exception.
14664 <ins>
14665 Otherwise <tt>E</tt> is a reference type and construction of the deleter
14666 <tt>D</tt> from an lvalue of type <tt>E</tt> shall be well formed and
14667 shall not throw an exception.
14668 </ins>
14669 If <tt>D</tt> is
14670 a reference type, then <tt>E</tt> shall be the same type as <tt>D</tt>
14671 (diagnostic required). <tt>unique_ptr&lt;U, E&gt;::pointer</tt> shall be
14672 implicitly convertible to <tt>pointer</tt>. [<tt>Note:</tt> These
14673 requirements imply that <tt>T</tt> and <tt>U</tt> are complete types.
14674 <i>-- end note</i>]
14675 </p>
14676
14677 <p>
14678 -21- <i>Effects:</i> Constructs a <tt>unique_ptr</tt> which owns the
14679 pointer which <tt>u</tt> owns (if any). If the deleter
14680 <ins><tt>E</tt></ins> is not a reference type, <del>it</del> <ins>this
14681 deleter</ins> is move constructed from <tt>u</tt>'s deleter, otherwise
14682 <del>the reference</del> <ins>this deleter</ins> is copy constructed
14683 from <tt>u</tt>.'s deleter. After the construction, <tt>u</tt> no longer
14684 owns a pointer. [<i>Note:</i> The deleter constructor can be implemented
14685 with <tt>std::forward&lt;<del>D</del><ins>E</ins>&gt;</tt>. <i>-- end
14686 note</i>]
14687 </p>
14688
14689 </blockquote>
14690 </blockquote>
14691
14692 <p>
14693 Change 20.8.14.2.3 [unique.ptr.single.asgn], p1-3
14694 </p>
14695
14696 <blockquote>
14697 <pre>unique_ptr&amp; operator=(unique_ptr&amp;&amp; u);
14698 </pre>
14699 <blockquote>
14700
14701 <p>
14702 -1- <i>Requires:</i> <ins>If the deleter <tt>D</tt> is not a reference type,</ins>
14703 <del>A</del><ins>a</ins>ssignment of the deleter <tt>D</tt> from an rvalue <tt>D</tt> shall not throw an exception.
14704 <ins>
14705 Otherwise the deleter <tt>D</tt> is a reference type,
14706 and assignment of the deleter <tt>D</tt> from an lvalue <tt>D</tt> shall not throw an exception.</ins>
14707 </p>
14708
14709 <p>
14710 -2- <i>Effects:</i> reset(u.release()) followed by
14711 a<ins>n</ins> <del>move</del> assignment from <del><tt>u</tt>'s deleter to this deleter</del>
14712 <ins><tt>std::forward&lt;D&gt;(u.get_deleter())</tt></ins>.
14713 </p>
14714
14715 <p>
14716 -3- <i>Postconditions:</i> This <tt>unique_ptr</tt> now owns the pointer
14717 which <tt>u</tt> owned, and <tt>u</tt> no longer owns it. <del>[<i>Note:</i> If
14718 <tt>D</tt> is a reference type, then the referenced lvalue deleters are
14719 move assigned. <i>-- end note</i>]</del>
14720 </p>
14721 </blockquote>
14722 </blockquote>
14723
14724 <p>
14725 Change 20.8.14.2.3 [unique.ptr.single.asgn], p6-7
14726 </p>
14727
14728 <blockquote>
14729 <pre>template &lt;class U, class E&gt; unique_ptr&amp; operator=(unique_ptr&lt;U, E&gt;&amp;&amp; u);
14730 </pre>
14731 <blockquote>
14732
14733 <p>
14734 <i>Requires:</i> <ins>If the deleter <tt>E</tt> is not a reference type,</ins>
14735 <del>A</del><ins>a</ins>ssignment of the deleter <tt>D</tt> from an rvalue
14736 <tt><del>D</del><ins>E</ins></tt> shall not throw an exception.
14737 <ins>
14738 Otherwise the deleter <tt>E</tt> is a reference type,
14739 and assignment of the deleter <tt>D</tt> from an lvalue <tt>E</tt> shall not throw an exception.</ins>
14740 <tt>unique_ptr&lt;U, E&gt;::pointer</tt> shall be implicitly convertible to <tt>pointer</tt>.
14741 [<i>Note:</i> These requirements imply that <tt>T</tt> and <tt>U&gt;</tt>
14742 are complete types. <i>-- end note</i>]
14743 </p>
14744
14745 <p>
14746 <i>Effects:</i> <tt>reset(u.release())</tt> followed by
14747 a<ins>n</ins> <del>move</del> assignment from <del><tt>u</tt>'s deleter to this deleter</del>
14748 <ins><tt>std::forward&lt;E&gt;(u.get_deleter())</tt></ins>.
14749 <del>If either
14750 <tt>D</tt> or <tt>E</tt> is a reference type, then the referenced lvalue
14751 deleter participates in the move assignment.</del>
14752 </p>
14753
14754 </blockquote>
14755 </blockquote>
14756
14757
14758
14759
14760
14761
14762 <hr>
14763 <h3><a name="985"></a>985. Allowing throwing move</h3>
14764 <p><b>Section:</b> 23.2.1 [container.requirements.general] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
14765  <b>Submitter:</b> Rani Sharoni <b>Opened:</b> 2009-02-12  <b>Last modified:</b> 2009-10-20</p>
14766 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.requirements.general">active issues</a> in [container.requirements.general].</p>
14767 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.requirements.general">issues</a> in [container.requirements.general].</p>
14768 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
14769 <p><b>Discussion:</b></p>
14770 <p>
14771 <b>Introduction</b>
14772 </p>
14773
14774 <p>This proposal is meant to resolve potential regression of the
14775 <a href="" ref="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2800.pdf">N2800</a>
14776 draft, see
14777 next section, and to relax the requirements for containers of types with
14778 throwing move constructors.</p>
14779
14780 <p>The basic problem is that some containers operations, like <tt>push_back</tt>,
14781 have a strong exception safety
14782 guarantee (i.e. no side effects upon exception) that are not achievable when
14783 throwing move constructors are used since there is no way to guarantee revert
14784 after partial move. For such operations the implementation can at most provide
14785 the basic guarantee (i.e. valid but unpredictable) as it does with multi
14786 copying operations (e.g. range insert).</p>
14787
14788 <p>For example, <tt>vector&lt;T&gt;::push_back()</tt> (where <tt>T</tt> has a move
14789 constructor) might resize the <tt>vector</tt> and move the objects to the new underlying
14790 buffer. If move constructor throws it might
14791 not be possible to recover the throwing object or to move the old objects back to
14792 the original buffer.</p>
14793
14794 <p>The current draft is explicit by disallowing throwing move
14795 for some operations (e.g. <tt>vector&lt;&gt;::reserve</tt>) and not clear about other
14796 operations mentioned in 23.2.1 [container.requirements.general]/10
14797 (e.g. single element <tt>insert</tt>): it guarantees strong exception
14798 safety without explicitly disallowing a throwing move constructor.
14799 </p>
14800
14801 <p>
14802 <b>Regression</b>
14803 </p>
14804
14805 <p>This section only refers to cases in which the contained object
14806 is by itself a standard container.</p>
14807
14808 <p>Move constructors of standard containers are allowed to throw and therefore
14809 existing operations are broken, compared with C++03, due to move optimization.
14810 (In fact existing implementations like Dinkumware are actually throwing).</p>
14811
14812 <p>For example, <tt>vector&lt; list&lt;int&gt; &gt;::reserve</tt> yields
14813 undefined behavior since <tt>list&lt;int&gt;</tt>'s move constructor is allowed to throw.
14814 On the other hand, the same operation has strong exception safety guarantee in
14815 C++03.</p>
14816
14817 <p>There are few options to solve this regression:</p>
14818
14819 <ol>
14820 <li>
14821 Disallow throwing move and throwing default constructor
14822 </li>
14823
14824 <li>
14825 Disallow throwing move but disallowing usage after move
14826 </li>
14827
14828 <li>
14829 Special casing
14830 </li>
14831
14832 <li>
14833 Disallow throwing move and making it optional
14834 </li>
14835
14836 </ol>
14837
14838 <p>Option 1 is suggested by proposal
14839 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2815.html">N2815</a>
14840 but it might not be applicable for existing implementations for which
14841 containers default constructors are throwing.</p>
14842
14843 <p>Option 2 limits the usage significantly and it's error prone
14844 by allowing zombie objects that are nothing but destructible (e.g. no <tt>clear()</tt>
14845 is allowed after move). It also potentially complicates the implementation by
14846 introducing special state.</p>
14847
14848 <p>Option 3 is possible, for example, using default
14849 construction and <tt>swap</tt> instead of move for standard containers case. The
14850 implementation is also free to provide special hidden operation for non
14851 throwing move without forcing the user the cope with the limitation of option-2
14852 when using the public move.</p>
14853
14854 <p>Option 4 impact the efficiency in all use cases due to rare throwing move.</p>
14855
14856 <p>The proposed wording will imply option 1 or 3 though option 2 is also
14857 achievable using more wording. I personally oppose to option 2 that has impact
14858 on usability.</p>
14859
14860 <p>
14861 <b>Relaxation for user types</b>
14862 </p>
14863
14864 <p>Disallowing throwing move constructors in general seems very restrictive
14865 since, for example, common implementation of move will be default construction
14866 + <tt>swap</tt> so move will throw if the
14867 default constructor will throw. This is currently the case with the Dinkumware
14868 implementation of node based containers (e.g. <tt>std::list</tt>)
14869 though this section doesn't refer to standard types.</p>
14870
14871 <p>For throwing move constructors it seem that the implementation should have
14872 no problems to provide the basic guarantee instead of the strong one. It's
14873 better to allow throwing move constructors with basic guarantee than to
14874 disallow it silently (compile and run), via undefined behavior.</p>
14875
14876 <p>There might still be cases in which the relaxation will break existing generic
14877 code that assumes the strong guarantee but it's broken either way given a
14878 throwing move constructor since this is not a preserving optimization. </p>
14879
14880 <p><i>[
14881 Batavia (2009-05):
14882 ]</i></p>
14883
14884 <blockquote>
14885 <p>
14886 Bjarne comments (referring to his draft paper):
14887 "I believe that my suggestion simply solves that.
14888 Thus, we don't need a throwing move."
14889 </p>
14890 <p>
14891 Move to Open and recommend it be deferred until after the next
14892 Committee Draft is issued.
14893 </p>
14894 </blockquote>
14895
14896 <p><i>[
14897 2009-10 Santa Cruz:
14898 ]</i></p>
14899
14900
14901 <blockquote>
14902 Should wait to get direction from Dave/Rani
14903 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2983.html">N2983</a>).
14904 </blockquote>
14905
14906
14907
14908 <p><b>Proposed resolution:</b></p>
14909
14910 <p>
14911 23.2.1 [container.requirements.general]  paragraph 10 add footnote:
14912 </p>
14913
14914 <blockquote>
14915 <p>
14916 -10- Unless otherwise specified (see 23.1.4.1, 23.1.5.1, 23.2.2.3, and
14917 23.2.6.4) all container types defined in this Clause meet the following
14918 additional requirements:
14919 </p>
14920 <ul>
14921 <li>...</li>
14922 </ul>
14923
14924 <p>
14925 <ins>[<i>Note</i>: for compatibility with C++
14926 2003, when "no effect" is required, standard containers should not use the
14927 value_type's throwing move constructor when the contained object is by itself a
14928 standard container. -- <i>end note</i>]</ins>
14929 </p>
14930
14931 </blockquote>
14932
14933 <p>23.2.5.1 [unord.req.except] change paragraph 2 to say: </p>
14934
14935 <blockquote>
14936 <p>
14937 -2- For unordered associative containers, if an exception is
14938 thrown by any operation other than the container's hash function from within an
14939 <tt>insert()</tt> function inserting a single element, the <tt>insert()</tt>
14940 function has no effect<ins> unless the exception is thrown by the contained
14941 object move constructor</ins>.
14942 </p>
14943
14944 <p>
14945 -4- For unordered associative containers, if an exception is
14946 thrown from within a <tt>rehash()</tt> function other than by the container's hash
14947 function or comparison function, the <tt>rehash()</tt> function has no effect
14948 <ins>unless the exception is thrown by the contained
14949 object move constructor</ins>.</p>
14950
14951 </blockquote>
14952
14953 <p>
14954 23.3.2.3 [deque.modifiers] change paragraph 2 to say:
14955 </p>
14956
14957 <blockquote>
14958 -2- <i>Remarks:</i> If an exception is thrown other than by
14959 the copy constructor<ins>, move constructor</ins>
14960 or assignment operator of <tt>T</tt>
14961 there are no effects.
14962 <ins>If an exception is thrown by <tt>push_back()</tt> or <tt>emplace_back()</tt>
14963 function, that function has no effects unless the exception is thrown by
14964 the move constructor of <tt>T</tt>.</ins>
14965 </blockquote>
14966
14967 <p>
14968 23.3.2.3 [deque.modifiers] change paragraph 6 to say:
14969 </p>
14970
14971 <blockquote>
14972 -6- <i>Throws:</i> Nothing unless an exception is thrown by the copy
14973 constructor<ins>, move constructor</ins> or assignment operator of <tt>T</tt>.
14974 </blockquote>
14975
14976 <p>
14977 23.3.6.2 [vector.capacity] remove paragraph 2
14978 </p>
14979
14980 <blockquote>
14981 <del>-2- <i>Requires:</i> If <tt>value_type</tt> has a move constructor,
14982 that constructor shall not throw any exceptions.</del>
14983 </blockquote>
14984
14985 <p>
14986 23.3.6.2 [vector.capacity] paragraph 3 change to say:
14987 </p>
14988
14989 <blockquote>
14990 -3- <i>Effects:</i> A directive that informs a <tt>vector</tt>
14991 of a planned change in size, so
14992 that it can manage the storage allocation accordingly. After <tt>reserve()</tt>,
14993 <tt>capacity()</tt> is greater or equal to the argument of <tt>reserve</tt>
14994 if reallocation happens; and equal
14995 to the previous value of <tt>capacity()</tt>
14996 otherwise. Reallocation happens at this point if and only if the current
14997 capacity is less than the argument of <tt>reserve()</tt>.
14998 If an exception is thrown, there are no effects<ins>
14999 unless the exception is thrown by the contained object move constructor</ins>.
15000 </blockquote>
15001
15002 <p>
15003 23.3.6.2 [vector.capacity] paragraph 12 change to say:
15004 </p>
15005
15006 <blockquote>
15007 -12- <i>Requires:</i> <del>If <tt>value_type</tt> has a move constructor,
15008 that constructor shall not throw any exceptions.</del>
15009 <ins>If an exception is thrown, there are no effects unless the exception is thrown by
15010 the contained object move constructor.</ins>
15011 </blockquote>
15012
15013 <p>
15014 23.3.6.4 [vector.modifiers] change paragraph 1 to say:
15015 </p>
15016
15017 <blockquote>
15018 -1- <del><i>Requires:</i> If <tt>value_type</tt> has a move constructor,
15019 that constructor shall not throw any exceptions.</del>
15020 <ins><i>Remarks:</i> If an exception is thrown by <tt>push_back()</tt>
15021 or <tt>emplace_back()</tt> function, that function has no effect unless the
15022 exception is thrown by the move constructor of <tt>T</tt>.</ins>
15023 </blockquote>
15024
15025 <p>
15026 23.3.6.4 [vector.modifiers] change paragraph 2 to say:
15027 </p>
15028
15029 <blockquote>
15030 -2- <i>Remarks:</i> Causes reallocation if the new size is greater than
15031 the old capacity. If no reallocation happens, all the iterators and
15032 references before the insertion point remain valid. If an exception is
15033 thrown other than by the copy constructor<ins>, move constructor</ins>
15034 or assignment operator of <tt>T</tt> or by any <tt>InputIterator</tt>
15035 operation there are no effects.
15036 </blockquote>
15037
15038 <p>
15039 23.3.6.4 [vector.modifiers] change paragraph 6 to say:
15040 </p>
15041
15042 <blockquote>
15043 -6- <i>Throws:</i> Nothing unless an exception is thrown by the copy
15044 constructor<ins>, move constructor</ins> or assignment operator of <tt>T</tt>.
15045 </blockquote>
15046
15047
15048
15049
15050
15051
15052 <hr>
15053 <h3><a name="987"></a>987. <tt>reference_wrapper</tt> and function types</h3>
15054 <p><b>Section:</b> 20.7.5 [refwrap] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
15055  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-02-18  <b>Last modified:</b> 2009-10-26</p>
15056 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#refwrap">issues</a> in [refwrap].</p>
15057 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
15058 <p><b>Discussion:</b></p>
15059 <p>
15060 The synopsis in 20.7.5 [refwrap] says:
15061 </p>
15062
15063 <blockquote><pre>template &lt;<b>ObjectType</b> T&gt; class reference_wrapper
15064 ...
15065 </pre></blockquote>
15066
15067 <p>
15068 And then paragraph 3 says:
15069 </p>
15070
15071 <blockquote>
15072 <p>
15073 The template instantiation <tt>reference_wrapper&lt;T&gt;</tt> shall be
15074 derived from <tt>std::unary_function&lt;T1, R&gt;</tt> only if the type
15075 <tt>T</tt> is any of the following:
15076 </p>
15077
15078 <ul>
15079 <li>
15080 a <b>function type</b> or a pointer to function type taking one argument of
15081 type <tt>T1</tt> and returning <tt>R</tt>
15082 </li>
15083 </ul>
15084 </blockquote>
15085
15086 <p>
15087 But function types are not <tt>ObjectType</tt>s.
15088 </p>
15089
15090 <p>
15091 Paragraph 4 contains the same contradiction.
15092 </p>
15093
15094 <p><i>[
15095 Post Summit:
15096 ]</i></p>
15097
15098
15099 <blockquote>
15100 <p>
15101 Jens: restricted reference to ObjectType
15102 </p>
15103 <p>
15104 Recommend Review.
15105 </p>
15106 </blockquote>
15107
15108 <p><i>[
15109 Post Summit, Peter adds:
15110 ]</i></p>
15111
15112
15113 <blockquote>
15114 <p>
15115 In <a href="https://svn.boost.org/trac/boost/ticket/1846">https://svn.boost.org/trac/boost/ticket/1846</a>
15116 however Eric Niebler makes the very reasonable point that <tt>reference_wrapper&lt;F&gt;</tt>,
15117 where <tt>F</tt> is a function type, represents a reference to a function,
15118 a legitimate entity. So <tt>boost::ref</tt> was changed to allow it.
15119 </p>
15120 <p>
15121 <a href="https://svn.boost.org/trac/boost/browser/trunk/libs/bind/test/ref_fn_test.cpp">https://svn.boost.org/trac/boost/browser/trunk/libs/bind/test/ref_fn_test.cpp</a>
15122 </p>
15123 <p>
15124 Therefore, I believe an alternative proposed resolution for issue 987 could simply
15125 allow <tt>reference_wrapper</tt> to be used with function types.
15126 </p>
15127 </blockquote>
15128
15129 <p><i>[
15130 Post Summit, Howard adds:
15131 ]</i></p>
15132
15133
15134 <blockquote>
15135 <p>
15136 I agree with Peter (and Eric).  I got this one wrong on my first try.  Here
15137 is code that demonstrates how easy (and useful) it is to instantiate
15138 <tt>reference_wrapper</tt> with a function type:
15139 </p>
15140
15141 <blockquote><pre>#include &lt;functional&gt;
15142
15143 template &lt;class F&gt;
15144 void test(F f);
15145
15146 void f() {}
15147
15148 int main()
15149 {
15150     test(std::ref(f));
15151 }
15152 </pre></blockquote>
15153
15154 <p>
15155 Output (link time error shows type of <tt>reference_wrapper</tt> instantiated
15156 with function type):
15157 </p>
15158
15159 <blockquote><pre>Undefined symbols:
15160   "void test&lt;std::reference_wrapper&lt;void ()()&gt; &gt;(std::reference_wrapper&lt;void ()()&gt;)",...
15161 </pre></blockquote>
15162
15163 <p>
15164 I've taken the liberty of changing the proposed wording to allow function types
15165 and set to Open.  I'll also freely admit that I'm not positive <tt>ReferentType</tt>
15166 is the correct concept.
15167 </p>
15168
15169 </blockquote>
15170
15171
15172
15173 <p><i>[
15174 Batavia (2009-05):
15175 ]</i></p>
15176
15177 <blockquote>
15178 <p>
15179 Howard observed that <tt>FunctionType</tt>,
15180 a concept not (yet?) in the Working Paper,
15181 is likely the correct constraint to be applied.
15182 However, the proposed resolution provides an adequate approximation.
15183 </p>
15184 <p>
15185 Move to Review.
15186 </p>
15187 </blockquote>
15188
15189 <p><i>[
15190 2009-05-23 Alisdair adds:
15191 ]</i></p>
15192
15193
15194 <blockquote>
15195 <p>
15196 By constraining to <tt>PointeeType</tt> we rule out the ability for <tt>T</tt> to be a
15197 reference, and call in reference-collapsing.  I'm not sure if this is
15198 correct and intended, but would like to be sure the case was considered.
15199 </p>
15200 <p>
15201 Is dis-allowing reference types and the
15202 implied reference collapsing the intended result?
15203 </p>
15204 </blockquote>
15205
15206 <p><i>[
15207 2009-07 Frankfurt
15208 ]</i></p>
15209
15210
15211 <blockquote>
15212 Moved from Review to Open only because the wording needs to be
15213 tweaked for concepts removal.
15214 </blockquote>
15215
15216 <p><i>[
15217 2009-10-14 Daniel provided de-conceptified wording.
15218 ]</i></p>
15219
15220
15221 <p><i>[
15222 2009-10 post-Santa Cruz:
15223 ]</i></p>
15224
15225
15226 <blockquote>
15227 Move to Tentatively Ready.
15228 </blockquote>
15229
15230
15231
15232 <p><b>Proposed resolution:</b></p>
15233 <p>
15234 Change 20.7.5 [refwrap]/1 as indicated:
15235 </p>
15236
15237 <blockquote>
15238 <tt>reference_wrapper&lt;T&gt;</tt> is a <tt>CopyConstructible</tt> and
15239 <tt><ins>Copy</ins>Assignable</tt> wrapper around a
15240 reference to an object <ins>or function</ins> of type <tt>T</tt>.
15241 </blockquote>
15242
15243
15244
15245
15246
15247 <p><b>Rationale:</b></p>
15248 <p>
15249 a) The occurrence of <tt>T&amp;</tt> in the function signature auto-implies
15250 <tt>std::ReferentType</tt>,
15251 this is due to  [temp.req.impl]/4 bullet 4
15252 </p>
15253 <p>
15254 b) The occurrence of the constrained template <tt>reference_wrapper&lt;T&gt;</tt> in
15255 the remaining
15256 signatures lets kick in  [temp.req.impl]/4 bullet 1 and adds *all* requirements of
15257 this template. But we need to add at least *one* requirement (and it
15258 was an arbitrary,
15259 but natural decision to require <tt>std::PointeeType</tt> here) to *activate*
15260 this. If we hadn't done
15261 this, we were in unconstrained mode!
15262 </p>
15263
15264
15265
15266
15267
15268 <hr>
15269 <h3><a name="996"></a>996. Move operation not well specified</h3>
15270 <p><b>Section:</b> 17 [library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
15271  <b>Submitter:</b> David Abrahams <b>Opened:</b> 2009-03-06  <b>Last modified:</b> 2009-05-23</p>
15272 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
15273 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
15274 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
15275 <p><b>Discussion:</b></p>
15276 <p>
15277 There are lots of places in the standard where we talk about "the move
15278 constructor" but where we mean "the move operation," i.e.  <tt>T( move( x ) )</tt>.
15279 </p>
15280 <p>
15281 We also don't account for whether that operation modifies <tt>x</tt> or not, and
15282 we need to.
15283 </p>
15284
15285 <p><i>[
15286 Batavia (2009-05):
15287 ]</i></p>
15288
15289 <blockquote>
15290 Move to Open, pending proposed wording from Dave for further
15291 review.
15292 </blockquote>
15293
15294
15295 <p><b>Proposed resolution:</b></p>
15296 <p>
15297 </p>
15298
15299
15300
15301
15302
15303 <hr>
15304 <h3><a name="999"></a>999. Taking the address of a function</h3>
15305 <p><b>Section:</b> 20.8.13 [specialized.algorithms] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
15306  <b>Submitter:</b> Peter Dimov <b>Opened:</b> 2009-03-09  <b>Last modified:</b> 2009-10-26</p>
15307 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#specialized.algorithms">issues</a> in [specialized.algorithms].</p>
15308 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
15309 <p><b>Discussion:</b></p>
15310 <p>
15311 The same fix (reference <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#987">987</a>) may be applied to <tt>addressof</tt>, which is also constrained to
15312 <tt>ObjectType</tt>. (That was why <tt>boost::ref</tt> didn't work with functions - it
15313 tried to apply <tt>boost::addressof</tt> and the <tt>reinterpret_cast&lt;char&amp;&gt;</tt>
15314 implementation of <tt>addressof</tt> failed.)
15315 </p>
15316
15317
15318
15319 <p><i>[
15320 Batavia (2009-05):
15321 ]</i></p>
15322
15323 <blockquote>
15324 <p>
15325 We agree.
15326 </p>
15327 <p>
15328 Move to Tentatively Ready.
15329 </p>
15330 </blockquote>
15331
15332 <p><i>[
15333 2009-07 Frankfurt
15334 ]</i></p>
15335
15336
15337 <blockquote>
15338 Moved from Tentatively Ready to Open only because the wording needs to be
15339 tweaked for concepts removal.
15340 </blockquote>
15341
15342 <p><i>[
15343 2009-10-10 Daniel updates wording to concept-free.
15344 ]</i></p>
15345
15346
15347 <p><i>[
15348 2009-10 post-Santa Cruz:
15349 ]</i></p>
15350
15351
15352 <blockquote>
15353 Move to Tentatively Ready.
15354 </blockquote>
15355
15356
15357
15358 <p><b>Proposed resolution:</b></p>
15359 <p><i>[
15360 The resolution assumes that <tt>addressof</tt> is reintroduced as described in
15361 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2946.pdf">n2946</a>
15362 ]</i></p>
15363
15364
15365 <p>
15366 In 20.8.13 [specialized.algorithms] change as described:
15367 </p>
15368
15369 <blockquote><pre>template &lt;class T&gt; T* addressof(T&amp; r);
15370 </pre>
15371 <blockquote>
15372 <i>Returns:</i> The actual address of the object <ins>or function</ins>
15373 referenced by <tt>r</tt>, even in the
15374 presence of an overloaded <tt>operator&amp;</tt>.
15375 </blockquote>
15376 </blockquote>
15377
15378
15379
15380
15381 <p><b>Rationale:</b></p>
15382 <p>
15383 a) The occurrence of <tt>T&amp;</tt> in the function signature auto-implies
15384 <tt>std::ReferentType</tt>,
15385 this is due to  [temp.req.impl]/4 bullet 4
15386 </p>
15387
15388
15389
15390
15391
15392 <hr>
15393 <h3><a name="1008"></a>1008. <tt>nested_exception</tt> wording unclear</h3>
15394 <p><b>Section:</b> 18.8.6 [except.nested] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
15395  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-11  <b>Last modified:</b> 2009-10-22</p>
15396 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#except.nested">active issues</a> in [except.nested].</p>
15397 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#except.nested">issues</a> in [except.nested].</p>
15398 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
15399 <p><b>Discussion:</b></p>
15400
15401 <p><b>Addresses JP 31</b></p>
15402
15403 <p>
15404 It is difficult to understand in which case <tt>nested_exception</tt> is applied.
15405 </p>
15406
15407 <p><i>[
15408 Summit:
15409 ]</i></p>
15410
15411
15412 <blockquote>
15413 Alisdair will add an example in an update to
15414 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2619.pdf">N2619</a>.
15415 </blockquote>
15416
15417 <p><i>[
15418 2009-10 Santa Cruz:
15419 ]</i></p>
15420
15421
15422 <blockquote>
15423 It doesn't appear that N2619 really addresses this. Alisdair to propose wording.
15424 </blockquote>
15425
15426
15427
15428 <p><b>Proposed resolution:</b></p>
15429
15430
15431
15432
15433
15434 <hr>
15435 <h3><a name="1011"></a>1011. <tt>next/prev</tt> wrong iterator type</h3>
15436 <p><b>Section:</b> 24.4.4 [iterator.operations] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
15437  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-11  <b>Last modified:</b> 2009-10-22</p>
15438 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.operations">active issues</a> in [iterator.operations].</p>
15439 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.operations">issues</a> in [iterator.operations].</p>
15440 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
15441 <p><b>Discussion:</b></p>
15442
15443 <p><b>Addresses UK 271</b></p>
15444
15445 <p>
15446 <tt>next/prev</tt> return an incremented iterator without changing the value of
15447 the original iterator. However, even this may invalidate an
15448 <tt>InputIterator</tt>. A <tt>ForwardIterator</tt> is required to guarantee the
15449 'multipass' property.
15450 </p>
15451
15452 <p><i>[
15453 Batavia (2009-05):
15454 ]</i></p>
15455
15456 <blockquote>
15457 We agree with the proposed resolution.
15458 Move to Tentatively Ready.
15459 </blockquote>
15460
15461 <p><i>[
15462 2009-07 Frankfurt
15463 ]</i></p>
15464
15465
15466 <blockquote>
15467 Moved from Tentatively Ready to Open only because the wording needs to be
15468 tweaked for concepts removal.
15469 </blockquote>
15470
15471 <p><i>[
15472 2009-10-14 Daniel provided de-conceptified wording.
15473 ]</i></p>
15474
15475
15476 <p><i>[
15477 2009-10 Santa Cruz:
15478 ]</i></p>
15479
15480
15481 <blockquote>
15482 Moved to Ready.
15483 </blockquote>
15484
15485
15486
15487 <p><b>Proposed resolution:</b></p>
15488
15489
15490 <ol>
15491 <li>
15492 <p>
15493 Change header <tt>&lt;iterator&gt;</tt> synopsis 24.3 [iterator.synopsis] as indicated:
15494 </p>
15495
15496 <blockquote><pre>// 24.4.4, iterator operations:
15497 ...
15498 template &lt;class <del>Input</del><ins>Forward</ins>Iterator&gt;
15499   <del>Input</del><ins>Forward</ins>Iterator
15500   next(<del>Input</del><ins>Forward</ins>Iterator x, typename std::iterator_traits&lt;<del>Input</del><ins>Forward</ins>Iterator&gt;::difference_type n = 1);
15501 </pre></blockquote>
15502 </li>
15503
15504 <li>
15505 <p>
15506 Change 24.4.4 [iterator.operations] before p.6 as indicated:
15507 </p>
15508
15509 <blockquote><pre>template &lt;class <del>Input</del><ins>Forward</ins>Iterator&gt;
15510   <del>Input</del><ins>Forward</ins>Iterator
15511   next(<del>Input</del><ins>Forward</ins>Iterator x, typename std::iterator_traits&lt;<del>Input</del><ins>Forward</ins>Iterator&gt;::difference_type n = 1);
15512 </pre></blockquote>
15513 </li>
15514 </ol>
15515
15516
15517
15518
15519
15520
15521 <hr>
15522 <h3><a name="1030"></a>1030. Response to JP 44</h3>
15523 <p><b>Section:</b> 20.8.15.5 [util.smartptr.shared.atomic] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
15524  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-11  <b>Last modified:</b> 2009-10-23</p>
15525 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
15526 <p><b>Discussion:</b></p>
15527
15528 <p><b>Addresses JP 44</b></p>
15529
15530 <p>
15531 The 1st parameter <tt>p</tt> and 2nd parameter <tt>v</tt> is now
15532 <tt>shared_ptr&lt;T&gt;*</tt>.
15533 </p>
15534 <p>
15535 It should be <tt>shared_ptr&lt;T&gt;&amp;</tt>, or if these are
15536 <tt>shared_ptr&lt;T&gt;*</tt> then add the "<tt>p</tt> shall not be a
15537 null pointer" at the requires.
15538 </p>
15539
15540 <p><i>[
15541 Summit:
15542 ]</i></p>
15543
15544
15545 <blockquote>
15546 Agree. All of the functions need a requirement that <tt>p</tt> (or
15547 <tt>v</tt>) is a pointer to a valid object.
15548 </blockquote>
15549
15550 <p><i>[
15551 2009-07 post-Frankfurt:
15552 ]</i></p>
15553
15554
15555 <blockquote>
15556 <p>
15557 Lawrence explained that these signatures match the regular atomics. The
15558 regular atomics must not use references because these signatures are
15559 shared with C. The decision to pass shared_ptrs by pointer rather than
15560 by reference was deliberate and was motivated by the principle of least
15561 surprise.
15562 </p>
15563 <p>
15564 Lawrence to write wording that requires that the pointers not be null.
15565 </p>
15566 </blockquote>
15567
15568 <p><i>[
15569 2009-09-20 Lawrence provided wording:
15570 ]</i></p>
15571
15572
15573 <blockquote>
15574 <p>
15575 The parameter types for atomic shared pointer access
15576 were deliberately chosen to be pointers
15577 to match the corresponding parameters of the atomics chapter.
15578 Those in turn were deliberately chosen
15579 to match C functions,
15580 which do not have reference parameters.
15581 </p>
15582 <p>
15583 We adopt the second suggestion,
15584 to require that such pointers not be null.
15585 </p>
15586 </blockquote>
15587
15588 <p><i>[
15589 2009-10 Santa Cruz:
15590 ]</i></p>
15591
15592
15593 <blockquote>
15594 Moved to Ready.
15595 </blockquote>
15596
15597
15598
15599 <p><b>Proposed resolution:</b></p>
15600 <p>
15601 In section "<code>shared_ptr</code> atomic access"
15602 20.8.15.5 [util.smartptr.shared.atomic], add to each function the
15603 following clause.
15604 </p>
15605 <blockquote><p>
15606 <i>Requires:</i> <code>p</code> shall not be null.
15607 </p></blockquote>
15608
15609
15610
15611
15612
15613 <hr>
15614 <h3><a name="1033"></a>1033. <tt>thread::join()</tt> effects?</h3>
15615 <p><b>Section:</b> 30.3.1.5 [thread.thread.member] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
15616  <b>Submitter:</b> Alberto Ganesh Barbati <b>Opened:</b> 2009-03-12  <b>Last modified:</b> 2009-05-23</p>
15617 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.member">active issues</a> in [thread.thread.member].</p>
15618 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.member">issues</a> in [thread.thread.member].</p>
15619 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
15620 <p><b>Discussion:</b></p>
15621
15622 <p>
15623 While looking at <tt>thread::join()</tt> I think I spotted a couple of
15624 possible defects in the specifications. I could not find a previous
15625 issue or NB comment about that, but I might have missed it.
15626 </p>
15627
15628 <p>
15629 The postconditions clause for <tt>thread::join()</tt> is:
15630 </p>
15631
15632 <blockquote>
15633 <i>Postconditions:</i> If <tt>join()</tt> throws an exception, the value
15634 returned by <tt>get_id()</tt> is unchanged. Otherwise, <tt>get_id() == id()</tt>.
15635 </blockquote>
15636
15637 <p>
15638 and the throws clause is:
15639 </p>
15640
15641 <blockquote>
15642 <i>Throws:</i> <tt>std::system_error</tt> when the postconditions cannot be achieved.
15643 </blockquote>
15644
15645 <p>
15646 Now... how could the postconditions <em>not</em> be achieved?
15647 It's just a matter of resetting the value of <tt>get_id()</tt> or leave it
15648 unchanged! I bet we can always do that. Moreover, it's a chicken-and-egg
15649 problem: in order to decide whether to throw or not I depend on the
15650 postconditions, but the postconditions are different in the two cases.
15651 </p>
15652
15653 <p>
15654 I believe the throws clause should be:
15655 </p>
15656
15657 <blockquote>
15658 <i>Throws:</i> <tt>std::system_error</tt> when the effects or postconditions
15659 cannot be achieved.
15660 </blockquote>
15661
15662 <p>
15663 as it is in <tt>detach()</tt>, or, even better, as the postcondition is
15664 trivially satisfiable and to remove the circular dependency:
15665 </p>
15666
15667
15668 <blockquote>
15669 <i>Throws:</i> <tt>std::system_error</tt> if the effects cannot be achieved.
15670 </blockquote>
15671
15672 <p>
15673 Problem is that... ehm... <tt>join()</tt> has no "Effects" clause. Is that intentional?
15674 </p>
15675
15676 <p><i>[
15677 See the thread starting at c++std-lib-23204 for more discussion.
15678 ]</i></p>
15679
15680
15681 <p><i>[
15682 Batavia (2009-05):
15683 ]</i></p>
15684
15685 <blockquote>
15686 <p>
15687 Pete believes there may be some more general language (in frontmatter)
15688 that can address this and related issues such as <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#962">962</a>.
15689 </p>
15690 <p>
15691 Move to Open.
15692 </p>
15693 </blockquote>
15694
15695
15696 <p><b>Proposed resolution:</b></p>
15697
15698
15699
15700
15701
15702 <hr>
15703 <h3><a name="1034"></a>1034. Response to UK 222</h3>
15704 <p><b>Section:</b> 23.2.1 [container.requirements.general] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
15705  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-12  <b>Last modified:</b> 2009-10-20</p>
15706 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.requirements.general">active issues</a> in [container.requirements.general].</p>
15707 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.requirements.general">issues</a> in [container.requirements.general].</p>
15708 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
15709 <p><b>Discussion:</b></p>
15710
15711 <p><b>Addresses UK 222</b></p>
15712
15713 <p>
15714 It is not clear what purpose the Requirement tables serve in the
15715 Containers clause. Are they the definition of a library Container? Or
15716 simply a conventient shorthand to factor common semantics into a single
15717 place, simplifying the description of each subsequent container? This
15718 becomes an issue for 'containers' like <tt>array</tt>, which does not meet the
15719 default-construct-to-empty requirement, or <tt>forward_list</tt> which does not
15720 support the size operation. Are these components no longer containers?
15721 Does that mean the remaining requirements don't apply? Or are these
15722 contradictions that need fixing, despite being a clear design decision?
15723 </p>
15724
15725 <p>
15726 Recommend:
15727 </p>
15728
15729 <p>
15730 Clarify all the tables in 23.2 [container.requirements] are
15731 there as a convenience for documentation, rather than a strict set of
15732 requirements. Containers should be allowed to relax specific
15733 requirements if they call attention to them in their documentation. The
15734 introductory text for <tt>array</tt> should be expanded to mention a
15735 default constructed <tt>array</tt> is not empty, and
15736 <tt>forward_list</tt> introduction should mention it does not provide
15737 the required <tt>size</tt> operation as it cannot be implemented
15738 efficiently.
15739 </p>
15740
15741 <p><i>[
15742 Summit:
15743 ]</i></p>
15744
15745
15746 <blockquote>
15747 Agree in principle.
15748 </blockquote>
15749
15750 <p><i>[
15751 2009-07 post-Frankfurt:
15752 ]</i></p>
15753
15754
15755 <blockquote>
15756 We agree in principle, but we have a timetable. This group feels that
15757 the issue should be closed as NAD unless a proposed resolution is
15758 submitted prior to the March 2010 meeting.
15759 </blockquote>
15760
15761 <p><i>[
15762 2009-10 Santa Cruz:
15763 ]</i></p>
15764
15765
15766 <blockquote>
15767 Looked at this and still intend to close as NAD in March
15768 2010 unless there is proposed wording that we like.
15769 </blockquote>
15770
15771
15772
15773 <p><b>Proposed resolution:</b></p>
15774
15775
15776
15777
15778
15779 <hr>
15780 <h3><a name="1052"></a>1052. Response to UK 281</h3>
15781 <p><b>Section:</b> 24.5.1.3.5 [reverse.iter.opref] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
15782  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-12  <b>Last modified:</b> 2009-10-22</p>
15783 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
15784 <p><b>Discussion:</b></p>
15785
15786 <p><b>Addresses UK 281</b></p>
15787
15788 <p>
15789 The current specification for return value for <tt>reverse_iterator::operator-&gt;</tt>
15790 will always be a true pointer type, but <tt>reverse_iterator</tt> supports proxy
15791 iterators where the pointer type may be some kind of 'smart pointer'.
15792 </p>
15793
15794 <p><i>[
15795 Summit:
15796 ]</i></p>
15797
15798
15799 <blockquote>
15800 <p>
15801 <tt>move_iterator</tt> avoids this problem by returning a value of the wrapped
15802 Iterator type.
15803 study group formed to come up with a suggested resolution.
15804 </p>
15805 <p>
15806 <tt>move_iterator</tt> solution shown in proposed wording.
15807 </p>
15808 </blockquote>
15809
15810 <p><i>[
15811 2009-07 post-Frankfurt:
15812 ]</i></p>
15813
15814
15815 <blockquote>
15816 Howard to deconceptize. Move to Review after that happens.
15817 </blockquote>
15818
15819 <p><i>[
15820 2009-08-01 Howard deconceptized:
15821 ]</i></p>
15822
15823
15824 <blockquote>
15825 </blockquote>
15826
15827 <p><i>[
15828 2009-10 Santa Cruz:
15829 ]</i></p>
15830
15831
15832 <blockquote>
15833 <p>
15834 We can't think of any reason we can't just define reverse
15835 iterator's pointer types to be the same as the underlying iterator's
15836 pointer type, and get it by calling the right arrow directly.
15837 </p>
15838 <p>
15839 Here is the proposed wording that was replaced:
15840 </p>
15841 <blockquote><pre>template &lt;class Iterator&gt;
15842 class reverse_iterator {
15843   ...
15844   typedef <del>typename iterator_traits&lt;</del>Iterator<del>&gt;::pointer</del> pointer;
15845 </pre></blockquote>
15846
15847 <p>
15848 Change 24.5.1.3.5 [reverse.iter.opref]:
15849 </p>
15850
15851 <blockquote><pre>pointer operator-&gt;() const;
15852 </pre>
15853 <blockquote>
15854 <i>Returns:</i>
15855 <blockquote><pre><del>&amp;(operator*());</del>
15856 <ins>this-&gt;tmp = current;</ins>
15857 <ins>--this-&gt;tmp;</ins>
15858 <ins>return this-&gt;tmp;</ins>
15859 </pre></blockquote>
15860 </blockquote>
15861 </blockquote>
15862 </blockquote>
15863
15864
15865
15866 <p><b>Proposed resolution:</b></p>
15867 <p>
15868 Change 24.5.1.3.5 [reverse.iter.opref]:
15869 </p>
15870
15871 <blockquote><pre>pointer operator-&gt;() const;
15872 </pre>
15873 <blockquote>
15874 <i>Returns:</i>
15875 <blockquote><pre><del>&amp;(operator*());</del>
15876 <ins>deref_tmp = current;
15877 --deref_tmp;
15878 return deref_tmp::operator-&gt;();</ins>
15879 </pre></blockquote>
15880 </blockquote>
15881 </blockquote>
15882
15883
15884
15885
15886
15887
15888
15889
15890
15891 <hr>
15892 <h3><a name="1056"></a>1056. Must all Engines and Distributions be Streamable?</h3>
15893 <p><b>Section:</b> 26.5 [rand] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD">Tentatively NAD</a>
15894  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-12  <b>Last modified:</b> 2009-11-03</p>
15895 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#rand">issues</a> in [rand].</p>
15896 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD">Tentatively NAD</a> status.</p>
15897 <p><b>Discussion:</b></p>
15898
15899 <p>
15900 Both the concepts <tt>RandomNumberEngine</tt> and <tt>RandomNumberDistribution</tt> have
15901 requirements to be <tt>InputStreamable</tt> and <tt>OutputStreamable</tt>.
15902 </p>
15903 <p>
15904 I have no problems leaving the WP in an inconsistent state on the best-faith
15905 assumption these concepts will be provided later, however disagree with the
15906 proposers that these constraints are not separable, orthogonal to the basic
15907 concepts of generating random number distributions.
15908 </p>
15909 <p>
15910 These constraints should be dropped, and applied to specific algorithms as
15911 needed.
15912 </p>
15913 <p>
15914 If a more refined concept (certainly deemed useful by the proposers) is
15915 proposed there is no objection, but the basic concept should not require
15916 persistence via streaming.
15917 </p>
15918
15919 <p><i>[
15920 Batavia (2009-05):
15921 ]</i></p>
15922
15923 <blockquote>
15924 Move to Open.
15925 </blockquote>
15926
15927 <p><i>[
15928 2009-05-31 Alisdair adds:
15929 ]</i></p>
15930
15931
15932 <blockquote>
15933 <p>
15934 Working on constraining the stream iterators, I have a few more observations
15935 to make on the concepts proposed while constraining the random number
15936 facility.
15937 </p>
15938 <p>
15939 While I still believe the concerns are orthogonal, I don't believe the
15940 existing constraints go far enough either!  The goal we want to achieve is
15941 not that a <tt>RandomNumberEngine</tt> / <tt>RandomNumberDistribution</tt> supports the stream
15942 operators, but that it is <tt>Serializable</tt>.  I.e. there is a relationship
15943 between the insert and extract operations that guarantees to restore the
15944 state of the original object.  This implies a coupling of the concepts
15945 together in a broader concept (<tt>Serializable</tt>) with at least one axiom to
15946 assert the semantics.
15947 </p>
15948 <p>
15949 One problem is that <tt>istream</tt> and <tt>ostream</tt> may be fundamentally different
15950 types, although we can hook a relation if we are prepared to drop down to
15951 the <tt>char</tt> type and <tt>char_traits</tt> template parameters.  Doing so ties us to a
15952 form of serialization that demands implementation via the std iostreams
15953 framework, which seems overly prescriptive.  I believe the goal is generally
15954 to support serialization without regard to how it is expressed - although
15955 this is getting even more inventive in terms of concepts we do not have
15956 today.
15957 </p>
15958 </blockquote>
15959
15960 <p><i>[
15961 2009-11-03 Alisdair adds:
15962 ]</i></p>
15963
15964
15965 <blockquote>
15966 <p>
15967 I can't find the record in the wiki minutes, but it was agreed at both
15968 Frankfurt and Santa Cruz that this issue is NAD.
15969 </p>
15970 <p>
15971 The agreement in SC was that I would provide you with the rationale (see
15972 below) to include when moving to NAD.
15973 </p>
15974 </blockquote>
15975
15976 <p><i>[
15977 2009-11-03 Howard adds:
15978 ]</i></p>
15979
15980
15981 <blockquote>
15982 Moved to Tentatively NAD after 5 positive votes on c++std-lib.
15983 </blockquote>
15984
15985
15986 <p><b>Proposed resolution:</b></p>
15987
15988
15989 <p><b>Rationale:</b></p>
15990 <p>
15991 The issue suggests a more refined concept should be used if we want to
15992 require streaming, to separate concerns from the basic
15993 <tt>RandomNumberEngine</tt> behaviour.  In Frankfurt it was observed
15994 that <tt>RandomNumberEngine</tt> <em>is</em> that more refined concept,
15995 and the basic concept used in the framework is
15996 <tt>UniformRandomNumberGenerator</tt>, which it refines.
15997 </p>
15998
15999 <p>
16000 We concur, and expect this to have no repurcussions re-writing this
16001 clause now concepts are removed.
16002 </p>
16003
16004
16005
16006
16007
16008 <hr>
16009 <h3><a name="1068"></a>1068. class random_device should be movable</h3>
16010 <p><b>Section:</b> 26.5.6 [rand.device] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
16011  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-18  <b>Last modified:</b> 2009-10-26</p>
16012 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#rand.device">issues</a> in [rand.device].</p>
16013 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
16014 <p><b>Discussion:</b></p>
16015
16016 <p>
16017 class <tt>random_device</tt> should be movable.
16018 </p>
16019
16020 <p><i>[
16021 Batavia (2009-05):
16022 ]</i></p>
16023
16024 <blockquote>
16025 Move to Open, and recommend this issue be deferred until after the next
16026 Committee Draft is issued.
16027 </blockquote>
16028
16029 <p><i>[
16030 2009-10 post-Santa Cruz:
16031 ]</i></p>
16032
16033
16034 <blockquote>
16035 Leave open. Walter to provide drafting as part of his planned paper.
16036 </blockquote>
16037
16038
16039
16040 <p><b>Proposed resolution:</b></p>
16041
16042
16043
16044
16045
16046 <hr>
16047 <h3><a name="1069"></a>1069. class seed_seq should support efficient move operations</h3>
16048 <p><b>Section:</b> 26.5.7.1 [rand.util.seedseq] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
16049  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-18  <b>Last modified:</b> 2009-10-26</p>
16050 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#rand.util.seedseq">issues</a> in [rand.util.seedseq].</p>
16051 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
16052 <p><b>Discussion:</b></p>
16053
16054 <p>
16055 class <tt>seed_seq</tt> should support efficient move operations.
16056 </p>
16057
16058 <p><i>[
16059 Batavia (2009-05):
16060 ]</i></p>
16061
16062 <blockquote>
16063 Move to Open, and recommend this issue be deferred until after the next
16064 Committee Draft is issued.
16065 </blockquote>
16066
16067 <p><i>[
16068 2009-10 post-Santa Cruz:
16069 ]</i></p>
16070
16071
16072 <blockquote>
16073 Leave open. Walter to provide drafting as part of his planned paper.
16074 </blockquote>
16075
16076
16077
16078 <p><b>Proposed resolution:</b></p>
16079
16080
16081
16082
16083
16084 <hr>
16085 <h3><a name="1071"></a>1071. is_bind_expression should derive from integral_constant&lt;bool&gt;</h3>
16086 <p><b>Section:</b> 20.7.11.1.1 [func.bind.isbind] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
16087  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-19  <b>Last modified:</b> 2009-10-26</p>
16088 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
16089 <p><b>Discussion:</b></p>
16090
16091 <p>
16092 Class template is_bind_expression 20.7.11.1.1 [func.bind.isbind]:
16093 </p>
16094
16095 <blockquote><pre>namespace std {
16096   template&lt;class T&gt; struct is_bind_expression {
16097     static const bool value = see below;
16098   };
16099 }
16100 </pre></blockquote>
16101 <p>
16102 <tt>is_bind_expression</tt> should derive from <tt>std::integral_constant&lt;bool&gt;</tt> like
16103 other similar trait types.
16104 </p>
16105
16106 <p><i>[
16107 Daniel adds:
16108 ]</i></p>
16109
16110 <blockquote>
16111 We need the same thing for the trait <tt>is_placeholder</tt> as well.
16112 </blockquote>
16113 <p><i>[
16114 2009-03-22 Daniel provided wording.
16115 ]</i></p>
16116
16117
16118 <p><i>[
16119 Batavia (2009-05):
16120 ]</i></p>
16121
16122 <blockquote>
16123 <p>
16124 We recommend this be deferred until after the next Committee Draft is issued.
16125 </p>
16126 <p>
16127 Move to Open.
16128 </p>
16129 </blockquote>
16130
16131 <p><i>[
16132 2009-05-31 Peter adds:
16133 ]</i></p>
16134
16135
16136 <blockquote>
16137 <p>
16138 I am opposed to the proposed resolution and to the premise of the issue
16139 in general. The traits's default definitions should NOT derive from
16140 <tt>integral_constant</tt>, because this is harmful, as it misleads people into
16141 thinking that <tt>is_bind_expression&lt;E&gt;</tt> always derives from
16142 <tt>integral_constant</tt>, whereas it may not.
16143 </p>
16144 <p>
16145 <tt>is_bind_expression</tt> and <tt>is_placeholder</tt> allow user
16146 specializations, and in fact, this is their primary purpose. Such user
16147 specializations may not derive from <tt>integral_constant</tt>, and the
16148 places where <tt>is_bind_expression</tt> and <tt>is_placeholder</tt> are
16149 used intentionally do not require such derivation.
16150 </p>
16151 <p>
16152 The long-term approach here is to switch to
16153 <tt>BindExpression&lt;E&gt;</tt> and <tt>Placeholder&lt;P&gt;</tt>
16154 explicit concepts, of course, but until that happens, I say leave them
16155 alone.
16156 </p>
16157 </blockquote>
16158
16159 <p><i>[
16160 2009-10 post-Santa Cruz:
16161 ]</i></p>
16162
16163
16164 <blockquote>
16165 Move to Tentatively Ready.  We are comfortable with requiring user specializations
16166 to derive from <tt>integral_constant</tt>.
16167 </blockquote>
16168
16169
16170
16171 <p><b>Proposed resolution:</b></p>
16172 <ol>
16173 <li>
16174 <p>
16175 In 20.7.11.1.1 [func.bind.isbind] change as indicated:
16176 </p>
16177 <blockquote><pre>namespace std {
16178  template&lt;class T&gt; struct is_bind_expression <ins>: integral_constant&lt;bool, <i>see below</i>&gt; { };</ins><del>{
16179    static const bool value = <i>see below</i>;
16180  };</del>
16181 }
16182 </pre></blockquote>
16183 </li>
16184 <li>
16185 <p>
16186 In 20.7.11.1.1 [func.bind.isbind]/2 change as indicated:
16187 </p>
16188 <blockquote><pre><del>static const bool value;</del>
16189 </pre>
16190 <blockquote>
16191 -2- <del><tt>true</tt> if <tt>T</tt> is a type returned from <tt>bind</tt>, <tt>false</tt> otherwise.</del>
16192   <ins>If <tt>T</tt> is a type returned from <tt>bind</tt>, <tt>is_bind_expression&lt;T&gt;</tt> shall
16193 be publicly derived from
16194         <tt>integral_constant&lt;bool, true&gt;</tt>, otherwise it shall be
16195 publicly derived from
16196           <tt>integral_constant&lt;bool, false&gt;</tt>.</ins>
16197 </blockquote>
16198 </blockquote>
16199 </li>
16200 <li>
16201 <p>
16202 In 20.7.11.1.2 [func.bind.isplace] change as indicated:
16203 </p>
16204 <blockquote><pre>namespace std {
16205  template&lt;class T&gt; struct is_placeholder <ins>: integral_constant&lt;int, <i>see below</i>&gt; { };</ins><del>{
16206    static const int value = <i>see below</i>;
16207  };</del>
16208 }
16209 </pre></blockquote>
16210 </li>
16211 <li>
16212 <p>
16213 In 20.7.11.1.2 [func.bind.isplace]/2 change as indicated:
16214 </p>
16215 <blockquote><pre><del>static const int value;</del>
16216 </pre>
16217 <blockquote>
16218 -2- <del>value is <tt>J</tt> if <tt>T</tt> is the type of <tt>std::placeholders::_J</tt>, 0 otherwise.</del>
16219   <ins>If <tt>T</tt> is the type of <tt>std::placeholders::_J</tt>, <tt>is_placeholder&lt;T&gt;</tt>
16220 shall be publicly
16221           derived from <tt>integral_constant&lt;int, J&gt;</tt> otherwise it shall
16222 be publicly derived
16223           from <tt>integral_constant&lt;int, 0&gt;</tt>.</ins>
16224 </blockquote>
16225 </blockquote>
16226 </li>
16227 </ol>
16228
16229
16230
16231
16232
16233 <hr>
16234 <h3><a name="1076"></a>1076. unary/binary_negate need constraining and move support</h3>
16235 <p><b>Section:</b> 20.7.10 [negators] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
16236  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-20  <b>Last modified:</b> 2009-10-26</p>
16237 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
16238 <p><b>Discussion:</b></p>
16239 <p>
16240 The class templates <tt>unary/binary_negate</tt> need constraining and move support.
16241 </p>
16242 <p>
16243 Ideally these classes would be deprecated, allowing <tt>unary/binary_function</tt> to
16244 also be deprecated.  However, until a generic negate adaptor is introduced
16245 that can negate any <tt>Callable</tt> type, they must be supported so should be
16246 constrained.  Likewise, they should be movable, and support adopting a
16247 move-only predicate type.
16248 </p>
16249 <p>
16250 In order to preserve ABI compatibility, new rvalue overloads are supplied in
16251 preference to changing the existing pass-by-const-ref to pass-by-value.
16252 </p>
16253 <p>
16254 Do not consider the issue of forwarding mutable lvalues at this point,
16255 although remain open to another issue on the topic.
16256 </p>
16257
16258 <p><i>[
16259 2009-05-01 Daniel adds:
16260 ]</i></p>
16261
16262 <blockquote>
16263 <p>
16264 IMO the currently proposed resolution needs some updates
16265 because it is ill-formed at several places:
16266 </p>
16267
16268 <ol>
16269 <li>
16270 <p>
16271 In concept AdaptableUnaryFunction change
16272 </p>
16273 <blockquote><pre>typename X::result_type;
16274 typename X::argument_type;
16275 </pre></blockquote>
16276 <p>
16277 to
16278 </p>
16279 <blockquote><pre>Returnable result_type = typename X::result_type;
16280 typename argument_type = typename X::argument_type;
16281 </pre></blockquote>
16282 <p>
16283 [The replacement "Returnable result_type" instead of "typename
16284 result_type" is non-editorial, but maybe you prefer that as well]
16285 </p>
16286 </li>
16287 <li>
16288 <p>
16289 In concept AdaptableBinaryFunction change
16290 </p>
16291 <blockquote><pre>typename X::result_type;
16292 typename X::first_argument_type;
16293 typename X::second_argument_type;
16294 </pre></blockquote>
16295 <p>
16296 to
16297 </p>
16298 <blockquote><pre>Returnable result_type = typename X::result_type;
16299 typename first_argument_type = typename X::first_argument_type;
16300 typename second_argument_type = typename X::second_argument_type;
16301 </pre></blockquote>
16302 <p>
16303 [The replacement "Returnable result_type" instead of "typename
16304 result_type" is non-editorial, but maybe you prefer that as well.]
16305 </p>
16306 </li>
16307
16308 <li>
16309 <p>
16310 In class unary/binary_function
16311 </p>
16312 <ol type="a">
16313 <li>
16314 I suggest to change "ReturnType" to "Returnable" in both cases.
16315 </li>
16316 <li>
16317 I think you want to replace the remaining occurrences of "Predicate" by "P"
16318 (in both classes in copy/move from a predicate)
16319 </li>
16320 </ol>
16321 </li>
16322 <li>
16323 <p>
16324 I think you need to change the proposed signatures of not1 and not2, because
16325 they would still remain unconstrained: To make them constrained at least a
16326 single requirement needs to be added to enable requirement implication. This
16327 could be done via a dummy ("requires True&lt;true&gt;") or just explicit as follows:
16328 </p>
16329 <ol type="a">
16330 <li>
16331 <blockquote><pre>template &lt;AdaptableUnaryFunction P&gt;
16332 requires Predicate&lt; P, P::argument_type&gt;
16333 unary_negate&lt;P&gt; not1(const P&amp;&amp; pred);
16334 template &lt;AdaptableUnaryFunction P&gt;
16335 requires Predicate&lt; P, P::argument_type &gt;
16336 unary_negate&lt;P&gt; not1(P&amp;&amp; pred);
16337 </pre>
16338 <blockquote>
16339 -3- Returns: unary_negate&lt;P&gt;(pred).
16340 </blockquote>
16341 </blockquote>
16342 <p>
16343 [Don't we want a move call for the second overload as in
16344 </p>
16345 <blockquote><pre>unary_negate&lt;P&gt;(std::move(pred))
16346 </pre></blockquote>
16347 <p>
16348 in the Returns clause ?]
16349 </p>
16350 </li>
16351 <li>
16352 <pre>template &lt;AdaptableBinaryFunction P&gt;
16353 requires Predicate&lt; P, P::first_argument_type, P::second_argument_type &gt;
16354 binary_negate&lt;P&gt; not2(const P&amp; pred);
16355 template &lt;AdaptableBinaryFunction P&gt;
16356 requires Predicate&lt; P, P::first_argument_type, P::second_argument_type &gt;
16357 binary_negate&lt;P&gt; not2(P&amp;&amp; pred);
16358 </pre>
16359 <p>
16360 -5- Returns: binary_negate&lt;P&gt;(pred).
16361 </p>
16362 <p>
16363 [Don't we want a move call for the second overload as in
16364 </p>
16365 <blockquote><pre>binary_negate&lt;P&gt;(std::move(pred))
16366 </pre></blockquote>
16367 <p>
16368 in the Returns clause ?]
16369 </p>
16370 </li>
16371 </ol>
16372 </li>
16373 </ol>
16374 </blockquote>
16375
16376 <p><i>[
16377 Batavia (2009-05):
16378 ]</i></p>
16379
16380 <blockquote>
16381 <p>
16382 There is concern that complicating the solution
16383 to preserve the ABI seems unnecessary,
16384 since we're not in general preserving the ABI.
16385 </p>
16386 <p>
16387 We would prefer a separate paper consolidating all Clause 20
16388 issues that are for the purpose of providing constrained versions
16389 of the existing facilities.
16390 </p>
16391 <p>
16392 Move to Open.
16393 </p>
16394 </blockquote>
16395
16396 <p><i>[
16397 2009-10 post-Santa Cruz:
16398 ]</i></p>
16399
16400
16401 <blockquote>
16402 Leave open pending the potential move constructor paper. Note that
16403 we consider the "constraining" part NAD Concepts.
16404 </blockquote>
16405
16406
16407
16408 <p><b>Proposed resolution:</b></p>
16409 <p>
16410 Add new concepts where appropriate::
16411 </p>
16412
16413 <blockquote><pre>auto concept AdaptableUnaryFunction&lt; typename X &gt; {
16414   typename X::result_type;
16415   typename X::argument_type;
16416 }
16417
16418 auto concept AdaptableBinaryFunction&lt; typename X &gt; {
16419   typename X::result_type;
16420   typename X::first_argument_type;
16421   typename X::second_argument_type;
16422 }
16423 </pre></blockquote>
16424
16425 <p>
16426 Revise as follows:
16427 </p>
16428
16429 <p>
16430 Base 20.7.3 [base] (Only change is constrained Result)
16431 </p>
16432
16433 <blockquote>
16434 <p>
16435 -1-  The following classes are provided to simplify the typedefs of the
16436 argument and result types:
16437 </p>
16438 <pre>namespace std {
16439   template &lt;class Arg, <del>class</del> <ins>ReturnType</ins> Result&gt;
16440   struct unary_function {
16441      typedef Arg    argument_type;
16442      typedef Result result_type;
16443   };
16444
16445   template &lt;class Arg1, class Arg2, <del>class</del> <ins>ReturnType</ins> Result&gt;
16446   struct binary_function {
16447      typedef Arg1   first_argument_type;
16448      typedef Arg2   second_argument_type;
16449      typedef Result result_type;
16450   };
16451 }
16452 </pre></blockquote>
16453
16454 <p>
16455 Negators 20.7.10 [negators]:
16456 </p>
16457
16458 <blockquote>
16459 <p>
16460 -1- Negators <tt>not1</tt> and <tt>not2</tt> take a unary and a binary predicate,
16461 respectively, and return their complements (5.3.1).
16462 </p>
16463
16464 <pre>template &lt;<del>class</del> <ins>AdaptableUnaryFunction</ins> P<del>redicate</del>&gt;
16465   <ins>requires Predicate&lt; P, P::argument_type &gt;</ins>
16466   class unary_negate
16467     : public unary_function&lt;<del>typename</del> P<del>redicate</del>::argument_type,bool&gt; {
16468   public:
16469     <ins>unary_negate(const unary_negate &amp; ) = default;</ins>
16470     <ins>unary_negate(unary_negate &amp;&amp; );</ins>
16471
16472     <ins>requires CopyConstructible&lt; P &gt;</ins>
16473        explicit unary_negate(const Predicate&amp; pred);
16474     <ins>requires MoveConstructible&lt; P &gt;
16475        explicit unary_negate(Predicate &amp;&amp; pred);</ins>
16476
16477     bool operator()(const <del>typename</del> P<del>redicate</del>::argument_type&amp; x) const;
16478   };
16479 </pre>
16480 <blockquote>
16481 -2 <tt>operator()</tt> returns <tt>!pred(x)</tt>.
16482 </blockquote>
16483
16484 <pre>template &lt;class Predicate&gt;
16485   unary_negate&lt;Predicate&gt; not1(const Predicate&amp;amp; pred);
16486 <ins>template &lt;class Predicate&gt;
16487   unary_negate&lt;Predicate&gt; not1(Predicate&amp;&amp; pred);</ins>
16488 </pre>
16489 <blockquote>
16490 -3-  <i>Returns:</i> <tt>unary_negate&lt;Predicate&gt;(pred)</tt>.
16491 </blockquote>
16492
16493 <pre>template &lt;<del>class</del> <ins>AdaptableBinaryFunction</ins> P<del>redicate</del> &gt;
16494   <ins>requires Predicate&lt; P, P::first_argument_type, P::second_argument_type &gt;</ins>
16495   class binary_negate
16496     : public binary_function&lt;<del>typename</del> P<del>redicate</del>::first_argument_type,
16497                               <del>typename</del> P<del>redicate</del>::second_argument_type, bool&gt; {
16498   public:
16499     <ins>biary_negate(const binary_negate &amp; ) = default;</ins>
16500     <ins>binary_negate(binary_negate &amp;&amp; );</ins>
16501
16502     <ins>requires CopyConstructible&lt; P &gt;</ins>
16503        explicit binary_negate(const Predicate&amp; pred);
16504     <ins>requires MoveConstructible&lt; P &gt;
16505        explicit binary_negate(const Predicate&amp; pred);</ins>
16506
16507     bool operator()(const <del>typename</del> P<del>redicate</del>::first_argument_type&amp; x,
16508                     const <del>typename</del> P<del>redicate</del>::second_argument_type&amp; y) const;
16509   };
16510 </pre>
16511 <blockquote>
16512 -4- <tt>operator()</tt> returns <tt>!pred(x,y)</tt>.
16513 </blockquote>
16514
16515 <pre>template &lt;class Predicate&gt;
16516   binary_negate&lt;Predicate&gt; not2(const Predicate&amp; pred);
16517 <ins>template &lt;class Predicate&gt;
16518   binary_negate&lt;Predicate&gt; not2(Predicate&amp;&amp; pred);</ins>
16519 </pre>
16520
16521 <blockquote>
16522 -5- <i>Returns:</i> <tt>binary_negate&lt;Predicate&gt;(pred)</tt>.
16523 </blockquote>
16524 </blockquote>
16525
16526
16527
16528
16529
16530
16531 <hr>
16532 <h3><a name="1079"></a>1079. UK-265: <code>RandomAccessIterator</code>'s <code>operator-</code> has nonsensical effects clause</h3>
16533 <p><b>Section:</b> 24.2.5 [random.access.iterators] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
16534  <b>Submitter:</b> Doug Gregor <b>Opened:</b> 2009-03-20  <b>Last modified:</b> 2009-10-23</p>
16535 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#random.access.iterators">issues</a> in [random.access.iterators].</p>
16536 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
16537 <p><b>Discussion:</b></p>
16538 <p><b>Addresses UK 265</b></p>
16539
16540 <p>UK-265:</p>
16541 <p>
16542 This effects clause is nonesense. It looks more like an axiom stating
16543 equivalence, and certainly an effects clause cannot change the state of
16544 two arguments passed by const reference
16545 </p>
16546
16547 <p><i>[
16548 2009-09-18 Alisdair adds:
16549 ]</i></p>
16550
16551
16552 <blockquote>
16553 <p>
16554 For random access iterators, the definitions of <tt>(b-a)</tt> and
16555 <tt>(a&lt;b)</tt> are circular:
16556 </p>
16557
16558 <p>
16559 From table Table 104 -- Random access iterator requirements:
16560 </p>
16561
16562 <blockquote><pre>b - a :==&gt;  (a &lt; b) ? distance(a,b) : -distance(b,a)
16563
16564 a &lt; b :==&gt;  b - a &gt; 0
16565 </pre></blockquote>
16566 </blockquote>
16567
16568 <p><i>[
16569 2009-10 Santa Cruz:
16570 ]</i></p>
16571
16572
16573 <blockquote>
16574 Moved to Ready.
16575 </blockquote>
16576
16577
16578
16579 <p><b>Proposed resolution:</b></p>
16580
16581 <p>Modify 24.2.5 [random.access.iterators]p7-9 as follows:</p>
16582
16583 <blockquote><pre>difference_type operator-(const X&amp; a, const X&amp; b);
16584 </pre>
16585 <ol start="7">
16586   <li><i>Precondition</i>: there exists a value <code>n</code> of
16587   <code>difference_type</code> such that <code>a == b + n</code>.</li>
16588   <li><del><i>Effects</i>: <code>b == a + (b - a)</code></del></li>
16589   <li><i>Returns</i>: <del><code>(a &lt; b) ? distance(a,b) :
16590   -distance(b,a)</code></del><ins><code>n</code></ins></li>
16591 </ol>
16592 </blockquote>
16593
16594
16595
16596
16597
16598 <hr>
16599 <h3><a name="1089"></a>1089. Response to JP 76</h3>
16600 <p><b>Section:</b> 30 [thread] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
16601  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-03-22  <b>Last modified:</b> 2009-10-23</p>
16602 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread">issues</a> in [thread].</p>
16603 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
16604 <p><b>Discussion:</b></p>
16605 <p><b>Addresses JP 76</b></p>
16606
16607 <p>
16608 A description for "Throws: Nothing." are not unified.
16609 </p>
16610
16611 <p>
16612 At the part without throw, "Throws: Nothing." should be described.
16613 </p>
16614
16615 <p>
16616 Add "Throws: Nothing." to the following.
16617 </p>
16618
16619 <ul>
16620 <li>
16621 30.3.1.6 [thread.thread.static] p1
16622 </li>
16623 <li>
16624 30.4.3.1 [thread.lock.guard] p4
16625 </li>
16626 <li>
16627 30.4.3.2.1 [thread.lock.unique.cons] p6
16628 </li>
16629 <li>
16630 30.5.1 [thread.condition.condvar] p7 and p8
16631 </li>
16632 <li>
16633 30.5.2 [thread.condition.condvarany] p6, p7, p19, p21 and p25
16634 </li>
16635 </ul>
16636
16637 <p><i>[
16638 Summit:
16639 ]</i></p>
16640
16641 <blockquote>
16642 Pass on to editor.
16643 </blockquote>
16644
16645 <p><i>[
16646 Post Summit:  Editor declares this non-editorial.
16647 ]</i></p>
16648
16649
16650 <p><i>[
16651 2009-08-01 Howard provided wording:
16652 ]</i></p>
16653
16654
16655 <blockquote>
16656
16657 <p>
16658 The definition of "<i>Throws:</i> Nothing." that I added is probably going to
16659 be controversial, but I beg you to consider it seriously.
16660 </p>
16661
16662 <blockquote>
16663 <p>
16664 In C++ there are three "flow control" options for a function:
16665 </p>
16666
16667 <ol>
16668 <li>
16669 It can return, either with a value, or with <tt>void</tt>.
16670 </li>
16671 <li>
16672 It can call a function which never returns, such as <tt>std::exit</tt> or
16673 <tt>std::terminate</tt>.
16674 </li>
16675 <li>
16676 It can throw an exception.
16677 </li>
16678 </ol>
16679
16680 The above list can be abbreviated with:
16681
16682 <ol>
16683 <li><b>R</b>eturns.</li>
16684 <li><b>E</b>nds program.</li>
16685 <li><b>T</b>hrows exception.</li>
16686 </ol>
16687
16688 <p>
16689 In general a function can have the behavior of any of these 3, or any combination
16690 of any of these three, depending upon run time data.
16691 </p>
16692
16693 <ol>
16694 <li><b>R</b></li>
16695 <li><b>E</b></li>
16696 <li><b>T</b></li>
16697 <li><b>RE</b></li>
16698 <li><b>RT</b></li>
16699 <li><b>ET</b></li>
16700 <li><b>RET</b></li>
16701 </ol>
16702
16703 <p>
16704 A function with no throw spec, and no documentation, is in general a <b>RET</b>
16705 function.  It may return, it may end the program, or it may throw.  When we
16706 specify a function with an empty throw spec:
16707 </p>
16708
16709 <blockquote><pre>void f() throw();
16710 </pre></blockquote>
16711
16712 <p>
16713 We are saying that <tt>f()</tt> is an <b>RE</b> function:  It may return or end
16714 the program, but it will not throw.
16715 </p>
16716
16717 <p>
16718 I posit that there are very few places in the library half of the standard
16719 where we intend for functions to be able to end the program (call <tt>terminate</tt>).
16720 And none of those places where we do say <tt>terminate</tt> could be called,
16721 do we currently say "<i>Throws:</i> Nothing.".
16722 </p>
16723
16724 <p>
16725 I believe that if we define "<i>Throws:</i> Nothing." to mean <b>R</b>,
16726 we will both clarify many, many places in the standard, <em>and</em> give us a
16727 good rationale for choosing between "<i>Throws:</i> Nothing." (<b>R</b>)
16728 and <tt>throw()</tt> (<b>RE</b>) in the future.  Indeed, this may give us motivation
16729 to change several <tt>throw()</tt>s to "<i>Throws:</i> Nothing.".
16730 </p>
16731 </blockquote>
16732
16733 <p>
16734 I did not add the following changes as JP 76 requested as I believe we want to
16735 allow these functions to throw:
16736 </p>
16737
16738 <blockquote>
16739 <p>
16740 Add a paragraph under 30.4.3.1 [thread.lock.guard] p4:
16741 </p>
16742
16743 <blockquote><pre>explicit lock_guard(mutex_type&amp; m);
16744 </pre>
16745
16746 <p><ins>
16747 <i>Throws:</i> Nothing.
16748 </ins></p>
16749 </blockquote>
16750
16751 <p>
16752 Add a paragraph under 30.4.3.2.1 [thread.lock.unique.cons] p6:
16753 </p>
16754
16755 <blockquote><pre>explicit unique_lock(mutex_type&amp; m);
16756 </pre>
16757
16758 <p><ins>
16759 <i>Throws:</i> Nothing.
16760 </ins></p>
16761 </blockquote>
16762
16763 <p>
16764 Add a paragraph under 30.5.2 [thread.condition.condvarany] p19, p21 and p25:
16765 </p>
16766
16767 <blockquote><pre>template &lt;class Lock, class Rep, class Period&gt;
16768   bool wait_for(Lock&amp; lock, const chrono::duration&lt;Rep, Period&gt;&amp; rel_time);
16769 </pre>
16770
16771 <p><ins>
16772 <i>Throws:</i> Nothing.
16773 </ins></p>
16774 </blockquote>
16775
16776 <blockquote><pre>template &lt;class Lock, class Duration, class Predicate&gt;
16777   bool wait_until(Lock&amp; lock, const chrono::time_point&lt;Clock, Duration&gt;&amp; rel_time, Predicate pred);
16778 </pre>
16779
16780 <p><ins>
16781 <i>Throws:</i> Nothing.
16782 </ins></p>
16783 </blockquote>
16784
16785 <blockquote><pre>template &lt;class Lock, class Rep, class Period, class Predicate&gt;
16786   bool wait_for(Lock&amp; lock, const chrono::duration&lt;Rep, Period&gt;&amp; rel_time, Predicate pred);
16787 </pre>
16788
16789 <p><ins>
16790 <i>Throws:</i> Nothing.
16791 </ins></p>
16792 </blockquote>
16793
16794 </blockquote>
16795
16796 </blockquote>
16797
16798 <p><i>[
16799 2009-10 Santa Cruz:
16800 ]</i></p>
16801
16802
16803 <blockquote>
16804 Defer pending further developments with exception restriction annotations.
16805 </blockquote>
16806
16807
16808
16809 <p><b>Proposed resolution:</b></p>
16810 <p>
16811 Add a paragraph after 17.5.1.4 [structure.specifications] p4:
16812 </p>
16813
16814 <blockquote>
16815 <p>
16816 -3- Descriptions of function semantics contain the following elements
16817 (as appropriate):<sup>158</sup>
16818 </p>
16819 <ul>
16820 <li>...</li>
16821 <li>
16822 <i>Throws:</i> any exceptions thrown by the function, and the conditions
16823 that would cause the exception
16824 </li>
16825 <li>...</li>
16826 </ul>
16827
16828 <p>
16829 -4- For non-reserved replacement and handler functions, ...
16830 </p>
16831
16832 <p><ins>
16833 A "<i>Throws:</i> Nothing." element indicates that the function shall
16834 return ordinarily, and not exit via an exception.  This element also
16835 indicates that the function <em>shall</em> return. [<i>Note:</i> This
16836 differs from an empty throw specification which may cause a function to
16837 call <tt>unexpected</tt> and subsequently <tt>terminate</tt>. &#8212;
16838 <i>end note</i>]
16839 </ins></p>
16840 </blockquote>
16841
16842 <p>
16843 Add a paragraph under 30.3.1.6 [thread.thread.static] p1:
16844 </p>
16845
16846 <blockquote><pre>unsigned hardware_concurrency();
16847 </pre>
16848
16849 <p>
16850 -1- <i>Returns:</i> ...
16851 </p>
16852
16853 <p><ins>
16854 <i>Throws:</i> Nothing.
16855 </ins></p>
16856 </blockquote>
16857
16858 <p>
16859 Add a paragraph under 30.5.1 [thread.condition.condvar] p7 and p8:
16860 </p>
16861
16862 <blockquote>
16863 <p>
16864 <i>[Informational, not to be incluced in the WP: The POSIX spec allows only:</i>
16865 </p>
16866 <dl>
16867 <dt><i>[EINVAL]</i></dt>
16868 <dd><i>The value <tt>cond</tt> does not refer to an initialized condition variable. &#8212; end informational]</i></dd>
16869 </dl>
16870
16871 <pre>void notify_one();
16872 </pre>
16873
16874 <p>
16875 -7- <i>Effects:</i> ...
16876 </p>
16877
16878 <p><ins>
16879 <i>Throws:</i> Nothing.
16880 </ins></p>
16881 </blockquote>
16882
16883 <blockquote><pre>void notify_all();
16884 </pre>
16885
16886 <p>
16887 -8- <i>Effects:</i> ...
16888 </p>
16889
16890 <p><ins>
16891 <i>Throws:</i> Nothing.
16892 </ins></p>
16893 </blockquote>
16894
16895
16896 <p>
16897 Add a paragraph under 30.5.2 [thread.condition.condvarany] p6 and p7:
16898 </p>
16899
16900 <blockquote>
16901 <pre>void notify_one();
16902 </pre>
16903
16904 <p>
16905 -6- <i>Effects:</i> ...
16906 </p>
16907
16908 <p><ins>
16909 <i>Throws:</i> Nothing.
16910 </ins></p>
16911 </blockquote>
16912
16913 <blockquote><pre>void notify_all();
16914 </pre>
16915
16916 <p>
16917 -7- <i>Effects:</i> ...
16918 </p>
16919
16920 <p><ins>
16921 <i>Throws:</i> Nothing.
16922 </ins></p>
16923 </blockquote>
16924
16925
16926
16927
16928
16929
16930
16931 <hr>
16932 <h3><a name="1090"></a>1090. Missing description of <tt>packaged_task</tt> member <tt>swap</tt>,  missing non-member <tt>swap</tt></h3>
16933 <p><b>Section:</b> 30.6.10 [futures.task] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
16934  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-03-22  <b>Last modified:</b> 2009-10-26</p>
16935 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
16936 <p><b>Discussion:</b></p>
16937 <p>
16938 Class template <tt>packaged_task</tt> in 30.6.10 [futures.task] shows a member <tt>swap</tt>
16939 declaration, but misses to
16940 document it's effects (No prototype provided). Further on this class
16941 misses to provide a non-member
16942 swap.
16943 </p>
16944
16945 <p><i>[
16946 Batavia (2009-05):
16947 ]</i></p>
16948
16949 <blockquote>
16950 <p>
16951 Alisdair notes that paragraph 2 of the proposed resolution has already been
16952 applied in the current Working Draft.
16953 </p>
16954 <p>
16955 We note a pending <tt>future</tt>-related paper by Detlef;
16956 we would like to wait for this paper before proceeding.
16957 </p>
16958 <p>
16959 Move to Open.
16960 </p>
16961 </blockquote>
16962
16963 <p><i>[
16964 2009-05-24 Daniel removed part 2 of the proposed resolution.
16965 ]</i></p>
16966
16967
16968 <p><i>[
16969 2009-10 post-Santa Cruz:
16970 ]</i></p>
16971
16972
16973 <blockquote>
16974 Move to Tentatively Ready, removing bullet 3 from the proposed
16975 resolution but keeping the other two bullets.
16976 </blockquote>
16977
16978
16979
16980 <p><b>Proposed resolution:</b></p>
16981 <ol>
16982 <li>
16983 <p>
16984 In 30.6.10 [futures.task], immediately after the definition of class
16985 template packaged_task add:
16986 </p>
16987 <blockquote><pre><ins>
16988 template&lt;class R, class... Argtypes&gt;
16989 void swap(packaged_task&lt;R(ArgTypes...)&gt;&amp;, packaged_task&lt;R(ArgTypes...)&gt;&amp;);
16990 </ins>
16991 </pre></blockquote>
16992 </li>
16993 </ol>
16994
16995 <ol start="4">
16996
16997 <li>
16998 <p>
16999 At the end of 30.6.10 [futures.task] (after p. 20), add add the following
17000 prototype description:
17001 </p>
17002
17003 <blockquote><pre><ins>
17004 template&lt;class R, class... Argtypes&gt;
17005 void swap(packaged_task&lt;R(ArgTypes...)&gt;&amp; x, packaged_task&lt;R(ArgTypes...)&gt;&amp; y);
17006 </ins></pre>
17007 <blockquote>
17008 <p><ins>
17009 <i>Effects:</i> <tt>x.swap(y)</tt>
17010 </ins></p>
17011 <p><ins>
17012 <i>Throws:</i> Nothing.
17013 </ins></p>
17014 </blockquote>
17015 </blockquote>
17016 </li>
17017 </ol>
17018
17019
17020
17021
17022
17023 <hr>
17024 <h3><a name="1093"></a>1093. Multiple definitions for random_shuffle algorithm</h3>
17025 <p><b>Section:</b> 25.3.12 [alg.random.shuffle] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
17026  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-03-22  <b>Last modified:</b> 2009-10-26</p>
17027 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#alg.random.shuffle">issues</a> in [alg.random.shuffle].</p>
17028 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
17029 <p><b>Discussion:</b></p>
17030
17031 <p>
17032 There are a couple of issues with the declaration of the <tt>random_shuffle</tt>
17033 algorithm accepting a random number engine.
17034 </p>
17035
17036 <ol type="i">
17037 <li>
17038 The Iterators must be shuffle iterators, yet this requirement is missing.
17039 </li>
17040 <li>
17041 The <tt>RandomNumberEngine</tt> concept is now provided by the random number
17042 library
17043 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2836.pdf">n2836</a>)
17044 and the placeholder should be removed.
17045 </li>
17046 </ol>
17047
17048 <p><i>[
17049 2009-05-02 Daniel adds:
17050 ]</i></p>
17051
17052
17053 <blockquote>
17054 <p>
17055 this issue completes adding necessary requirement to the
17056 third new <tt>random_shuffle</tt> overload. The current suggestion is:
17057 </p>
17058
17059 <blockquote><pre>template&lt;RandomAccessIterator Iter, UniformRandomNumberGenerator Rand&gt;
17060 requires ShuffleIterator&lt;Iter&gt;
17061 void random_shuffle(Iter first, Iter last, Rand&amp;&amp; g);
17062 </pre></blockquote>
17063
17064 <p>
17065 IMO this is still insufficient and I suggest to add the requirement
17066 </p>
17067 <blockquote><pre>Convertible&lt;Rand::result_type, Iter::difference_type&gt;
17068 </pre></blockquote>
17069 <p>
17070 to the list (as the two other overloads already have).
17071 </p>
17072
17073 <p>
17074 Rationale:
17075 </p>
17076
17077 <blockquote>
17078 <p>
17079 Its true that this third overload is somewhat different from the remaining
17080 two. Nevertheless we know from <tt>UniformRandomNumberGenerator</tt>, that
17081 it's <tt>result_type</tt> is an integral type and that it satisfies
17082 <tt>UnsignedIntegralLike&lt;result_type&gt;</tt>.
17083 </p>
17084 <p>
17085 To realize it's designated task, the algorithm has to invoke the
17086 <tt>Callable</tt> aspect of <tt>g</tt> and needs to perform some algebra involving
17087 it's <tt>min()/max()</tt> limits to compute another index value that
17088 at this point is converted into <tt>Iter::difference_type</tt>. This is so,
17089 because 24.2.5 [random.access.iterators] uses this type as argument
17090 of it's algebraic operators. Alternatively consider the equivalent
17091 iterator algorithms in 24.4.4 [iterator.operations] with the same result.
17092 </p>
17093 <p>
17094 This argument leads us to the conclusion that we also need
17095 <tt>Convertible&lt;Rand::result_type, Iter::difference_type&gt;</tt> here.
17096 </p>
17097 </blockquote>
17098
17099 </blockquote>
17100
17101 <p><i>[
17102 Batavia (2009-05):
17103 ]</i></p>
17104
17105 <blockquote>
17106 <p>
17107 Alisdair notes that point (ii) has already been addressed.
17108 </p>
17109 <p>
17110 We agree with the proposed resolution to point (i)
17111 with Daniel's added requirement.
17112 </p>
17113 <p>
17114 Move to Review.
17115 </p>
17116 </blockquote>
17117
17118 <p><i>[
17119 2009-06-05 Daniel updated proposed wording as recommended in Batavia.
17120 ]</i></p>
17121
17122
17123 <p><i>[
17124 2009-07-28 Alisdair adds:
17125 ]</i></p>
17126
17127
17128 <blockquote>
17129 Revert to Open, with a note there is consensus on direction but the
17130 wording needs updating to reflect removal of concepts.
17131 </blockquote>
17132
17133 <p><i>[
17134 2009-10 post-Santa Cruz:
17135 ]</i></p>
17136
17137
17138 <blockquote>
17139 Leave Open, Walter to work on it.
17140 </blockquote>
17141
17142
17143
17144 <p><b>Proposed resolution:</b></p>
17145 <p>
17146 Change in  [algorithms.syn] and 25.3.12 [alg.random.shuffle]:
17147 </p>
17148
17149 <blockquote><pre><del>concept UniformRandomNumberGenerator&lt;typename Rand&gt; { }</del>
17150 template&lt;RandomAccessIterator Iter, UniformRandomNumberGenerator Rand&gt;
17151   <ins>requires ShuffleIterator&lt;Iter&gt; &amp;&amp;
17152   Convertible&lt;Rand::result_type, Iter::difference_type&gt;</ins>
17153   void random_shuffle(Iter first, Iter last, Rand&amp;&amp; g);
17154 </pre></blockquote>
17155
17156
17157
17158
17159
17160
17161 <hr>
17162 <h3><a name="1094"></a>1094. Response to JP 65 and JP 66</h3>
17163 <p><b>Section:</b> 27.5.4.3 [iostate.flags] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17164  <b>Submitter:</b> P.J. Plauger <b>Opened:</b> 2009-03-24  <b>Last modified:</b> 2009-10-20</p>
17165 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iostate.flags">issues</a> in [iostate.flags].</p>
17166 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17167 <p><b>Discussion:</b></p>
17168 <p><b>Addresses JP 65 and JP 66</b></p>
17169
17170 <p>
17171 Switch from "unspecified-bool-type" to "explicit operator bool() const".
17172 </p>
17173
17174 <p>
17175 Replace <tt>operator unspecified-bool-type() const;</tt>" with <tt>explicit operator bool() const;</tt>
17176 </p>
17177
17178 <p><i>[
17179 Batavia (2009-05):
17180 ]</i></p>
17181
17182 <blockquote>
17183 We agree with the proposed resolution.
17184 Move to Review.
17185 </blockquote>
17186
17187 <p><i>[
17188 2009 Santa Cruz:
17189 ]</i></p>
17190
17191
17192 <blockquote>
17193 Moved to Ready.
17194 </blockquote>
17195
17196
17197
17198 <p><b>Proposed resolution:</b></p>
17199 <p>
17200 Change the synopis in 27.5.4 [ios]:
17201 </p>
17202
17203 <blockquote><pre><ins>explicit</ins> operator <del><i>unspecified-bool-type</i></del> <ins>bool</ins>() const;
17204 </pre></blockquote>
17205
17206 <p>
17207 Change 27.5.4.3 [iostate.flags]:
17208 </p>
17209
17210 <blockquote><pre><ins>explicit</ins> operator <del><i>unspecified-bool-type</i></del> <ins>bool</ins>() const;
17211 </pre>
17212
17213 <blockquote>
17214 <p>
17215 -1- <i>Returns:</i> <ins><tt>!fail()</tt></ins> <del>If <tt>fail()</tt> then a value that will evaluate
17216 false in a boolean context; otherwise a value that will evaluate true in
17217 a boolean context. The value type returned shall not be convertible to
17218 int.</del>
17219 </p>
17220 <p>
17221 <del>[<i>Note:</i> This conversion can be used in contexts where a bool is expected
17222 (e.g., an <tt>if</tt> condition); however, implicit conversions (e.g.,
17223 to <tt>int</tt>) that can occur with <tt>bool</tt> are not allowed,
17224 eliminating some sources of user error. One possible implementation
17225 choice for this type is pointer-to-member. <i>-- end note</i>]</del>
17226 </p>
17227 </blockquote>
17228 </blockquote>
17229
17230
17231
17232
17233
17234
17235
17236 <hr>
17237 <h3><a name="1095"></a>1095. <i>Shared objects and the library</i> wording unclear</h3>
17238 <p><b>Section:</b> 17.6.3.10 [res.on.objects] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17239  <b>Submitter:</b> Beman Dawes <b>Opened:</b> 2009-03-27  <b>Last modified:</b> 2009-10-21</p>
17240 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17241 <p><b>Discussion:</b></p>
17242 <p>
17243 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2775.htm">N2775</a>,
17244 <i>Small library thread-safety revisions</i>, among other changes, removed a note from
17245 17.6.3.10 [res.on.objects] that read:
17246 </p>
17247
17248 <blockquote>
17249 [<i>Note:</i> This prohibition against concurrent non-const access means that
17250 modifying an object of a standard library type shared between threads
17251 without using a locking mechanism may result in a data race. <i>--end note</i>.]
17252 </blockquote>
17253
17254 <p>
17255 That resulted in wording which is technically correct but can only be
17256 understood by reading the lengthy and complex 17.6.4.8 [res.on.data.races]
17257 Data race avoidance. This has the effect of making
17258 17.6.3.10 [res.on.objects] unclear, and has already resulted in a query
17259 to the LWG reflector. See c++std-lib-23194.
17260 </p>
17261
17262 <p><i>[
17263 Batavia (2009-05):
17264 ]</i></p>
17265
17266 <blockquote>
17267 <p>
17268 The proposed wording seems to need a bit of tweaking
17269 ("really bad idea" isn't quite up to standardese).
17270 We would like feedback
17271 as to whether the original Note's removal was intentional.
17272 </p>
17273 <p>
17274 Change the phrase "is a really bad idea"
17275 to "risks undefined behavior" and
17276 move to Review status.
17277 </p>
17278 </blockquote>
17279
17280 <p><i>[
17281 2009-10 Santa Cruz:
17282 ]</i></p>
17283
17284
17285 <blockquote>
17286 Note: Change to read: "Modifying...", Delete 'thus', move to Ready
17287 </blockquote>
17288
17289
17290
17291 <p><b>Proposed resolution:</b></p>
17292 <p>
17293 Change 17.6.3.10 [res.on.objects] as indicated:
17294 </p>
17295
17296 <blockquote>
17297 <p>
17298 The behavior of a program is undefined if calls to standard library
17299 functions from different threads may introduce a data race. The
17300 conditions under which this may occur are specified in 17.6.4.7.
17301 </p>
17302 <p><ins>
17303 [<i>Note:</i> Modifying an object of a standard library type shared between
17304 threads risks undefined behavior unless objects of the type are explicitly
17305 specified as being sharable without data races or the user supplies a
17306 locking mechanism. <i>--end note</i>]
17307 </ins></p>
17308 </blockquote>
17309
17310
17311
17312
17313
17314 <hr>
17315 <h3><a name="1097"></a>1097. #define __STDCPP_THREADS</h3>
17316 <p><b>Section:</b> 18.2 [support.types] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17317  <b>Submitter:</b> Jens Maurer <b>Opened:</b> 2009-04-03  <b>Last modified:</b> 2009-10-21</p>
17318 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#support.types">issues</a> in [support.types].</p>
17319 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17320 <p><b>Discussion:</b></p>
17321 <p><b>Addresses DE 18</b></p>
17322
17323 <p>
17324 Freestanding implementations do not (necessarily) have
17325   support for multiple threads (see 1.10 [intro.multithread]).
17326   Applications and libraries may want to optimize for the
17327   absence of threads.  I therefore propose a preprocessor
17328   macro to indicate whether multiple threads can occur.
17329 </p>
17330
17331 <p>
17332 There is ample prior implementation experience for this
17333   feature with various spellings of the macro name.  For
17334   example, gcc implicitly defines <tt>_REENTRANT</tt>
17335   if multi-threading support is selected on the compiler
17336   command-line.
17337 </p>
17338
17339 <p>
17340 While this is submitted as a library issue, it may be more
17341   appropriate to add the macro in 16.8 cpp.predefined in the
17342   core language.
17343 </p>
17344
17345 <p>
17346 See also
17347 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2693.html">N2693</a>.
17348 </p>
17349
17350 <p><i>[
17351 Batavia (2009-05):
17352 ]</i></p>
17353
17354 <blockquote>
17355 <p>
17356 We agree with the issue, and believe it is properly a library issue.
17357 </p>
17358 <p>
17359 We prefer that the macro be conditionally defined
17360 as part of the <tt>&lt;thread&gt;</tt> header.
17361 </p>
17362 <p>
17363 Move to Review.
17364 </p>
17365 </blockquote>
17366
17367 <p><i>[
17368 2009-10 Santa Cruz:
17369 ]</i></p>
17370
17371
17372 <blockquote>
17373 Move to Ready.
17374 </blockquote>
17375
17376
17377
17378 <p><b>Proposed resolution:</b></p>
17379 <p>
17380 Insert a new subsection before 18.2 [support.types], entitled
17381 "Feature Macros" (support.macros):
17382 </p>
17383 <blockquote>
17384 <p>
17385 The standard library defines the following macros; no explicit
17386 prior inclusion of any header file is necessary.
17387 </p>
17388 <blockquote>
17389 <dl>
17390 <dt><tt>__STDCPP_THREADS</tt></dt>
17391 <dd>
17392 The macro <tt>__STDCPP_THREADS</tt> shall be defined if and only if a
17393     program can have more than one thread of execution (1.10 [intro.multithread]).
17394 If the macro is defined, it shall have the same
17395     value as the predefined macro <tt>__cplusplus</tt> (16.8 [cpp.predefined]).
17396 </dd>
17397 </dl>
17398 </blockquote>
17399 </blockquote>
17400
17401
17402
17403
17404
17405
17406 <hr>
17407 <h3><a name="1098"></a>1098. definition of get_pointer_safety()</h3>
17408 <p><b>Section:</b> 20.8.15.6 [util.dynamic.safety] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17409  <b>Submitter:</b> Jens Maurer <b>Opened:</b> 2009-04-03  <b>Last modified:</b> 2009-10-23</p>
17410 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#util.dynamic.safety">issues</a> in [util.dynamic.safety].</p>
17411 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17412 <p><b>Discussion:</b></p>
17413 <p><b>Addresses DE 18</b></p>
17414
17415 <p>
17416  In 20.8.15.6 [util.dynamic.safety], <tt>get_pointer_safety()</tt> purports
17417 to define behavior for
17418  non-safely derived pointers (3.7.4.3 [basic.stc.dynamic.safety]).  However,
17419  the cited core-language section in paragraph 4 specifies undefined behavior
17420  for the use of such pointer values.  This seems an unfortunate near-contradiction.
17421  I suggest to specify the term <i>relaxed pointer safety</i> in
17422  the core language section and refer to it from the library description.
17423  This issue deals with the library part, the corresponding core issue (c++std-core-13940)
17424  deals with the core modifications.
17425 </p>
17426
17427 <p>
17428 See also
17429 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2693.html">N2693</a>.
17430 </p>
17431
17432 <p><i>[
17433 Batavia (2009-05):
17434 ]</i></p>
17435
17436 <blockquote>
17437 <p>
17438 We recommend if this issue is to be moved,
17439 the issue be moved concurrently with the cited Core issue.
17440 </p>
17441 <p>
17442 We agree with the intent of the proposed resolution.
17443 We would like input from garbage collection specialists.
17444 </p>
17445 <p>
17446 Move to Open.
17447 </p>
17448 </blockquote>
17449
17450 <p><i>[
17451 2009-10 Santa Cruz:
17452 ]</i></p>
17453
17454
17455 <blockquote>
17456 The core issue is 853 and is in Ready status.
17457 </blockquote>
17458
17459
17460
17461 <p><b>Proposed resolution:</b></p>
17462 <p>
17463 In 20.8.15.6 [util.dynamic.safety] p16, replace the description of
17464 <tt>get_pointer_safety()</tt> with:
17465 </p>
17466
17467 <blockquote>
17468 <p>
17469 <tt>pointer_safety get_pointer_safety();</tt>
17470 </p>
17471 <blockquote>
17472 <p>
17473 <del><i>Returns:</i> an enumeration value indicating the implementation's treatment
17474 of pointers that are not safely derived (3.7.4.3). Returns
17475 <tt>pointer_safety::relaxed</tt> if pointers that are not safely derived will be
17476 treated the same as pointers that are safely derived for the duration of
17477 the program. Returns <tt>pointer_safety::preferred</tt> if pointers that are not
17478 safely derived will be treated the same as pointers that are safely
17479 derived for the duration of the program but allows the implementation to
17480 hint that it could be desirable to avoid dereferencing pointers that are
17481 not safely derived as described. [<i>Example:</i> <tt>pointer_safety::preferred</tt>
17482 might be returned to detect if a leak detector is running to avoid
17483 spurious leak reports. -- <i>end note</i>] Returns <tt>pointer_safety::strict</tt> if
17484 pointers that are not safely derived might be treated differently than
17485 pointers that are safely derived.</del>
17486 </p>
17487 <p><ins>
17488 <i>Returns:</i> Returns <tt>pointer_safety::strict</tt> if the implementation has
17489    strict pointer safety (3.7.4.3 [basic.stc.dynamic.safety]). It is
17490    implementation-defined whether <tt>get_pointer_safety</tt> returns
17491    <tt>pointer_safety::relaxed</tt> or <tt>pointer_safety::preferred</tt> if the
17492    implementation has relaxed pointer safety
17493    (3.7.4.3 [basic.stc.dynamic.safety]).<sup>Footnote</sup>
17494 </ins></p>
17495
17496 <p><ins>
17497 <i>Throws:</i> nothing
17498 </ins></p>
17499
17500 <p><ins>
17501 Footnote) <tt>pointer_safety::preferred</tt> might be returned to indicate to the
17502    program that a leak detector is running so that the program can avoid
17503    spurious leak reports.
17504 </ins>
17505 </p>
17506
17507 </blockquote>
17508 </blockquote>
17509
17510
17511
17512
17513
17514 <hr>
17515 <h3><a name="1099"></a>1099. Various issues</h3>
17516 <p><b>Section:</b> 17 [library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD">Tentatively NAD</a>
17517  <b>Submitter:</b> David Abrahams <b>Opened:</b> 2009-03-21  <b>Last modified:</b> 2009-10-26</p>
17518 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
17519 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
17520 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD">Tentatively NAD</a> status.</p>
17521 <p><b>Discussion:</b></p>
17522 <p>
17523 Notes
17524 </p>
17525 <blockquote>
17526 <p>
17527 [2009-03-21 Sat] p. 535 at the top we need MoveConstructible V1,
17528 MoveConstructible V2 (where V1,V2 are defined on 539).  Also make_tuple
17529 on 550
17530 </p>
17531
17532 <blockquote>
17533 <p>
17534 CD-1 reads:
17535 </p>
17536
17537 <blockquote><pre>template &lt;MoveConstructible T1, MoveConstructible T2&gt;
17538 pair&lt;V1, V2&gt; make_pair(T1&amp;&amp;, T2&amp;&amp;);
17539 </pre></blockquote>
17540
17541 <p>
17542 Actually I'm guessing we need something like <tt>MoveConstructible&lt;V1,T1&gt;</tt>,
17543 i.e. "<tt>V1</tt> can be constructed from an rvalue of type <tt>T1</tt>."
17544 </p>
17545
17546 <p>
17547 Ditto for <tt>make_tuple</tt>
17548 </p>
17549 </blockquote>
17550
17551 <p>
17552 [2009-03-21 Sat] p1183 thread ctor, and in general, we need a way to
17553 talk about "copiable from generalized rvalue ref argument" for cases
17554 where we're going to forward and copy.
17555 </p>
17556 <blockquote>
17557 <p>
17558    This issue may well be quite large.  Language in para 4 about "if
17559    an lvalue" is wrong because types aren't expressions.
17560 </p>
17561
17562 <blockquote>
17563 <p>
17564 Maybe we should define the term "move" so we can just say in the
17565 effects, "<tt>f</tt> is moved into the newly-created thread" or something, and
17566 agree (and ideally document) that saying "<tt>f</tt> is moved" implies
17567 </p>
17568
17569 <blockquote><pre>F x(move(f))
17570 </pre></blockquote>
17571
17572 <p>
17573 is required to work.  That would cover both ctors at once.
17574 </p>
17575 </blockquote>
17576
17577 <p>
17578    p1199, call_once has all the same issues.
17579 </p>
17580 </blockquote>
17581 <p>
17582 [2009-03-21 Sat] p869 InputIterator pointer type should not be required
17583 to be convertible to const value_type*, rather it needs to have a
17584 operator-&gt; of its own that can be used for the value type.
17585 </p>
17586
17587 <blockquote>
17588 This one is serious and unrelated to the move issue.
17589 </blockquote>
17590
17591 <p>
17592 [2009-03-21 Sat] p818 stack has the same problem with default ctor.
17593 </p>
17594 <p>
17595 [2009-03-21 Sat] p816 priority_queue has the same sorts of problems as queue, only more so
17596 </p>
17597 <blockquote><pre>   requires MoveConstructible&lt;Cont&gt;
17598      explicit priority_queue(const Compare&amp; x = Compare(), Cont&amp;&amp; = Cont());
17599 </pre>
17600 <p>
17601    Don't require MoveConstructible when default constructing Cont.
17602    Also missing semantics for move ctor.
17603 </p>
17604 </blockquote>
17605 <p>
17606  [2009-03-21 Sat] Why are Allocators required to be CopyConstructible as
17607  opposed to MoveConstructible?
17608 </p>
17609 <p>
17610  [2009-03-21 Sat] p813 queue needs a separate default ctor (Cont needn't
17611  be MoveConstructible).  No documented semantics for move c'tor.  Or
17612  *any* of its 7 ctors!
17613 </p>
17614 <p>
17615  [2009-03-21 Sat] std::array should have constructors for C++0x,
17616  consequently must consider move construction.
17617 </p>
17618
17619 <p><i>[
17620 2009-05-01 Daniel adds:
17621 ]</i></p>
17622
17623
17624 <blockquote>
17625 This could be done as part of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1035">1035</a>, which already handles
17626 deviation of <tt>std::array</tt> from container tables.
17627 </blockquote>
17628
17629 <p>
17630  [2009-03-21 Sat] p622 all messed up.
17631 </p>
17632 <blockquote>
17633 <p>
17634    para 8 "implementation-defined" is the wrong term; should be "see
17635    below" or something.
17636 </p>
17637 <p>
17638    para 12 "will be selected" doesn't make any sense because we're not
17639    talking about actual arg types.
17640 </p>
17641 <p>
17642    paras 9-13 need to be totally rewritten for concepts.
17643 </p>
17644 </blockquote>
17645
17646 <p>
17647  [2009-03-21 Sat] Null pointer comparisons (p587) have all become
17648  unconstrained.  Need to fix that
17649 </p>
17650 <p>
17651  [2009-03-21 Sat] mem_fun_t etc. definition doesn't match declaration.
17652   We think CopyConstructible is the right reqt.
17653 </p>
17654 <p>
17655  make_pair needs Constructible&lt;V1, T1&amp;&amp;&gt; requirements!
17656 </p>
17657 <p>
17658  make_tuple needs something similar
17659 </p>
17660 <p>
17661  tuple bug in synopsis:
17662 </p>
17663 <blockquote><pre>   template &lt;class... UTypes&gt;
17664    requires Constructible&lt;Types, const UTypes&amp;&gt;...
17665    template &lt;class... UTypes&gt;
17666    requires Constructible&lt;Types, RvalueOf&lt;UTypes&gt;::type&gt;...
17667 </pre>
17668 <p>
17669    Note: removal of MoveConstructible requirements in std::function makes
17670    these routines unconstrained!
17671 </p>
17672 </blockquote>
17673
17674 <p><i>[
17675 2009-05-02 Daniel adds:
17676 ]</i></p>
17677
17678
17679 <blockquote>
17680 This part of the issue is already covered by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1077">1077</a>.
17681 </blockquote>
17682
17683 <p>
17684  these unique_ptr constructors are broken [ I think this is covered in "p622 all messed up" ]
17685 </p>
17686 <blockquote><pre> unique_ptr(pointer p, implementation-defined d);
17687  unique_ptr(pointer p, implementation-defined d);
17688 </pre></blockquote>
17689 <p>
17690  multimap range constructor should not have MoveConstructible&lt;value_type&gt; requirement.
17691 </p>
17692 <blockquote>
17693    same with insert(..., P&amp;&amp;); multiset has the same issue, as do
17694    unordered_multiset and unordered_multimap. Review these!
17695 </blockquote>
17696
17697 </blockquote>
17698
17699 <p><i>[
17700 Batavia (2009-05):
17701 ]</i></p>
17702
17703 <blockquote>
17704 Move to Open, pending proposed wording from Dave for further review.
17705 </blockquote>
17706
17707 <p><i>[
17708 2009-10 post-Santa Cruz:
17709 ]</i></p>
17710
17711
17712 <blockquote>
17713 Tentatively NAD.  We are not sure what has been addressed and what hasn't.
17714 Recommend closing unless someone sorts this out into something more readable.
17715 </blockquote>
17716
17717
17718
17719 <p><b>Proposed resolution:</b></p>
17720 <p>
17721 </p>
17722
17723
17724
17725
17726
17727 <hr>
17728 <h3><a name="1100"></a>1100. <tt>auto_ptr</tt> to <tt>unique_ptr</tt> conversion</h3>
17729 <p><b>Section:</b> 20.8.14.2.1 [unique.ptr.single.ctor] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17730  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-04-25  <b>Last modified:</b> 2009-10-21</p>
17731 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unique.ptr.single.ctor">active issues</a> in [unique.ptr.single.ctor].</p>
17732 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unique.ptr.single.ctor">issues</a> in [unique.ptr.single.ctor].</p>
17733 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17734 <p><b>Discussion:</b></p>
17735 <p>
17736 Message c++std-lib-23182 led to a discussion in which several people
17737 expressed interest in being able to convert an <tt>auto_ptr</tt> to a
17738 <tt>unique_ptr</tt> without the need to call <tt>release</tt>.  Below is
17739 wording to accomplish this.
17740 </p>
17741
17742 <p><i>[
17743 Batavia (2009-05):
17744 ]</i></p>
17745
17746 <blockquote>
17747 <p>
17748 Pete believes it not a good idea to separate parts of a class's definition.
17749 Therefore, if we do this,
17750 it should be part of <tt>unique-ptr</tt>'s specification.
17751 </p>
17752 <p>
17753 Alisdair believes the lvalue overload may be not necessary.
17754 </p>
17755 <p>
17756 Marc believes it is more than just sugar,
17757 as it does ease the transition to <tt>unique-ptr</tt>.
17758 </p>
17759 <p>
17760 We agree with the resolution as presented.
17761 Move to Tentatively Ready.
17762 </p>
17763 </blockquote>
17764
17765 <p><i>[
17766 2009-07 Frankfurt
17767 ]</i></p>
17768
17769
17770 <blockquote>
17771 Moved from Tentatively Ready to Open only because the wording needs to be
17772 tweaked for concepts removal.
17773 </blockquote>
17774
17775 <p><i>[
17776 2009-08-01 Howard deconceptifies wording:
17777 ]</i></p>
17778
17779
17780 <blockquote>
17781 I also moved the change from D.10 [depr.auto.ptr]
17782 to 20.8.14.2 [unique.ptr.single] per the Editor's request
17783 in Batavia (as long as I was making changes anyway).  Set back
17784 to Review.
17785 </blockquote>
17786
17787 <p><i>[
17788 2009-10 Santa Cruz:
17789 ]</i></p>
17790
17791
17792 <blockquote>
17793 Move to Ready.
17794 </blockquote>
17795
17796
17797
17798 <p><b>Proposed resolution:</b></p>
17799 <p>
17800 Add to 20.8.14.2 [unique.ptr.single]:
17801 </p>
17802
17803 <blockquote><pre>template &lt;class T, class D&gt;
17804 class unique_ptr
17805 {
17806 public:
17807 <ins>    template &lt;class U&gt;
17808       unique_ptr(auto_ptr&lt;U&gt;&amp; u);
17809     template &lt;class U&gt;
17810       unique_ptr(auto_ptr&lt;U&gt;&amp;&amp; u);</ins>
17811 };
17812 </pre></blockquote>
17813
17814 <p>
17815 Add to 20.8.14.2.1 [unique.ptr.single.ctor]:
17816 </p>
17817
17818 <blockquote><pre>template &lt;class U&gt;
17819   unique_ptr(auto_ptr&lt;U&gt;&amp; u);
17820 template &lt;class U&gt;
17821   unique_ptr(auto_ptr&lt;U&gt;&amp;&amp; u);
17822 </pre>
17823 <blockquote>
17824 <p>
17825 <i>Effects:</i> Constructs a <tt>unique_ptr</tt> with <tt>u.release()</tt>.
17826 </p>
17827
17828 <p>
17829 <i>Postconditions:</i> <tt>get() == </tt> the value <tt>u.get()</tt> had before
17830 the construciton, modulo any required offset adjustments resulting from the cast from
17831 <tt>U*</tt> to <tt>T*</tt>.  <tt>u.get() == nullptr</tt>.
17832 </p>
17833
17834 <p>
17835 <i>Throws:</i> nothing.
17836 </p>
17837
17838 <p>
17839 <i>Remarks:</i> <tt>U*</tt> shall be implicitly convertible to <tt>T*</tt> and
17840 <tt>D</tt> shall be the same type as <tt>default_delete&lt;T&gt;</tt>, else these
17841 constructors shall not participate in overload resolution.
17842 </p>
17843 </blockquote>
17844 </blockquote>
17845
17846
17847
17848
17849
17850
17851 <hr>
17852 <h3><a name="1104"></a>1104. <tt>basic_ios::move</tt> should accept lvalues</h3>
17853 <p><b>Section:</b> 27.5.4.2 [basic.ios.members] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
17854  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-04-25  <b>Last modified:</b> 2009-10-20</p>
17855 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#basic.ios.members">active issues</a> in [basic.ios.members].</p>
17856 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#basic.ios.members">issues</a> in [basic.ios.members].</p>
17857 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
17858 <p><b>Discussion:</b></p>
17859 <p>
17860 With the rvalue reference changes in
17861 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2844.html">N2844</a>
17862 <tt>basic_ios::move</tt> no longer has the most convenient signature:
17863 </p>
17864
17865 <blockquote><pre>void move(basic_ios&amp;&amp; rhs);
17866 </pre></blockquote>
17867
17868 <p>
17869 This signature should be changed to accept lvalues.  It does not need to be
17870 overloaded to accept rvalues.  This is a special case that only derived clients
17871 will see.  The generic <tt>move</tt> still needs to accept rvalues.
17872 </p>
17873
17874 <p><i>[
17875 Batavia (2009-05):
17876 ]</i></p>
17877
17878 <blockquote>
17879 <p>
17880 Tom prefers, on general principles, to provide both overloads.
17881 Alisdair agrees.
17882 </p>
17883 <p>
17884 Howard points out that there is no backward compatibility issue
17885 as this is new to C++0X.
17886 </p>
17887 <p>
17888 We agree that both overloads should be provided,
17889 and Howard will provide the additional wording.
17890 Move to Open.
17891 </p>
17892 </blockquote>
17893
17894 <p><i>[
17895 2009-05-23 Howard adds:
17896 ]</i></p>
17897
17898
17899 <blockquote>
17900 Added overload, moved to Review.
17901 </blockquote>
17902
17903 <p><i>[
17904 2009 Santa Cruz:
17905 ]</i></p>
17906
17907
17908 <blockquote>
17909 Move to Ready.
17910 </blockquote>
17911
17912
17913
17914 <p><b>Proposed resolution:</b></p>
17915 <p>
17916 Add a signature to the existing prototype in the synopsis of 27.5.4 [ios]
17917 and in 27.5.4.2 [basic.ios.members]:
17918 </p>
17919
17920 <blockquote><pre><ins>void move(basic_ios&amp; rhs);</ins>
17921 void move(basic_ios&amp;&amp; rhs);
17922 </pre></blockquote>
17923
17924
17925
17926
17927
17928 <hr>
17929 <h3><a name="1106"></a>1106. Multiple exceptions from connected <tt>shared_future::get()</tt>?</h3>
17930 <p><b>Section:</b> 30.6.7 [future.shared_future] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
17931  <b>Submitter:</b> Thomas J. Gritzan <b>Opened:</b> 2009-04-03  <b>Last modified:</b> 2009-05-23</p>
17932 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#future.shared_future">issues</a> in [future.shared_future].</p>
17933 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
17934 <p><b>Discussion:</b></p>
17935 <p>
17936 It is not clear, if multiple threads are waiting in a
17937 <tt>shared_future::get()</tt> call, if each will rethrow the stored exception.
17938 </p>
17939 <p>
17940 Paragraph 9 reads:
17941 </p>
17942 <blockquote>
17943 <i>Throws:</i> the stored exception, if an exception was stored and not
17944 retrieved before.
17945 </blockquote>
17946 <p>
17947 The "not retrieved before" suggests that only one exception is thrown,
17948 but one exception for each call to <tt>get()</tt> is needed, and multiple calls
17949 to <tt>get()</tt> even on the same <tt>shared_future</tt> object seem to be allowed.
17950 </p>
17951 <p>
17952 I suggest removing "and not retrieved before" from the Throws paragraph.
17953 I recommend adding a note that explains that multiple calls on <tt>get()</tt> are
17954 allowed, and each call would result in an exception if an exception was
17955 stored.
17956 </p>
17957
17958 <p><i>[
17959 Batavia (2009-05):
17960 ]</i></p>
17961
17962 <blockquote>
17963 <p>
17964 We note there is a pending paper by Detlef
17965 on such <tt>future</tt>-related issues;
17966 we would like to wait for his paper before proceeding.
17967 </p>
17968 <p>
17969 Alisdair suggests we may want language to clarify that this
17970 <tt>get()</tt> function can be called from several threads
17971 with no need for explicit locking.
17972 </p>
17973 <p>
17974 Move to Open.
17975 </p>
17976 </blockquote>
17977
17978
17979 <p><b>Proposed resolution:</b></p>
17980 <p>
17981 Change 30.6.7 [future.shared_future]:
17982 </p>
17983
17984 <blockquote><pre>const R&amp; shared_future::get() const;
17985 R&amp; shared_future&lt;R&amp;&gt;::get() const;
17986 void shared_future&lt;void&gt;::get() const;
17987 </pre>
17988 <blockquote>
17989 <p>...</p>
17990 <p>
17991 -9- <i>Throws:</i> the stored exception, if an exception was stored<del> and not retrieved before</del>.
17992 <ins>
17993 [<i>Note:</i> Multiple calls on <tt>get()</tt> are
17994 allowed, and each call would result in an exception if an exception was
17995 stored. &#8212; <i>end note</i>]
17996 </ins>
17997 </p>
17998 </blockquote>
17999 </blockquote>
18000
18001
18002
18003
18004
18005
18006 <hr>
18007 <h3><a name="1108"></a>1108. thread.req.exception overly constrains implementations</h3>
18008 <p><b>Section:</b> 30.2.2 [thread.req.exception] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
18009  <b>Submitter:</b> Christopher Kohlhoff <b>Opened:</b> 2009-04-25  <b>Last modified:</b> 2009-10-26</p>
18010 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
18011 <p><b>Discussion:</b></p>
18012 <p>
18013 The current formulation of 30.2.2 [thread.req.exception]/2 reads:
18014 </p>
18015 <blockquote>
18016 The error_category of the <tt>error_code</tt> reported by such an
18017 exception's <tt>code()</tt> member function is as specified in the error
18018 condition Clause.
18019 </blockquote>
18020 <p>
18021 This constraint on the code's associated <tt>error_categor</tt> means an
18022 implementation must perform a mapping from the system-generated
18023 error to a <tt>generic_category()</tt> error code. The problems with this
18024 include:
18025 </p>
18026
18027 <ul>
18028 <li>
18029 The mapping is always performed, even if the resultant value is
18030  never used.
18031 </li>
18032 <li>
18033 <p>
18034 The original error produced by the operating system is lost.
18035 </p>
18036 </li>
18037 </ul>
18038 <p>
18039 The latter was one of Peter Dimov's main objections (in a private
18040 email discussion) to the original <tt>error_code</tt>-only design, and led to
18041 the creation of <tt>error_condition</tt> in the first place. Specifically,
18042 <tt>error_code</tt> and <tt>error_condition</tt> are intended to perform the following
18043 roles:
18044 </p>
18045 <ul>
18046 <li>
18047 <tt>error_code</tt> holds the original error produced by the operating
18048  system.
18049 </li>
18050 <li>
18051 <tt>error_condition</tt> and the generic category provide a set of well
18052  known error constants that error codes may be tested against.
18053 </li>
18054 </ul>
18055 <p>
18056 Any mapping determining correspondence of the returned error code to
18057 the conditions listed in the error condition clause falls under the
18058 "latitude" granted to implementors in 19.5.1.5 [syserr.errcat.objects].
18059 (Although obviously their latitude is restricted a little by the
18060 need to match the right error condition when returning an error code
18061 from a library function.)
18062 </p>
18063 <p>
18064 It is important that this <tt>error_code/error_condition</tt> usage is done
18065 correctly for the thread library since it is likely to set the
18066 pattern for future TR libraries that interact with the operating
18067 system.
18068 </p>
18069
18070 <p><i>[
18071 Batavia (2009-05):
18072 ]</i></p>
18073
18074 <blockquote>
18075 Move to Open, and recommend the issue be deferred until after the next
18076 Committee Draft is issued.
18077 </blockquote>
18078
18079 <p><i>[
18080 2009-10 post-Santa Cruz:
18081 ]</i></p>
18082
18083
18084 <blockquote>
18085 Move to Tentatively Ready.
18086 </blockquote>
18087
18088
18089
18090 <p><b>Proposed resolution:</b></p>
18091 <p>
18092 Change 30.2.2 [thread.req.exception]/2:
18093 </p>
18094
18095 <blockquote>
18096 <p>
18097 -2- <del>The <tt>error_category</tt> (19.5.1.1) of the <tt>error_code</tt> reported by
18098 such an exception's <tt>code()</tt> member function
18099 is as specified in the error condition Clause.</del>
18100 <ins>
18101 The <tt>error_code</tt> reported by such an exception's <tt>code()</tt> member
18102 function shall compare equal to one of the conditions specified in
18103 the function's error condition Clause. [<i>Example:</i> When the thread
18104 constructor fails:
18105 </ins>
18106 </p>
18107 <blockquote><pre><ins>
18108 ec.category() == implementation-defined // probably system_category
18109 ec == errc::resource_unavailable_try_again // holds true
18110 </ins></pre></blockquote>
18111
18112 <p><ins>
18113 &#8212; <i>end example</i>]
18114 </ins></p>
18115
18116 </blockquote>
18117
18118
18119
18120
18121
18122
18123 <hr>
18124 <h3><a name="1110"></a>1110. Is <tt>for_each</tt> overconstrained?</h3>
18125 <p><b>Section:</b> 25.2.4 [alg.foreach] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
18126  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-04-29  <b>Last modified:</b> 2009-10-27</p>
18127 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#alg.foreach">issues</a> in [alg.foreach].</p>
18128 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
18129 <p><b>Discussion:</b></p>
18130 <p>
18131 Quoting working paper for reference (25.2.4 [alg.foreach]):
18132 </p>
18133
18134 <blockquote>
18135 <pre>template&lt;InputIterator Iter, Callable&lt;auto, Iter::reference&gt; Function&gt;
18136   requires CopyConstructible&lt;Function&gt;
18137   Function for_each(Iter first, Iter last, Function f);
18138 </pre>
18139 <blockquote>
18140 <p>
18141 1 Effects: Applies f to the result of dereferencing every iterator in the
18142  range [first,last), starting from first and proceeding to last - 1.
18143 </p>
18144 <p>
18145 2 Returns: f.
18146 </p>
18147 <p>
18148 3 Complexity: Applies f exactly last - first times.
18149 </p>
18150 </blockquote>
18151 </blockquote>
18152
18153 <p>
18154 P2 implies the passed object <tt>f</tt> should be invoked at each stage, rather than
18155 some copy of <tt>f</tt>.  This is important if the return value is to usefully
18156 accumulate changes.  So the requirements are an object of type <tt>Function</tt> can
18157 be passed-by-value, invoked multiple times, and then return by value.  In
18158 this case, <tt>MoveConstructible</tt> is sufficient. This would open support for
18159 move-only functors, which might become important in concurrent code as you
18160 can assume there are no other references (copies) of a move-only type and so
18161 freely use them concurrently without additional locks.
18162 </p>
18163
18164 <p><i>[
18165 See further discussion starting with c++std-lib-23686.
18166 ]</i></p>
18167
18168
18169 <p><i>[
18170 Batavia (2009-05):
18171 ]</i></p>
18172
18173 <blockquote>
18174 <p>
18175 Pete suggests we may want to look at this in a broader context
18176 involving other algorithms.
18177 We should also consider the implications of parallelism.
18178 </p>
18179 <p>
18180 Move to Open, and recommend the issue be deferred until after the next
18181 Committee Draft is issued.
18182 </p>
18183 </blockquote>
18184
18185 <p><i>[
18186 2009-10-14 Daniel de-conceptified the proposed resolution.
18187 ]</i></p>
18188
18189
18190 <blockquote>
18191 <p>
18192 The note in 25.1 [algorithms.general]/9 already says the right thing:
18193 </p>
18194 <blockquote>
18195 Unless otherwise specified, algorithms that take function objects
18196 as arguments are permitted to copy those function objects freely.
18197 </blockquote>
18198 <p>
18199 So we only need to ensure that the wording for <tt>for_each</tt> is sufficiently
18200 clear, which is the intend of the following rewording.
18201 </p>
18202 </blockquote>
18203
18204 <p><i>[
18205 2009-10-15 Daniel proposes:
18206 ]</i></p>
18207
18208
18209 <blockquote>
18210 <ul>
18211 <li>
18212 <p>
18213 Add a new Requires clause just after the prototype declaration (25.2.4 [alg.foreach]):
18214 </p>
18215 <blockquote>
18216 <p>
18217 <ins><i>Requires:</i> <tt>Function</tt> shall be <tt>MoveConstructible</tt>
18218 ( [moveconstructible]), <tt>CopyConstructible</tt> is not required.</ins>
18219 </p>
18220 </blockquote>
18221 </li>
18222 <li>
18223 <p>
18224 Change 25.2.4 [alg.foreach]/2 as indicated:
18225 </p>
18226
18227 <blockquote>
18228 <i>Returns:</i> <ins>std::move(</ins>f<ins>)</ins>.
18229 </blockquote>
18230 </li>
18231 </ul>
18232 </blockquote>
18233
18234 <p><i>[
18235 2009-10 post-Santa Cruz:
18236 ]</i></p>
18237
18238
18239 <blockquote>
18240 Move to Tentatively Ready, using Daniel's wording without the portion
18241 saying "CopyConstructible is not required".
18242 </blockquote>
18243
18244 <p><i>[
18245 2009-10-27 Daniel adds:
18246 ]</i></p>
18247
18248
18249 <blockquote>
18250 <p>
18251 I see that during the Santa Cruz meeting the originally proposed addition
18252 </p>
18253
18254 <blockquote>
18255 , <tt>CopyConstructible</tt> is not required.
18256 </blockquote>
18257
18258 <p>
18259 was removed. I don't think that this removal was a good idea. The combination
18260 of 25.1 [algorithms.general]/9
18261 </p>
18262
18263 <blockquote>
18264 [<i>Note:</i> Unless otherwise specified, algorithms that take function objects
18265 as arguments are permitted to copy those function objects freely.[..]
18266 </blockquote>
18267
18268 <p>
18269 with the fact that <tt>CopyConstructible</tt> is a refinement <tt>MoveConstructible</tt>
18270 makes it necessary that such an explicit statement is given. Even the
18271 existence of the usage of <tt>std::move</tt> in the <i>Returns</i> clause doesn't
18272 help much, because this would still be well-formed for a <tt>CopyConstructible</tt>
18273 without move constructor. Let me add that the originally proposed
18274 addition reflects current practice in the standard, e.g. 25.3.9 [alg.unique]/5
18275 usages a similar terminology.
18276 </p>
18277
18278 <p>
18279 For similar wording need in case for auto_ptr see <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#973">973</a>.
18280 </p>
18281
18282 <p><i>[
18283 Howard: Moved from Tentatively Ready to Open.
18284 ]</i></p>
18285
18286 </blockquote>
18287
18288
18289
18290 <p><b>Proposed resolution:</b></p>
18291 <ul>
18292 <li>
18293 <p>
18294 Add a new Requires clause just after the prototype declaration (25.2.4 [alg.foreach]):
18295 </p>
18296 <blockquote>
18297 <p>
18298 <ins><i>Requires:</i> <tt>Function</tt> shall be <tt>MoveConstructible</tt>
18299 ( [moveconstructible]).</ins>
18300 </p>
18301 </blockquote>
18302 </li>
18303 <li>
18304 <p>
18305 Change 25.2.4 [alg.foreach]/2 as indicated:
18306 </p>
18307
18308 <blockquote>
18309 <i>Returns:</i> <ins>std::move(</ins>f<ins>)</ins>.
18310 </blockquote>
18311 </li>
18312 </ul>
18313
18314
18315
18316
18317
18318
18319
18320
18321 <hr>
18322 <h3><a name="1112"></a>1112. bitsets and new style for loop</h3>
18323 <p><b>Section:</b> 20.3.7 [template.bitset] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Future">Tentatively NAD Future</a>
18324  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-06  <b>Last modified:</b> 2009-10-26</p>
18325 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#template.bitset">active issues</a> in [template.bitset].</p>
18326 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#template.bitset">issues</a> in [template.bitset].</p>
18327 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Future">Tentatively NAD Future</a> status.</p>
18328 <p><b>Discussion:</b></p>
18329 <p>
18330 <tt>Std::bitset</tt> is a homogeneous container-like sequence of bits, yet it does
18331 not model the Range concept so cannot be used with the new for-loop syntax.
18332 It is the only such type in the library that does NOT support the new for
18333 loop.
18334 </p>
18335 <p>
18336 The obvious reason is that bitset does not support iterators.
18337 </p>
18338 <p>
18339 At least two reasonable solutions are available:
18340 </p>
18341 <ol type="i">
18342 <li>
18343 Add an iterator interface to <tt>bitset</tt>, bringing its interface close to that
18344 of <tt>std::array</tt>
18345 </li>
18346 <li>
18347 Provide an unspecified concept_map for <tt>Range&lt;bitset&gt;</tt>.
18348 </li>
18349 </ol>
18350 <p>
18351 The latter will still need some kind of iterator-like adapter for <tt>bitset</tt>,
18352 but gives implementers greater freedom on the details. E.g. begin/end return
18353 some type that simply invokes <tt>operator[]</tt> on the object it wraps, and
18354 increments its index on <tt>operator++</tt>.  A vendor can settle for <tt>InputIterator</tt>
18355 support, rather than wrapping up a full <tt>RandomAccessIterator</tt>.
18356 </p>
18357 <p>
18358 I have a mild preference for option (ii) as I think it is less work to
18359 specify at this stage of the process, although (i) is probably more useful
18360 in the long run.
18361 </p>
18362 <p>
18363 Hmm, my wording looks a little woolly, as it does not say what the element
18364 type of the range is.  Do I get a range of <tt>bool</tt>, <tt>bitset&lt;N&gt;::reference</tt>, or
18365 something else entirely?
18366 </p>
18367 <p>
18368 I guess most users will assume the behaviour of reference, but expect to
18369 work with <tt>bool</tt>.  <tt>Bool</tt> is OK for read-only traversal, but you really need to
18370 take a reference to a <tt>bitset::reference</tt> if you want to write back.
18371 </p>
18372
18373 <p><i>[
18374 Batavia (2009-05):
18375 ]</i></p>
18376
18377 <blockquote>
18378 Move to Open.
18379 We further recommend this be deferred until after the next Committee Draft.
18380 </blockquote>
18381
18382 <p><i>[
18383 2009-05-25 Alisdair adds:
18384 ]</i></p>
18385
18386
18387 <blockquote>
18388 <p>
18389 I just stumbled over the <tt>Range concept_map</tt> for <tt>valarray</tt> and this should
18390 probably set the precedent on how to write the wording.
18391 </p>
18392
18393 <p><i>[
18394 Howard: I've replaced the proposed wording with Alisdair's suggestion.
18395 ]</i></p>
18396
18397
18398 </blockquote>
18399
18400 <p><i>[
18401 2009-07-24 Daniel modifies the proposed wording for non-concepts.
18402 ]</i></p>
18403
18404
18405 <p><i>[
18406 2009-10 post-Santa Cruz:
18407 ]</i></p>
18408
18409
18410 <blockquote>
18411 Mark as Tentatively NAD Future due to the loss of concepts.
18412 </blockquote>
18413
18414
18415
18416 <p><b>Proposed resolution:</b></p>
18417 <ol>
18418 <li>
18419 <p>
18420 Modify the section 20.3.7 [template.bitset] <tt>&lt;bitset&gt;</tt> synopsis by adding
18421 the following at the end of the synopsis:
18422 </p>
18423 <blockquote><pre><ins>
18424 // XX.X.X bitset range access [bitset.range]
18425 template&lt;size_t N&gt; <i>unspecified-1</i> begin(bitset&lt;N&gt;&amp;);
18426 template&lt;size_t N&gt; <i>unspecified-2</i> begin(const bitset&lt;N&gt;&amp;);
18427 template&lt;size_t N&gt; <i>unspecified-1</i> end(bitset&lt;N&gt;&amp;);
18428 template&lt;size_t N&gt; <i>unspecified-2</i> end(const bitset&lt;N&gt;&amp;);
18429 </ins>
18430 </pre></blockquote>
18431 </li>
18432 <li>
18433 <p>
18434 Add a new section <ins>"bitset range access" [bitset.range]</ins>
18435 after the current section 20.3.7.3 [bitset.operators] with the following series of
18436 paragraphs:
18437 </p>
18438 <blockquote>
18439 <p>
18440 <ins>
18441 1.  In the <tt>begin</tt> and <tt>end</tt> function templates that follow, <i>unspecified-1</i>
18442 is a type that meets the requirements of a mutable random access
18443 iterator (24.2.5 [random.access.iterators]) whose <tt>value_type</tt> is <tt>bool</tt> and
18444 whose reference type is <tt>bitset&lt;N&gt;::reference</tt>.
18445 <i>unspecified-2</i> is a type that meets the requirements of a constant
18446 random access iterator (24.2.5 [random.access.iterators]) whose <tt>value_type</tt>
18447 is <tt>bool</tt> and whose reference type is <tt>bool</tt>.
18448 </ins>
18449 </p>
18450 <pre><ins>
18451 template&lt;size_t N&gt; <i>unspecified-1</i> begin(bitset&lt;N&gt;&amp;);
18452 template&lt;size_t N&gt; <i>unspecified-2</i> begin(const bitset&lt;N&gt;&amp;);
18453 </ins>
18454 </pre>
18455 <blockquote>
18456 <ins>2.  Returns: an iterator referencing the first bit in the bitset.</ins>
18457 </blockquote>
18458
18459 <pre><ins>
18460 template&lt;size_t N&gt; <i>unspecified-1</i> end(bitset&lt;N&gt;&amp;);
18461 template&lt;size_t N&gt; <i>unspecified-2</i> end(const bitset&lt;N&gt;&amp;);
18462 </ins></pre>
18463
18464 <blockquote>
18465 <ins>3.  Returns: an iterator referencing one past the last bit in the
18466 bitset.</ins>
18467 </blockquote>
18468 </blockquote>
18469 </li>
18470 </ol>
18471
18472
18473
18474
18475
18476
18477
18478
18479
18480
18481
18482
18483 <hr>
18484 <h3><a name="1113"></a>1113. <tt>bitset::to_string</tt> could be simplified</h3>
18485 <p><b>Section:</b> 20.3.7 [template.bitset] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
18486  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-09  <b>Last modified:</b> 2009-10-26</p>
18487 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#template.bitset">active issues</a> in [template.bitset].</p>
18488 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#template.bitset">issues</a> in [template.bitset].</p>
18489 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
18490 <p><b>Discussion:</b></p>
18491 <p>
18492 In <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a> our resolution is changing the signature by adding two
18493 defaulting arguments to 3 calls.  In principle, this means that ABI breakage
18494 is not an issue, while API is preserved.
18495 </p>
18496 <p>
18497 With that observation, it would be very nice to use the new ability to
18498 supply default template parameters to function templates to collapse all 3
18499 signatures into 1.  In that spirit, this issue offers an alternative resolution
18500 than that of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a>.
18501 </p>
18502
18503 <p><i>[
18504 Batavia (2009-05):
18505 ]</i></p>
18506
18507 <blockquote>
18508 Move to Open,
18509 and look at the issue again after <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#853">853</a> has been accepted.
18510 We further recommend this be deferred until after the next Committee Draft.
18511 </blockquote>
18512
18513 <p><i>[
18514 2009-10 post-Santa Cruz:
18515 ]</i></p>
18516
18517
18518 <blockquote>
18519 Move to Tentatively Ready.
18520 </blockquote>
18521
18522
18523
18524 <p><b>Proposed resolution:</b></p>
18525
18526 <ol type="A">
18527 <li>
18528 <p>
18529 In 20.3.7 [template.bitset]/1 (class bitset) ammend:
18530 </p>
18531 <blockquote><pre>template &lt;class charT <ins>= char</ins>,
18532             class traits <ins>= char_traits&lt;charT&gt;</ins>,
18533             class Allocator <ins>= allocator&lt;charT&gt;</ins>&gt;
18534   basic_string&lt;charT, traits, Allocator&gt;
18535   to_string(charT zero = charT('0'), charT one = charT('1')) const;
18536 <del>template &lt;class charT, class traits&gt;
18537   basic_string&lt;charT, traits, allocator&lt;charT&gt; &gt; to_string() const;
18538 template &lt;class charT&gt;
18539   basic_string&lt;charT, char_traits&lt;charT&gt;, allocator&lt;charT&gt; &gt; to_string() const;
18540 basic_string&lt;char, char_traits&lt;char&gt;, allocator&lt;char&gt; &gt; to_string() const;</del>
18541 </pre></blockquote>
18542 </li>
18543 <li>
18544 <p>
18545 In 20.3.7.2 [bitset.members] prior to p35 ammend:
18546 </p>
18547 <blockquote><pre>template &lt;class charT <ins>= char</ins>,
18548             class traits <ins>= char_traits&lt;charT&gt;</ins>,
18549             class Allocator <ins>= allocator&lt;charT&gt;</ins>&gt;
18550   basic_string&lt;charT, traits, Allocator&gt;
18551   to_string(charT zero = charT('0'), charT one = charT('1')) const;
18552 </pre></blockquote>
18553 </li>
18554 <li>
18555 Strike 20.3.7.2 [bitset.members] paragraphs 37 -&gt; 39 (including signature
18556 above 37)
18557 </li>
18558 </ol>
18559
18560
18561
18562
18563
18564
18565 <hr>
18566 <h3><a name="1114"></a>1114. Type traits underspecified</h3>
18567 <p><b>Section:</b> 20.6 [meta] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
18568  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-05-12  <b>Last modified:</b> 2009-10-26</p>
18569 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta">issues</a> in [meta].</p>
18570 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
18571 <p><b>Discussion:</b></p>
18572
18573 <p>
18574 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#975">975</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1023">1023</a>.
18575 </p>
18576
18577 <p>
18578 The current wording in 20.6.1 [meta.rqmts] is still unclear concerning
18579 it's requirements on the type traits classes regarding ambiguities.
18580 Specifically it's unclear
18581 </p>
18582
18583 <ul>
18584 <li>
18585 if a predicate trait (20.6.4 [meta.unary], 20.6.5 [meta.rel]) could derive from both
18586 <tt>true_type</tt>/<tt>false_type</tt>.
18587 </li>
18588 <li>
18589 if any of the type traits (20.6.1 [meta.rqmts], 20.6.4 [meta.unary], 20.6.5 [meta.rel]) could ambiguously derive
18590 from the same specified result type.
18591 </li>
18592 <li>
18593 if any of the type traits (20.6.1 [meta.rqmts], 20.6.4 [meta.unary], 20.6.5 [meta.rel]) could derive from other
18594 <tt>integral_constant</tt> types making the contained names ambiguous
18595 </li>
18596 <li>
18597 if any of the type traits (20.6.1 [meta.rqmts], 20.6.4 [meta.unary], 20.6.5 [meta.rel]) could have other base
18598 classes that contain members hiding the name of the result type members
18599 or make the contained member names ambiguous.
18600 </li>
18601 </ul>
18602
18603 <p><i>[
18604 Batavia (2009-05):
18605 ]</i></p>
18606
18607 <blockquote>
18608 <p>
18609 Alisdair would prefer to factor some of the repeated text,
18610 but modulo a corner case or two,
18611 he believes the proposed wording is otherwise substantially correct.
18612 </p>
18613 <p>
18614 Move to Open.
18615 </p>
18616 </blockquote>
18617
18618 <p><i>[
18619 2009-10 post-Santa Cruz:
18620 ]</i></p>
18621
18622
18623 <blockquote>
18624 Move to Tentatively Ready.
18625 </blockquote>
18626
18627
18628
18629 <p><b>Proposed resolution:</b></p>
18630 <p><i>[
18631 The usage of the notion of a <i>BaseCharacteristic</i> below
18632 might be
18633 useful in other places - e.g. to define the base class relation in
18634 20.7.5 [refwrap], 20.7.14 [func.memfn], or 20.7.15.2 [func.wrap.func].
18635 In this case it's definition should probably
18636 be moved to Clause 17
18637 ]</i></p>
18638
18639
18640 <ol>
18641 <li>
18642 <p>
18643 Change 20.6.1 [meta.rqmts]/1 as indicated:
18644 </p>
18645 <blockquote>
18646 [..] It shall be <tt>DefaultConstructible</tt>, <tt>CopyConstructible</tt>, and publicly
18647 <ins>and unambiguously</ins> derived, directly or indirectly, from
18648 <ins>its <i>BaseCharacteristic</i>, which is</ins> a specialization of the
18649 template <tt>integral_constant</tt> (20.6.3), with the arguments to the template
18650 <tt>integral_constant</tt> determined by the requirements for the particular
18651 property being described. <ins>The member names of the
18652 <i>BaseCharacteristic</i> shall be unhidden and unambiguously
18653 available in the <i>UnaryTypeTrait</i>.</ins>
18654 </blockquote>
18655 </li>
18656 <li>
18657 <p>
18658 Change 20.6.1 [meta.rqmts]/2 as indicated:
18659 </p>
18660 <blockquote>
18661 [..] It shall be <tt>DefaultConstructible</tt>, <tt>CopyConstructible</tt>, and publicly
18662 <ins>and unambiguously</ins> derived, directly or indirectly, from
18663 <del>an instance</del> <ins>its <i>BaseCharacteristic</i>, which is a
18664 specialization</ins> of the template <tt>integral_constant</tt> (20.6.3), with
18665 the arguments to the template <tt>integral_constant</tt> determined by the
18666 requirements for the particular relationship being described. <ins>The
18667 member names of the <i>BaseCharacteristic</i> shall be unhidden
18668 and unambiguously available in the <i>BinaryTypeTrait</i>.</ins>
18669 </blockquote>
18670 </li>
18671 <li>
18672 <p>
18673 Change 20.6.4 [meta.unary]/2 as indicated:
18674 </p>
18675 <blockquote>
18676 Each of these templates shall be a <i>UnaryTypeTrait</i> (20.6.1),
18677 <del>publicly derived directly or indirectly from <tt>true_type</tt> if the
18678 corresponding condition is true, otherwise from <tt>false_type</tt></del>
18679 <ins>where its <i>BaseCharacteristic</i> shall be <tt>true_type</tt> if the
18680 corresponding condition is true, otherwise <tt>false_type</tt></ins>.
18681 </blockquote>
18682 </li>
18683 <li>
18684 <p>
18685 Change 20.6.5 [meta.rel]/2 as indicated:
18686 </p>
18687
18688 <blockquote>
18689 Each of these templates shall be a <i>BinaryTypeTrait</i> (20.6.1),
18690 <del>publicly derived directly or indirectly from <tt>true_type</tt> if the
18691 corresponding condition is true, otherwise from <tt>false_type</tt></del>
18692 <ins>where its <i>BaseCharacteristic</i> shall be <tt>true_type</tt> if the
18693 corresponding condition is true, otherwise <tt>false_type</tt></ins>.
18694 </blockquote>
18695 </li>
18696 </ol>
18697
18698
18699
18700
18701
18702
18703 <hr>
18704 <h3><a name="1115"></a>1115. <tt>va_copy</tt> missing from Standard macros table</h3>
18705 <p><b>Section:</b> C.2 [diff.library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Editorial">Tentatively NAD Editorial</a>
18706  <b>Submitter:</b> Miles Zhao <b>Opened:</b> 2009-05-23  <b>Last modified:</b> 2009-10-26</p>
18707 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#diff.library">issues</a> in [diff.library].</p>
18708 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Editorial">Tentatively NAD Editorial</a> status.</p>
18709 <p><b>Discussion:</b></p>
18710 <p>
18711 In "Table 122 -- Standard macros" of C.2 [diff.library], which lists the 56 macros
18712 inherited from C library, <tt>va_copy</tt> seems to be missing. But in
18713 "Table 21 -- Header <tt>&lt;cstdarg&gt;</tt> synopsis" (18.10 [support.runtime]), there is.
18714 </p>
18715
18716 <p><i>[
18717 2009-10 post-Santa Cruz:
18718 ]</i></p>
18719
18720
18721 <blockquote>
18722 Mark as Tentatively NAD Editorial, if Pete disagrees, Howard
18723 will move to Tentatively Ready
18724 </blockquote>
18725
18726
18727
18728 <p><b>Proposed resolution:</b></p>
18729 <p>
18730 Add <tt>va_copy</tt> to Table 122 -- Standard macros in C.2 [diff.library].
18731 </p>
18732
18733
18734
18735
18736
18737 <hr>
18738 <h3><a name="1118"></a>1118. tuple query APIs do not support cv-qualification</h3>
18739 <p><b>Section:</b> 20.5.2.5 [tuple.helper] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
18740  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-23  <b>Last modified:</b> 2009-10-26</p>
18741 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#tuple.helper">active issues</a> in [tuple.helper].</p>
18742 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#tuple.helper">issues</a> in [tuple.helper].</p>
18743 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
18744 <p><b>Discussion:</b></p>
18745 <p>
18746 The APIs <tt>tuple_size</tt> and <tt>tuple_element</tt> do not support
18747 cv-qualified <tt>tuple</tt>s, <tt>pair</tt>s or <tt>array</tt>s.
18748 </p>
18749 <p>
18750 The most generic solution would be to supply partial specializations once
18751 for each cv-type in the <tt>tuple</tt> header.  However, requiring this header for
18752 cv-qualified <tt>pair</tt>s/<tt>array</tt>s seems unhelpful.  The BSI editorial
18753 suggestion (UK-198/US-69,
18754 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2533.html">N2533</a>)
18755 to merge <tt>tuple</tt> into <tt>&lt;utility&gt;</tt> would help with <tt>pair</tt>,
18756 but not <tt>array</tt>.  That might be resolved by making a dependency between the
18757 <tt>&lt;array&gt;</tt> header and <tt>&lt;utility&gt;</tt>, or simply recognising
18758 the dependency be fulfilled in a Remark.
18759 </p>
18760
18761 <p><i>[
18762 2009-05-24 Daniel adds:
18763 ]</i></p>
18764
18765
18766 <blockquote>
18767 <p>
18768 All <tt>tuple_size</tt> templates with a base class need to derive publicly, e.g.
18769 </p>
18770
18771 <blockquote><pre>template &lt;IdentityOf T&gt; class tuple_size&lt; const T &gt; :
18772    <ins>public</ins> tuple_size&lt;T&gt; {};
18773 </pre></blockquote>
18774
18775 <p>
18776 The same applies to the tuple_element class hierarchies.
18777 </p>
18778 <p>
18779 What is actually meant with the comment
18780 </p>
18781 <blockquote>
18782 this solution relies on 'metafunction forwarding' to inherit the
18783 nested typename type
18784 </blockquote>
18785 <p>
18786 ?
18787 </p>
18788 <p>
18789 I ask, because all base classes are currently unconstrained and their
18790 instantiation is invalid in the constrained context of the <tt>tuple_element</tt> partial
18791 template specializations.
18792 </p>
18793 </blockquote>
18794
18795 <p><i>[
18796 2009-05-24 Alisdair adds:
18797 ]</i></p>
18798
18799
18800 <blockquote>
18801 <p>
18802 I think a better solution might be to ask Pete editorially to change all
18803 declarations of tupling APIs to use the struct specifier instead of class.
18804 </p>
18805 <p>
18806 "metafunction forwarding" refers to the MPL metafunction protocol, where a
18807 metafunction result is declared as a nested typedef with the name "type",
18808 allowing metafunctions to be chained by means of inheritance.  It is a
18809 neater syntax than repeatedly declaring a typedef, and inheritance syntax is
18810 slightly nicer when it comes to additional typename keywords.
18811 </p>
18812 <p>
18813 The constrained template with an unconstrained base is a good observation
18814 though.
18815 </p>
18816 </blockquote>
18817
18818 <p><i>[
18819 2009-10 post-Santa Cruz:
18820 ]</i></p>
18821
18822
18823 <blockquote>
18824 Move to Open, Alisdair to provide wording. Once wording is
18825 provided, Howard will move to Review.
18826 </blockquote>
18827
18828
18829
18830 <p><b>Proposed resolution:</b></p>
18831 <p>
18832 Add to 20.5.1 [tuple.general] p2 (Header <tt>&lt;tuple&gt;</tt> synopsis)
18833 </p>
18834
18835 <blockquote><pre>// 20.5.2.3, tuple helper classes:
18836 template &lt;IdentityOf T&gt; class tuple_size; // undefined
18837 <ins>template &lt;IdentityOf T&gt; class tuple_size&lt; const T &gt; : tuple_size&lt;T&gt; {};</ins>
18838 <ins>template &lt;IdentityOf T&gt; class tuple_size&lt; volatile T &gt; : tuple_size&lt;T&gt; {};</ins>
18839 <ins>template &lt;IdentityOf T&gt; class tuple_size&lt; const volatile T &gt; : tuple_size&lt;T&gt; {};</ins>
18840
18841 template &lt;VariableType... Types&gt; class tuple_size&lt;tuple&lt;Types...&gt; &gt;;
18842
18843 template &lt;size_t I, IdentityOf T&gt; class tuple_element; // undefined
18844 <ins>template &lt;size_t I, IdentityOf T&gt; class tuple_element&lt;I, const T&gt;;</ins>
18845 <ins>template &lt;size_t I, IdentityOf T&gt; class tuple_element&lt;I, volatile T&gt;;</ins>
18846 <ins>template &lt;size_t I, IdentityOf T&gt; class tuple_element&lt;I, const volatile T&gt;;</ins>
18847
18848 template &lt;size_t I, VariableType... Types&gt;
18849   requires True&lt;(I &lt; sizeof...(Types))&gt; class tuple_element&lt;I, tuple&lt;Types...&gt; &gt;;
18850 </pre></blockquote>
18851
18852 <p>
18853 Add to 20.5.2.5 [tuple.helper]
18854 </p>
18855 <p><i>[
18856 (note that this solution relies on 'metafunction forwarding' to inherit the
18857 nested typename type)
18858 ]</i></p>
18859
18860
18861 <blockquote><pre>template &lt;class... Types&gt;
18862 class tuple_size&lt;tuple&lt;Types...&gt; &gt;
18863   : public integral_constant&lt;size_t, sizeof...(Types)&gt; { };
18864
18865 template &lt;size_t I, class... Types&gt;
18866 requires True&lt;(I &lt; sizeof...(Types))&gt;
18867 class tuple_element&lt;I, tuple&lt;Types...&gt; &gt; {
18868 public:
18869   typedef TI type;
18870 };
18871
18872 <ins>template &lt;size_t I, IdentityOf T&gt;
18873   class tuple_element&lt;I, const T&gt; : add_const&lt;tuple_element&lt;I,T&gt;&gt; {};</ins>
18874
18875 <ins>template &lt;size_t I, IdentityOf T&gt;
18876   class tuple_element&lt;I, volatile T&gt; : add_volatile&lt;tuple_element&lt;I,T&gt;&gt; {};</ins>
18877
18878 <ins>template &lt;size_t I, IdentityOf T&gt;
18879   class tuple_element&lt;I, const volatile T&gt; : add_cv&lt;tuple_element&lt;I,T&gt;&gt; {};</ins>
18880 </pre></blockquote>
18881
18882
18883
18884
18885
18886 <hr>
18887 <h3><a name="1119"></a>1119. tuple query APIs do not support references</h3>
18888 <p><b>Section:</b> 20.5.2.5 [tuple.helper] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
18889  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-23  <b>Last modified:</b> 2009-10-26</p>
18890 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#tuple.helper">active issues</a> in [tuple.helper].</p>
18891 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#tuple.helper">issues</a> in [tuple.helper].</p>
18892 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
18893 <p><b>Discussion:</b></p>
18894 <p>
18895 The <tt>tuple</tt> query APIs <tt>tuple_size</tt> and
18896 <tt>tuple_element</tt> do not support references-to-tuples.  This can be
18897 annoying when a template deduced a parameter type to be a reference,
18898 which must be explicitly stripped with <tt>remove_reference</tt> before calling
18899 these APIs.
18900 </p>
18901 <p>
18902 I am not proposing a resolution at this point, as there is a
18903 combinatorial explosion with lvalue/rvalue references and
18904 cv-qualification (see previous issue) that suggests some higher
18905 refactoring is in order.  This might be something to kick back over to
18906 Core/Evolution.
18907 </p>
18908 <p>
18909 Note that we have the same problem in numeric_limits.
18910 </p>
18911
18912 <p><i>[
18913 2009-10 post-Santa Cruz:
18914 ]</i></p>
18915
18916
18917 <blockquote>
18918 Move to Open. Alisdair to provide wording.
18919 </blockquote>
18920
18921
18922
18923 <p><b>Proposed resolution:</b></p>
18924
18925
18926
18927
18928
18929 <hr>
18930 <h3><a name="1121"></a>1121. Support for multiple arguments</h3>
18931 <p><b>Section:</b> 20.4.2 [ratio.arithmetic] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Future">Tentatively NAD Future</a>
18932  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-25  <b>Last modified:</b> 2009-11-02</p>
18933 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#ratio.arithmetic">issues</a> in [ratio.arithmetic].</p>
18934 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Future">Tentatively NAD Future</a> status.</p>
18935 <p><b>Discussion:</b></p>
18936 <p>
18937 Both add and multiply could sensibly be called with more than two arguments.
18938 The variadic template facility makes such declarations simple, and is likely
18939 to be frequently wrapped by end users if we do not supply the variant
18940 ourselves.
18941 </p>
18942 <p>
18943 We deliberately ignore divide at this point as it is not transitive.
18944 Likewise, subtract places special meaning on the first argument so I do not
18945 suggest extending that immediately.  Both could be supported with analogous
18946 wording to that for add/multiply below.
18947 </p>
18948 <p>
18949 Note that the proposed resolution is potentially incompatible with that
18950 proposed for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a>, although the addition of the typedef to ratio would be
18951 equally useful.
18952 </p>
18953
18954 <p><i>[
18955 2009-10-30 Alisdair adds:
18956 ]</i></p>
18957
18958
18959 <blockquote>
18960 <p>
18961 The consensus of the group when we reviewed this in Santa Cruz was that
18962 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#921">921</a> would proceed to Ready as planned, and the
18963 multi-paramater add/multiply templates should be renamed as
18964 <tt>ratio_sum</tt> and <tt>ratio_product</tt> to avoid the problem
18965 mixing template aliases with partial specializations.
18966 </p>
18967
18968 <p>
18969 It was also suggested to close this issue as NAD Future as it does not
18970 correspond directly to any NB comment.  NBs are free to submit a
18971 specific comment (and re-open) in CD2 though.
18972 </p>
18973
18974 <p>
18975 Walter Brown also had concerns on better directing the order of
18976 evaluation to avoid overflows if we do proceed for 0x rather than TR1,
18977 so wording may not be complete yet.
18978 </p>
18979
18980 <p><i>[
18981 Alisdair updates wording.
18982 ]</i></p>
18983
18984
18985 </blockquote>
18986
18987 <p><i>[
18988 2009-10-30 Howard:
18989 ]</i></p>
18990
18991
18992 <blockquote>
18993 Moved to Tentatively NAD Future after 5 positive votes on c++std-lib.
18994 </blockquote>
18995
18996
18997
18998 <p><b>Proposed resolution:</b></p>
18999
19000 <p>
19001 Add the following type traits to p3 20.4 [ratio]
19002 </p>
19003
19004 <blockquote><pre>// ratio arithmetic
19005 template &lt;class R1, class R2&gt; struct ratio_add;
19006 template &lt;class R1, class R2&gt; struct ratio_subtract;
19007 template &lt;class R1, class R2&gt; struct ratio_multiply;
19008 template &lt;class R1, class R2&gt; struct ratio_divide;
19009 <ins>template &lt;class R1, class ... RList&gt; struct ratio_sum;</ins>
19010 <ins>template &lt;class R1, class ... RList&gt; struct ratio_product;</ins>
19011 </pre></blockquote>
19012
19013 <p>
19014 after 20.4.2 [ratio.arithmetic] p1: add
19015 </p>
19016
19017 <blockquote><pre>template &lt;class R1, class ... RList&gt; struct ratio_sum; // declared, never defined
19018
19019 template &lt;class R1&gt; struct ratio_sum&lt;R1&gt; : R1 {};
19020 </pre>
19021
19022 <blockquote>
19023 <i>Requires:</i> <tt>R1</tt> is a specialization of class template <tt>ratio</tt>
19024 </blockquote>
19025
19026 <pre>template &lt;class R1, class R2, class ... RList&gt;
19027  struct ratio_sum&lt;R1, R2, RList...&gt;
19028    : ratio_add&lt; R1, ratio_sum&lt;R2, RList...&gt;&gt; {
19029 };
19030 </pre>
19031
19032 <blockquote>
19033 <i>Requires:</i> <tt>R1</tt> and each element in parmater pack
19034 <tt>RList</tt> is a specialization of class template <tt>ratio</tt>
19035 </blockquote>
19036 </blockquote>
19037
19038 <p>
19039 after 20.4.2 [ratio.arithmetic] p3: add
19040 </p>
19041
19042 <blockquote><pre>template &lt;class R1, class ... RList&gt; struct ratio_product; // declared, never defined
19043
19044 template &lt;class R1&gt; struct ratio_product&lt;R1&gt; : R1 {};
19045 </pre>
19046
19047 <blockquote>
19048 <i>Requires:</i> <tt>R1</tt> is a specialization of class template <tt>ratio</tt>
19049 </blockquote>
19050
19051 <pre>template &lt;class R1, class R2, class ... RList&gt;
19052  struct ratio_sum&lt;R1, R2, RList...&gt;
19053    : ratio_add&lt; R1, ratio_product&lt;R2, RList...&gt;&gt; {
19054 };
19055 </pre>
19056
19057 <blockquote>
19058 <i>Requires:</i> <tt>R1</tt> and each element in parmater pack
19059 <tt>RList</tt> is a specialization of class template <tt>ratio</tt>
19060 </blockquote>
19061 </blockquote>
19062
19063
19064
19065
19066
19067
19068
19069
19070 <hr>
19071 <h3><a name="1123"></a>1123. no requirement that standard streams be flushed</h3>
19072 <p><b>Section:</b> 27.5.2.1.6 [ios::Init] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
19073  <b>Submitter:</b> James Kanze <b>Opened:</b> 2009-05-14  <b>Last modified:</b> 2009-10-20</p>
19074 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#ios::Init">issues</a> in [ios::Init].</p>
19075 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
19076 <p><b>Discussion:</b></p>
19077 <p>
19078 As currently formulated, the standard doesn't require that there
19079 is ever a flush of <tt>cout</tt>, etc.  (This implies, for example, that
19080 the classical hello, world program may have no output.)  In the
19081 current draft
19082 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2798.pdf">N2798</a>),
19083 there is a requirement that the objects
19084 be constructed before <tt>main</tt>, and before the dynamic
19085 initialization of any non-local objects defined after the
19086 inclusion of <tt>&lt;iostream&gt;</tt> in the same translation unit.  The only
19087 requirement that I can find concerning flushing, however, is in
19088 27.5.2.1.6 [ios::Init], where the destructor of the last
19089 <tt>std::ios_base::Init</tt> object flushes.  But there is, as far as I
19090 can see, no guarantee that such an object ever exists.
19091 </p>
19092 <p>
19093 Also, the wording in  [iostreams.objects] says that:
19094 </p>
19095 <blockquote>
19096 The objects
19097 are constructed and the associations are established at some
19098 time prior to or during the first time an object of class
19099 <tt>ios_base::Init</tt> is constructed, and in any case before the body
19100 of main begins execution.
19101 </blockquote>
19102 <p>
19103 In 27.5.2.1.6 [ios::Init], however, as an
19104 effect of the constructor, it says that
19105 </p>
19106 <blockquote>
19107 If <tt>init_cnt</tt> is zero,
19108 the function stores the value one in <tt>init_cnt</tt>, then constructs
19109 and initializes the objects <tt>cin</tt>, <tt>cout</tt>, <tt>cerr</tt>, <tt>clog</tt>
19110 <tt>wcin</tt>, <tt>wcout</tt>, <tt>wcerr</tt>, and <tt>wclog</tt>"
19111 </blockquote>
19112
19113 <p>
19114 which seems to forbid earlier
19115 construction.
19116 </p>
19117
19118 <p>
19119 (Note that with these changes, the exposition only "<tt>static
19120 int init_cnt</tt>" in <tt>ios_base::Init</tt> can be dropped.)
19121 </p>
19122 <p>
19123 Of course, a determined programmer can still inhibit the
19124 flush with things like:
19125 </p>
19126 <blockquote><pre>new std::ios_base::Init ;       //  never deleted
19127 </pre></blockquote>
19128 <p>
19129 or (in a function):
19130 </p>
19131 <blockquote><pre>std::ios_base::Init ensureConstruction ;
19132 //  ...
19133 exit( EXIT_SUCCESS ) ;
19134 </pre></blockquote>
19135 <p>
19136 Perhaps some words somewhere to the effect that all
19137 <tt>std::ios_base::Init</tt> objects should have static lifetime
19138 would be in order.
19139 </p>
19140
19141 <p><i>[
19142 2009 Santa Cruz:
19143 ]</i></p>
19144
19145
19146 <blockquote>
19147 Moved to Ready.  Some editorial changes are expected (in addition to the
19148 proposed wording) to remove <tt>init_cnt</tt> from <tt>Init</tt>.
19149 </blockquote>
19150
19151
19152
19153 <p><b>Proposed resolution:</b></p>
19154 <p>
19155 Change 27.4 [iostream.objects]/2:
19156 </p>
19157
19158 <blockquote>
19159 -2- The objects are constructed and the associations are established at
19160 some time prior to or during the first time an object of class
19161 <tt>ios_base::Init</tt> is constructed, and in any case before the body
19162 of main begins execution.<sup>292</sup> The objects are not destroyed
19163 during program execution.<sup>293</sup>
19164 <del>If a translation unit includes
19165 <tt>&lt;iostream&gt;</tt> or explicitly constructs an
19166 <tt>ios_base::Init</tt> object, these stream objects shall be
19167 constructed before dynamic initialization of non-local objects defined
19168 later in that translation unit.</del>
19169 <ins>The results of including <tt>&lt;iostream&gt;</tt> in a translation
19170 unit shall be as if <tt>&lt;iostream&gt;</tt> defined an instance of
19171 <tt>ios_base::Init</tt> with static lifetime. Similarly, the entire
19172 program shall behave as if there were at least one instance of
19173 <tt>ios_base::Init</tt> with static lifetime.</ins>
19174 </blockquote>
19175
19176 <p>
19177 Change 27.5.2.1.6 [ios::Init]/3:
19178 </p>
19179
19180 <blockquote>
19181 <pre>Init();
19182 </pre>
19183 <blockquote>
19184 -3- <i>Effects:</i> Constructs an object of class <tt>Init</tt>.
19185 <del>If <tt>init_cnt</tt> is zero, the function stores the value one in
19186 <tt>init_cnt</tt>, then constructs and initializes the objects
19187 <tt>cin</tt>, <tt>cout</tt>, <tt>cerr</tt>, <tt>clog</tt> (27.4.1),
19188 <tt>wcin</tt>, <tt>wcout</tt>, <tt>wcerr</tt>, and <tt>wclog</tt>
19189 (27.4.2). In any case, the function then adds one to the value stored in
19190 <tt>init_cnt</tt>.</del>
19191 <ins>Constructs and initializes the objects <tt>cin</tt>, <tt>cout</tt>,
19192 <tt>cerr</tt>, <tt>clog</tt>, <tt>wcin</tt>, <tt>wcout</tt>,
19193 <tt>wcerr</tt> and <tt>wclog</tt> if they have not already been
19194 constructed and initialized.</ins>
19195 </blockquote>
19196 </blockquote>
19197
19198 <p>
19199 Change 27.5.2.1.6 [ios::Init]/4:
19200 </p>
19201
19202 <blockquote>
19203 <pre>~Init();
19204 </pre>
19205 <blockquote>
19206 -4- <i>Effects:</i> Destroys an object of class <tt>Init</tt>.
19207 <del>The function subtracts one from the value stored in <tt>init_cnt</tt> and,
19208 if the resulting stored value is one,</del>
19209 <ins>If there are no other instances of the class still in
19210 existance,</ins>
19211 calls <tt>cout.flush()</tt>,
19212 <tt>cerr.flush()</tt>, <tt>clog.flush()</tt>, <tt>wcout.flush()</tt>,
19213 <tt>wcerr.flush()</tt>, <tt>wclog.flush()</tt>.
19214 </blockquote>
19215 </blockquote>
19216
19217
19218
19219
19220
19221
19222 <hr>
19223 <h3><a name="1125"></a>1125. ostream_iterator does not work with movable types</h3>
19224 <p><b>Section:</b> 24.6.2.2 [ostream.iterator.ops] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
19225  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-28  <b>Last modified:</b> 2009-05-30</p>
19226 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
19227 <p><b>Discussion:</b></p>
19228 <p>
19229 <tt>ostream_iterator</tt> has not been updated to support moveable types, in a
19230 similar manner to the insert iterators.
19231 Note that this is not a problem for <tt>ostreambuf_iterator</tt>, as the types it is
19232 restricted to dealing with do not support extra-efficient moving.
19233 </p>
19234
19235
19236 <p><b>Proposed resolution:</b></p>
19237 <p>
19238 Add second <tt>operator=</tt> overload to class <tt>template ostream_iterator</tt>
19239 in 24.6.2 [ostream.iterator], para 2:
19240 </p>
19241
19242 <blockquote><pre>ostream_iterator&lt;T,charT,traits&gt;&amp; operator=(const T&amp; value);
19243 <ins>ostream_iterator&lt;T,charT,traits&gt;&amp; operator=(T&amp;&amp; value);</ins>
19244 </pre></blockquote>
19245
19246 <p>
19247 Add a new paragraph: in 24.6.2.2 [ostream.iterator.ops]:
19248 </p>
19249
19250 <blockquote>
19251 <pre>ostream_iterator&amp; operator=(T&amp;&amp; value);
19252 </pre>
19253 <blockquote>
19254 <p>
19255 -2- <i>Effects:</i>
19256 </p>
19257 <blockquote><pre>*out_stream &lt;&lt; std::move(value);
19258 if(delim != 0)
19259   *out_stream &lt;&lt; delim;
19260 return (*this);
19261 </pre></blockquote>
19262 </blockquote>
19263 </blockquote>
19264
19265
19266
19267
19268
19269
19270 <hr>
19271 <h3><a name="1126"></a>1126. <tt>istreambuff_iterator::equal</tt> needs a const &amp; parameter</h3>
19272 <p><b>Section:</b> 24.6.3.5 [istreambuf.iterator::equal] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
19273  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-28  <b>Last modified:</b> 2009-11-02</p>
19274 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#istreambuf.iterator::equal">issues</a> in [istreambuf.iterator::equal].</p>
19275 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
19276 <p><b>Discussion:</b></p>
19277 <p>
19278 The <tt>equal</tt> member function of <tt>istreambuf_iterator</tt> is
19279 declared <tt>const</tt>, but takes its argument by non-const reference.
19280 </p>
19281 <p>
19282 This is not compatible with the <tt>operator==</tt> free function overload, which is
19283 defined in terms of calling <tt>equal</tt> yet takes both arguments by reference to
19284 const.
19285 </p>
19286
19287 <p><i>[
19288 The proposed wording is consistent with <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#110">110</a> with status TC1.
19289 ]</i></p>
19290
19291
19292 <p><i>[
19293 2009-11-02 Howard adds:
19294 ]</i></p>
19295
19296
19297 <blockquote>
19298 Set to Tentatively Ready after 5 positive votes on c++std-lib.
19299 </blockquote>
19300
19301
19302
19303 <p><b>Proposed resolution:</b></p>
19304 <p>
19305 Ammend in both:<br>
19306 24.6.3 [istreambuf.iterator]<br>
19307 24.6.3.5 [istreambuf.iterator::equal]<br>
19308 </p>
19309
19310 <blockquote><pre>bool equal(<ins>const </ins>istreambuf_iterator&amp; b) const;
19311 </pre></blockquote>
19312
19313
19314
19315
19316
19317
19318 <hr>
19319 <h3><a name="1130"></a>1130. <tt>copy_exception</tt> name misleading</h3>
19320 <p><b>Section:</b> 18.8.5 [propagation] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
19321  <b>Submitter:</b> Peter Dimov <b>Opened:</b> 2009-05-13  <b>Last modified:</b> 2009-10-20</p>
19322 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#propagation">active issues</a> in [propagation].</p>
19323 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#propagation">issues</a> in [propagation].</p>
19324 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
19325 <p><b>Discussion:</b></p>
19326 <p>
19327 The naming of <tt>std::copy_exception</tt> misleads almost everyone
19328 (experts included!) to think that it is the function that copies an
19329 <tt>exception_ptr</tt>:
19330 </p>
19331
19332 <blockquote><pre>exception_ptr p1 = current_exception();
19333 exception_ptr p2 = copy_exception( p1 );
19334 </pre></blockquote>
19335
19336 <p>
19337 But this is not at all what it does. The above actually creates an
19338 <tt>exception_ptr p2</tt> that contains a copy of <tt>p1</tt>, not of
19339 the exception to which <tt>p1</tt> refers!
19340 </p>
19341 <p>
19342 This is, of course, all my fault; in my defence, I used <tt>copy_exception</tt>
19343 because I was unable to think of a better name.
19344 </p>
19345 <p>
19346 But I believe that, based on what we've seen so far, <em>any</em> other name would be better.
19347 </p>
19348 <p>
19349 Therefore, I propose <tt>copy_exception</tt> to be renamed to
19350 <tt>create_exception</tt>:
19351 </p>
19352
19353 <blockquote><pre>template&lt;class E&gt; exception_ptr create_exception(E e);
19354 </pre></blockquote>
19355
19356 <p>
19357 with the following explanatory paragraph after it:
19358 </p>
19359
19360 <blockquote>
19361 Creates an <tt>exception_ptr</tt> that refers to a copy of <tt>e</tt>.
19362 </blockquote>
19363
19364 <p><i>[
19365 2009-05-13 Daniel adds:
19366 ]</i></p>
19367
19368
19369 <blockquote>
19370 <p>
19371 What about
19372 </p>
19373 <blockquote><pre>make_exception_ptr
19374 </pre></blockquote>
19375 <p>
19376 in similarity to <tt>make_pair</tt> and <tt>make_tuple</tt>, <tt>make_error_code</tt> and
19377 <tt>make_error_condition</tt>, or <tt>make_shared</tt>? Or, if a stronger symmetry to
19378 <tt>current_exception</tt> is preferred:
19379 </p>
19380
19381 <blockquote><pre>make_exception
19382 </pre></blockquote>
19383 <p>
19384 We have not a single <tt>create_*</tt> function in the library, it was always
19385 <tt>make_*</tt> used.
19386 </p>
19387 </blockquote>
19388
19389 <p><i>[
19390 2009-05-13 Peter adds:
19391 ]</i></p>
19392
19393
19394 <blockquote>
19395 <tt>make_exception_ptr</tt> works for me.
19396 </blockquote>
19397
19398 <p><i>[
19399 2009-06-02 Thomas J. Gritzan adds:
19400 ]</i></p>
19401
19402
19403 <blockquote>
19404 <p>
19405 To avoid surprises and unwanted recursion, how about making a call to
19406 <tt>std::make_exception_ptr</tt> with an <tt>exception_ptr</tt> illegal?
19407 </p>
19408 <p>
19409 It might work like this:
19410 </p>
19411 <blockquote><pre>template&lt;class E&gt;
19412 exception_ptr make_exception_ptr(E e);
19413 template&lt;&gt;
19414 exception_ptr make_exception_ptr&lt;exception_ptr&gt;(exception_ptr e) = delete;
19415 </pre></blockquote>
19416 </blockquote>
19417
19418 <p><i>[
19419 2009 Santa Cruz:
19420 ]</i></p>
19421
19422
19423 <blockquote>
19424 Move to Review for the time being. The subgroup thinks this is a good
19425 idea, but doesn't want to break compatibility unnecessarily if someone
19426 is already shipping this. Let's talk to Benjamin and PJP tomorrow to
19427 make sure neither objects.
19428 </blockquote>
19429
19430
19431
19432 <p><b>Proposed resolution:</b></p>
19433 <p>
19434 Change 18.8.5 [propagation]:
19435 </p>
19436
19437 <blockquote>
19438 <pre>template&lt;class E&gt; exception_ptr <del>copy_exception</del><ins>make_exception_ptr</ins>(E e);
19439 </pre>
19440
19441 <blockquote>
19442 <p>
19443 -11- <i>Effects:</i> <ins>Creates an <tt>exception_ptr</tt> that refers
19444 to a copy of <tt>e</tt>,</ins> as if
19445 </p>
19446
19447 <blockquote><pre>try {
19448   throw e;
19449 } catch(...) {
19450   return current_exception();
19451 }
19452 </pre></blockquote>
19453
19454 <p>...</p>
19455 </blockquote>
19456
19457 </blockquote>
19458
19459
19460
19461
19462
19463
19464 <hr>
19465 <h3><a name="1131"></a>1131. C++0x does not need <tt>alignment_of</tt></h3>
19466 <p><b>Section:</b> 20.6.4.3 [meta.unary.prop] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
19467  <b>Submitter:</b> Niels Dekker <b>Opened:</b> 2009-06-01  <b>Last modified:</b> 2009-06-02</p>
19468 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#meta.unary.prop">active issues</a> in [meta.unary.prop].</p>
19469 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.unary.prop">issues</a> in [meta.unary.prop].</p>
19470 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
19471 <p><b>Discussion:</b></p>
19472 <p>
19473 The <tt>alignment_of</tt> template is no longer necessary, now that the
19474 core language will provide <tt>alignof</tt>. Scott Meyers raised this
19475 issue at comp.std.c++,
19476 <a href="http://groups.google.com/group/comp.std.c++/browse_thread/thread/9b020306e803f08a">C++0x: alignof vs. alignment_of</a>,
19477 May 21, 2009.  In a reply, Daniel Krügler pointed out that
19478 <tt>alignof</tt> was added to the working paper <i>after</i>
19479 <tt>alignment_of</tt>. So it appears that <tt>alignment_of</tt> is only
19480 part of the current Working Draft
19481 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2857.pdf">N2857</a>)
19482 because it is in TR1.
19483 </p>
19484 <p>
19485 Having both <tt>alignof</tt> and <tt>alignment_of</tt> would cause
19486 unwanted confusion. In general, I think TR1 functionality should not be
19487 brought into C++0x if it is entirely redundant with other C++0x language
19488 or library features.
19489 </p>
19490
19491
19492 <p><b>Proposed resolution:</b></p>
19493 <p>
19494 Remove from Header &lt;type_traits&gt; synopsis 20.6.2 [meta.type.synop]:
19495 </p>
19496 <blockquote><pre><del>template &lt;class T&gt; struct alignment_of;</del>
19497 </pre></blockquote>
19498
19499 <p>
19500 Remove the first row of Table 34 ("Type property queries"), from
19501 Type properties 20.6.4.3 [meta.unary.prop]:
19502 </p>
19503 <blockquote>
19504 <table border="1">
19505 <caption>Table 34 -- Type property queries</caption>
19506 <tbody><tr>
19507 <td><del><tt>template &lt;class T&gt; struct alignment_of;</tt></del></td>
19508 <td><del><tt>alignof(T)</tt>.</del><br>
19509 <del><i>Precondition:</i> <tt>T</tt> shall be a complete type, a reference
19510 type, or an array of unknown bound, but shall not be a function type or
19511 (possibly cv-qualified) <tt>void</tt>.</del>
19512 </td>
19513 </tr>
19514 </tbody></table>
19515 </blockquote>
19516
19517 <p>
19518 Change text in Table 41 ("Other transformations"), from Other
19519 transformations 20.6.7 [meta.trans.other], as follows:
19520 </p>
19521 <blockquote>
19522 <table border="1">
19523 <caption>Table 41 -- Other transformations</caption>
19524 <tbody><tr><td>...</td>
19525 <td>
19526  Align shall be equal to <tt>
19527  <del>alignment_of&lt;T&gt;::value</del>
19528  <ins>alignof(T)</ins>
19529  </tt> for some type <tt>T</tt> or to <i>default-alignment</i>.
19530 </td>
19531 <td>...</td>
19532 </tr></tbody></table>
19533 </blockquote>
19534
19535
19536
19537
19538
19539 <hr>
19540 <h3><a name="1133"></a>1133. Does N2844 break current specification of list::splice?</h3>
19541 <p><b>Section:</b> 23.3.3.5 [forwardlist.ops], 23.3.4.4 [list.ops] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
19542  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-05-09  <b>Last modified:</b> 2009-10-27</p>
19543 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#forwardlist.ops">issues</a> in [forwardlist.ops].</p>
19544 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
19545 <p><b>Discussion:</b></p>
19546 <p>
19547 IIUC,
19548 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2844.html">N2844</a>
19549 means that lvalues will no longer bind to rvalue references.
19550 Therefore, the current specification of <tt>list::splice</tt> (list
19551 operations 23.3.4.4 [list.ops]) will be a breaking change of behaviour for existing
19552 programs.  That is because we changed the signature to swallow via an rvalue
19553 reference rather than the lvalue reference used in 03.
19554 </p>
19555 <p>
19556 Retaining this form would be safer, requiring an explicit move when splicing
19557 from lvalues.  However, this will break existing programs.
19558 We have the same problem with <tt>forward_list</tt>, although without the risk of
19559 breaking programs so here it might be viewed as a positive feature.
19560 </p>
19561 <p>
19562 The problem signatures:
19563 </p>
19564 <blockquote><pre>void splice_after(const_iterator position, forward_list&lt;T,Alloc&gt;&amp;&amp; x);
19565 void splice_after(const_iterator position, forward_list&lt;T,Alloc&gt;&amp;&amp; x,
19566                   const_iterator i);
19567 void splice_after(const_iterator position, forward_list&lt;T,Alloc&gt;&amp;&amp; x,
19568                   const_iterator first, const_iterator last);
19569
19570 void splice(const_iterator position, list&lt;T,Alloc&gt;&amp;&amp; x);
19571 void splice(const_iterator position, list&lt;T,Alloc&gt;&amp;&amp; x,
19572             const_iterator i);
19573 void splice(const_iterator position, list&lt;T,Alloc&gt;&amp;&amp; x,
19574             const_iterator first, const_iterator last);
19575 </pre></blockquote>
19576
19577 <b>Possible resolutions:</b>
19578
19579 <p>
19580 Option A.   Add an additional (non-const) lvalue-reference
19581 overload in each case
19582 </p>
19583 <p>
19584 Option B.   Change rvalue reference back to (non-const)
19585 lvalue-reference overload in each case
19586 </p>
19587 <p>
19588 Option C.   Add an additional (non-const) lvalue-reference
19589 overload in just the <tt>std::list</tt> cases
19590 </p>
19591 <p>
19592 I think (B) would be very unfortunate, I really like the <tt>forward_list</tt>
19593 behaviour in (C) but feel (A) is needed for consistency.
19594 </p>
19595 <p>
19596 My actual preference would be NAD, ship with this as a breaking change as it
19597 is a more explicit interface.  I don't think that will fly though!
19598 </p>
19599
19600 <p>
19601 See the thread starting with c++std-lib-23725 for more discussion.
19602 </p>
19603
19604 <p><i>[
19605 2009-10-27 Christopher Jefferson provides proposed wording for Option C.
19606 ]</i></p>
19607
19608
19609
19610 <p><b>Proposed resolution:</b></p>
19611 <p>
19612 In 23.3.4 [list]
19613 </p>
19614
19615 <p>
19616 Add lvalue overloads before rvalue ones:
19617 </p>
19618
19619 <blockquote><pre><ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x);</ins>
19620 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x);
19621 <ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x, const_iterator i);</ins>
19622 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x, const_iterator i);
19623 <ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x,
19624             const_iterator first, const_iterator last);</ins>
19625 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x,
19626             const_iterator first, const_iterator last);
19627 </pre></blockquote>
19628
19629 <p>
19630 In 23.3.4.4 [list.ops], similarly add lvalue overload before each rvalue one:
19631 </p>
19632 <p>
19633 (After paragraph 2)
19634 </p>
19635
19636 <blockquote><pre><ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x);</ins>
19637 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x);
19638 </pre></blockquote>
19639
19640 <p>
19641 (After paragraph 6)
19642 </p>
19643
19644 <blockquote><pre><ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x, const_iterator i);</ins>
19645 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x, const_iterator i);
19646 </pre></blockquote>
19647
19648 <p>
19649 (After paragraph 10)
19650 </p>
19651
19652 <blockquote><pre><ins>void splice(const_iterator position, list&lt;T,Allocator&gt;&amp; x,
19653             const_iterator first, const_iterator last);</ins>
19654 void splice(const_iterator position, list&lt;T,Allocator&gt;&amp;&amp; x,
19655             const_iterator first, const_iterator last);
19656 </pre></blockquote>
19657
19658
19659
19660
19661
19662
19663 <hr>
19664 <h3><a name="1134"></a>1134. Redundant specification of stdint.h, fenv.h, tgmath.h, and maybe complex.h</h3>
19665 <p><b>Section:</b> 18.4.2 [stdinth], 26.3.2 [fenv], 26.8 [c.math], 26.4.11 [cmplxh] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
19666  <b>Submitter:</b> Robert Klarer <b>Opened:</b> 2009-05-26  <b>Last modified:</b> 2009-10-20</p>
19667 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
19668 <p><b>Discussion:</b></p>
19669 <p>
19670 This is probably editorial.
19671 </p>
19672 <p>
19673 The following items should be removed from the draft, because they're
19674 redundant with Annex D, and they arguably make some *.h headers
19675 non-deprecated:
19676 </p>
19677 <p>
19678 18.4.2 [stdinth] (regarding <tt>&lt;stdint.h&gt;</tt>)
19679 </p>
19680 <p>
19681 26.3.2 [fenv] (regarding <tt>&lt;fenv.h&gt;</tt>
19682 </p>
19683 <p>
19684 Line 3 of 26.8 [c.math] (regarding <tt>&lt;tgmath.h&gt;</tt>)
19685 </p>
19686 <p>
19687 26.4.11 [cmplxh] (regarding <tt>&lt;complex.h&gt;</tt>, though the note in this subclause is not redundant)
19688 </p>
19689
19690 <p><i>[
19691 2009-06-10 Ganesh adds:
19692 ]</i></p>
19693
19694
19695 <blockquote>
19696 While searching for <tt>stdint</tt> in the CD, I found that <tt>&lt;stdint.h&gt;</tt> is also
19697 mentioned in 3.9.1 [basic.fundamental] /5. It guess it should refer to
19698 <tt>&lt;cstdint&gt;</tt> instead.
19699 </blockquote>
19700
19701 <p><i>[
19702 2009 Santa Cruz:
19703 ]</i></p>
19704
19705
19706 <blockquote>
19707 Real issue. Maybe just editorial, maybe not. Move to Ready.
19708 </blockquote>
19709
19710
19711
19712 <p><b>Proposed resolution:</b></p>
19713 <p>
19714 Remove the section 18.4.2 [stdinth].
19715 </p>
19716 <p>
19717 Remove the section 26.3.2 [fenv].
19718 </p>
19719 <p>
19720 Remove 26.8 [c.math], p3:
19721 </p>
19722
19723 <blockquote>
19724 <del>-3- The header <tt>&lt;tgmath.h&gt;</tt> effectively includes the headers <tt>&lt;complex.h&gt;</tt>
19725 and <tt>&lt;math.h&gt;</tt>.</del>
19726 </blockquote>
19727 <p>
19728 Remove the section 26.4.11 [cmplxh].
19729 </p>
19730
19731
19732
19733
19734
19735 <hr>
19736 <h3><a name="1135"></a>1135. <tt>exception_ptr</tt> should support contextual conversion to <tt>bool</tt></h3>
19737 <p><b>Section:</b> 18.8.5 [propagation] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
19738  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2007-06-06  <b>Last modified:</b> 2009-10-20</p>
19739 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#propagation">active issues</a> in [propagation].</p>
19740 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#propagation">issues</a> in [propagation].</p>
19741 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
19742 <p><b>Discussion:</b></p>
19743 <p>
19744 As of
19745 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2857.pdf">N2857</a>
19746 18.8.5 [propagation]/5, the implementation-defined type
19747 <tt>exception_ptr</tt> does provide the following ways to check whether
19748 it is a null value:
19749 </p>
19750 <blockquote><pre>void f(std::exception_ptr p) {
19751   p == nullptr;
19752   p == 0;
19753   p == exception_ptr();
19754 }
19755 </pre></blockquote>
19756 <p>
19757 This is rather cumbersome way of checking for the null value
19758 and I suggest to require support for evaluation in a boolean
19759 context like so:
19760 </p>
19761
19762 <blockquote><pre>void g(std::exception_ptr p) {
19763   if (p) {}
19764   !p;
19765 }
19766 </pre></blockquote>
19767
19768 <p><i>[
19769 2009 Santa Cruz:
19770 ]</i></p>
19771
19772
19773 <blockquote>
19774 Move to Ready. Note to editor: considering putting in a cross-reference
19775 to 4 [conv], paragraph 3, which defines the phrase
19776 "contextually converted to <tt>bool</tt>".
19777 </blockquote>
19778
19779
19780
19781 <p><b>Proposed resolution:</b></p>
19782 <p>
19783 In section 18.8.5 [propagation] insert a new paragraph between p.5 and p.6:
19784 </p>
19785
19786 <blockquote>
19787 <ins>
19788 An object <tt>e</tt> of type <tt>exception_ptr</tt> can be contextually converted to <tt>bool</tt>.
19789 The effect shall be as if <tt>e != exception_ptr()</tt> had been evaluated in place
19790 of <tt>e</tt>. There shall be no implicit conversion to arithmetic type, to
19791 enumeration type or to pointer type.
19792 </ins>
19793 </blockquote>
19794
19795
19796
19797
19798
19799 <hr>
19800 <h3><a name="1136"></a>1136. Incomplete specification of <tt>nested_exception::rethrow_nested()</tt></h3>
19801 <p><b>Section:</b> 18.8.6 [except.nested] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
19802  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2007-06-06  <b>Last modified:</b> 2009-10-20</p>
19803 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#except.nested">active issues</a> in [except.nested].</p>
19804 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#except.nested">issues</a> in [except.nested].</p>
19805 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
19806 <p><b>Discussion:</b></p>
19807 <p>
19808 It was recently mentioned in a newsgroup article
19809 <a href="http://groups.google.de/group/comp.std.c++/msg/f82022aff68edf3d">http://groups.google.de/group/comp.std.c++/msg/f82022aff68edf3d</a>
19810 that the specification of the member function <tt>rethrow_nested()</tt> of the
19811 class <tt>nested_exception</tt> is incomplete, specifically it remains unclear
19812 what happens, if member <tt>nested_ptr()</tt> returns a null value. In
19813 18.8.6 [except.nested] we find only the following paragraph related to that:
19814 </p>
19815 <blockquote><pre>void rethrow_nested() const; // [[noreturn]]
19816 </pre>
19817 <blockquote>
19818 -4- <i>Throws:</i> the stored exception captured by this <tt>nested_exception</tt> object.
19819 </blockquote>
19820 </blockquote>
19821 <p>
19822 This is a problem, because it is possible to create an object of
19823 <tt>nested_exception</tt> with exactly such a state, e.g.
19824 </p>
19825 <blockquote><pre>#include &lt;exception&gt;
19826 #include &lt;iostream&gt;
19827
19828 int main() try {
19829   std::nested_exception e; // OK, calls current_exception() and stores it's null value
19830   e.rethrow_nested(); // ?
19831   std::cout &lt;&lt; "A" &lt;&lt; std::endl;
19832 }
19833 catch(...) {
19834   std::cout &lt;&lt; "B" &lt;&lt; std::endl;
19835 }
19836 </pre></blockquote>
19837 <p>
19838 I suggest to follow the proposal of the reporter, namely to invoke
19839 <tt>terminate()</tt> if <tt>nested_ptr()</tt> return a null value of <tt>exception_ptr</tt> instead
19840 of relying on the fallback position of undefined behavior. This would
19841 be consistent to the behavior of a <tt>throw;</tt> statement when no
19842 exception is being handled.
19843 </p>
19844
19845 <p><i>[
19846 2009 Santa Cruz:
19847 ]</i></p>
19848
19849
19850 <blockquote>
19851 Move to Ready.
19852 </blockquote>
19853
19854
19855
19856 <p><b>Proposed resolution:</b></p>
19857 <p>
19858 Change around 18.8.6 [except.nested]/4 as indicated:
19859 </p>
19860 <blockquote>
19861 <p>
19862 -4- <i>Throws:</i> the stored exception captured by this <tt>nested_exception</tt>
19863 object<ins>, if <tt>nested_ptr() != nullptr</tt></ins>
19864 </p>
19865 <p>
19866 <ins>- <i>Remarks:</i> If <tt>nested_ptr() == nullptr</tt>, <tt>terminate()</tt>
19867 shall be called.</ins>
19868 </p>
19869 </blockquote>
19870
19871
19872
19873
19874
19875 <hr>
19876 <h3><a name="1137"></a>1137. Return type of <tt>conj</tt> and <tt>proj</tt></h3>
19877 <p><b>Section:</b> 26.4.9 [cmplx.over] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
19878  <b>Submitter:</b> Marc Steinbach <b>Opened:</b> 2009-06-11  <b>Last modified:</b> 2009-06-27</p>
19879 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#cmplx.over">issues</a> in [cmplx.over].</p>
19880 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
19881 <p><b>Discussion:</b></p>
19882 <p>
19883 In clause 1, the Working Draft
19884 (<a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2857.pdf">N2857</a>)
19885 specifies overloads of the
19886 functions
19887 </p>
19888 <blockquote><pre>arg, conj, imag, norm, proj, real
19889 </pre></blockquote>
19890 <p>
19891 for non-complex arithmetic types (<tt>float</tt>, <tt>double</tt>,
19892 <tt>long double</tt>, and integers).
19893 The only requirement (clause 2) specifies effective type promotion of arguments.
19894 </p>
19895 <p>
19896 I strongly suggest to add the following requirement on the return types:
19897 </p>
19898 <blockquote>
19899 All the specified overloads must return real (i.e., non-complex) values,
19900 specifically, the nested <tt>value_type</tt> of effectively promoted arguments.
19901 </blockquote>
19902
19903 <p>
19904 (This has no effect on <tt>arg</tt>, <tt>imag</tt>, <tt>norm</tt>, <tt>real</tt>:
19905 they are real-valued anyway.)
19906 </p>
19907 <p><b>Rationale:</b></p>
19908 <p>
19909 Mathematically, <tt>conj()</tt> and <tt>proj()</tt>, like the transcendental functions, are
19910 complex-valued in general but map the (extended) real line to itself.
19911 In fact, both functions act as identity on the reals.
19912 A typical user will expect <tt>conj()</tt> and <tt>proj()</tt> to preserve this essential
19913 mathematical property in the same way as <tt>exp()</tt>, <tt>sin()</tt>, etc.
19914 A typical use of <tt>conj()</tt>, e.g., is the generic scalar product of n-vectors:
19915 </p>
19916
19917 <blockquote><pre>template&lt;typename T&gt;
19918 inline T
19919 scalar_product(size_t n, T const* x, T const* y) {
19920   T result = 0;
19921   for (size_t i = 0; i &lt; n; ++i)
19922     result += x[i] * std::conj(y[i]);
19923   return result;
19924 }
19925 </pre></blockquote>
19926 <p>
19927 This will work equally well for real and complex floating-point types <tt>T</tt> if
19928 <tt>conj()</tt> returns <tt>T</tt>. It will not work with real types if <tt>conj()</tt>
19929 returns complex values.
19930 </p>
19931 <p>
19932 Instead, the implementation of <tt>scalar_product</tt> becomes either less efficient
19933 and less useful (if a complex result is always returned), or unnecessarily
19934 complicated (if overloaded versions with proper return types are defined).
19935 In the second case, the real-argument overload of <tt>conj()</tt> cannot be used.
19936 In fact, it must be avoided.
19937 </p>
19938 <p>
19939 Overloaded <tt>conj()</tt> and <tt>proj()</tt> are principally needed in generic programming.
19940 All such use cases will benefit from the proposed return type requirement,
19941 in a similar way as the <tt>scalar_product</tt> example.
19942 The requirement will not harm use cases where a complex return value
19943 is expected, because of implicit conversion to complex.
19944 Without the proposed return type guarantee, I find overloaded versions
19945 of <tt>conj()</tt> and <tt>proj()</tt> not only useless but actually troublesome.
19946 </p>
19947
19948
19949
19950 <p><b>Proposed resolution:</b></p>
19951 <p>
19952 Insert a new paragraph after 26.4.9 [cmplx.over]/2:
19953 </p>
19954
19955 <blockquote>
19956 <ins>
19957 All of the specified overloads shall have a return type which is the nested <tt>value_type</tt> of
19958 the effectively promoted arguments.
19959 </ins>
19960 </blockquote>
19961
19962
19963
19964
19965
19966 <hr>
19967 <h3><a name="1138"></a>1138. unusual return value for operator+</h3>
19968 <p><b>Section:</b> 21.4.8.1 [string::op+] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
19969  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-06-12  <b>Last modified:</b> 2009-11-05</p>
19970 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
19971 <p><b>Discussion:</b></p>
19972 <p>
19973 Many of the <tt>basic_string operator+</tt> overloads return an rvalue-reference.  Is
19974 that really intended?
19975 </p>
19976 <p>
19977 I'm considering it might be a mild performance tweak to avoid making
19978 un-necessary copies of a cheaply movable type, but it opens risk to dangling
19979 references in code like:
19980 </p>
19981
19982 <blockquote><pre>auto &amp;&amp; s = string{"x"} + string{y};
19983 </pre></blockquote>
19984
19985 <p>
19986 and I'm not sure about:
19987 </p>
19988
19989 <blockquote><pre>auto s = string{"x"} + string{y};
19990 </pre></blockquote>
19991
19992 <p><i>[
19993 2009-10-11 Howard updated <i>Returns:</i> clause for each of these.
19994 ]</i></p>
19995
19996
19997 <p><i>[
19998 2009-11-05 Howard adds:
19999 ]</i></p>
20000
20001
20002 <blockquote>
20003 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
20004 </blockquote>
20005
20006
20007 <p><b>Proposed resolution:</b></p>
20008 <p>
20009 Strike the <tt>&amp;&amp;</tt> from the return type in the following function
20010 signatures:
20011 </p>
20012
20013 <blockquote>
20014 <p>
20015 21.3 [string.classes] p2 Header Synopsis
20016 </p>
20017
20018 <blockquote><pre>template&lt;class charT, class traits, class Allocator&gt;
20019   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20020     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20021               const basic_string&lt;charT,traits,Allocator&gt;&amp; rhs);
20022
20023 template&lt;class charT, class traits, class Allocator&gt;
20024   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20025     operator+(const basic_string&lt;charT,traits,Allocator&gt;&amp; lhs,
20026               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20027
20028 template&lt;class charT, class traits, class Allocator&gt;
20029   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20030     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20031               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20032
20033
20034 template&lt;class charT, class traits, class Allocator&gt;
20035   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20036     operator+(const charT* lhs,
20037               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20038
20039 template&lt;class charT, class traits, class Allocator&gt;
20040   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20041     operator+(charT lhs, basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20042
20043 template&lt;class charT, class traits, class Allocator&gt;
20044   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20045     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20046               const charT* rhs);
20047
20048 template&lt;class charT, class traits, class Allocator&gt;
20049   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20050     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs, charT rhs);
20051 </pre></blockquote>
20052
20053 <p>
20054 21.4.8.1 [string::op+]
20055 </p>
20056
20057 <blockquote><pre>template&lt;class charT, class traits, class Allocator&gt;
20058   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20059     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20060               const basic_string&lt;charT,traits,Allocator&gt;&amp; rhs);
20061 </pre>
20062 <blockquote>
20063 <i>Returns:</i> <tt><ins>std::move(</ins>lhs.append(rhs)<ins>)</ins></tt>
20064 </blockquote>
20065 <pre>template&lt;class charT, class traits, class Allocator&gt;
20066   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20067     operator+(const basic_string&lt;charT,traits,Allocator&gt;&amp; lhs,
20068               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20069 </pre>
20070 <blockquote>
20071 <i>Returns:</i> <tt><ins>std::move(</ins>rhs.insert(0, lhs)<ins>)</ins></tt>
20072 </blockquote>
20073 <pre>template&lt;class charT, class traits, class Allocator&gt;
20074   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20075     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20076               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20077 </pre>
20078 <blockquote>
20079 <i>Returns:</i> <tt><ins>std::move(</ins>lhs.append(rhs)<ins>)</ins></tt> [<i>Note:</i> Or equivalently
20080 <tt><ins>std::move(</ins>rhs.insert(0, lhs)<ins>)</ins></tt> &#8212; <i>end note</i>]
20081 </blockquote>
20082 <pre>template&lt;class charT, class traits, class Allocator&gt;
20083   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20084     operator+(const charT* lhs,
20085               basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20086 </pre>
20087 <blockquote>
20088 <i>Returns:</i> <tt><ins>std::move(</ins>rhs.insert(0, lhs)<ins>)</ins></tt>.
20089 </blockquote>
20090 <pre>template&lt;class charT, class traits, class Allocator&gt;
20091   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20092     operator+(charT lhs, basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; rhs);
20093 </pre>
20094 <blockquote>
20095 <i>Returns:</i> <tt><ins>std::move(</ins>rhs.insert(0, 1, lhs)<ins>)</ins></tt>.
20096 </blockquote>
20097 <pre>template&lt;class charT, class traits, class Allocator&gt;
20098   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20099     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs,
20100               const charT* rhs);
20101 </pre>
20102 <blockquote>
20103 <i>Returns:</i> <tt><ins>std::move(</ins>lhs.append(rhs)<ins>)</ins></tt>.
20104 </blockquote>
20105 <pre>template&lt;class charT, class traits, class Allocator&gt;
20106   basic_string&lt;charT,traits,Allocator&gt;<del>&amp;&amp;</del>
20107     operator+(basic_string&lt;charT,traits,Allocator&gt;&amp;&amp; lhs, charT rhs);
20108 </pre>
20109 <blockquote>
20110 <i>Returns:</i> <tt><ins>std::move(</ins>lhs.append(1, rhs)<ins>)</ins></tt>.
20111 </blockquote>
20112 </blockquote>
20113
20114 </blockquote>
20115
20116
20117
20118
20119
20120
20121 <hr>
20122 <h3><a name="1144"></a>1144. "thread safe" is undefined</h3>
20123 <p><b>Section:</b> 18.5 [support.start.term] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
20124  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-16  <b>Last modified:</b> 2009-10-20</p>
20125 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#support.start.term">issues</a> in [support.start.term].</p>
20126 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
20127 <p><b>Discussion:</b></p>
20128
20129 <p><b>Addresses UK 187</b></p>
20130
20131 <p>
20132 The term "thread safe" is not defined nor used in this context
20133 anywhere else in the standard.
20134 </p>
20135
20136 <p><b>Suggested action:</b></p>
20137 <p>
20138 Clarify the meaning of "thread safe".
20139 </p>
20140
20141 <p><i>[
20142 2009 Santa Cruz:
20143 ]</i></p>
20144
20145
20146 <blockquote>
20147 <p>
20148 The "thread safe" language has already been change in the WP. It was
20149 changed to "happen before", but the current WP text is still a little
20150 incomplete: "happen before" is binary, but the current WP text only
20151 mentions one thing.
20152 </p>
20153 <p>
20154 Move to Ready.
20155 </p>
20156 </blockquote>
20157
20158
20159
20160 <p><b>Proposed resolution:</b></p>
20161 <p>
20162 For the following functions in 18.5 [support.start.term].
20163 </p>
20164 <blockquote><pre><code>
20165 extern "C" int at_quick_exit(void (*f)(void));
20166 extern "C++" int at_quick_exit(void (*f)(void));
20167 </code></pre></blockquote>
20168
20169 <p>
20170 Edit paragraph 10 as follows.
20171 The intent is
20172 to provide the other half of the happens before relation;
20173 to note indeterminate ordering;
20174 and to clean up some formatting.
20175 </p>
20176 <blockquote><p>
20177 <i>Effects:</i>
20178 The <code>at_quick_exit()</code> functions
20179 register the function pointed to by <code>f</code>
20180 to be called without arguments when <code>quick_exit</code> is called.
20181 It is unspecified whether a call to <code>at_quick_exit()</code>
20182 that does not <del>happen-before</del> <ins>happen before</ins> (1.10)
20183 <ins>all calls to <code>quick_exit</code></ins>
20184 will succeed.
20185 [<i>Note:</i>
20186 the <code>at_quick_exit()</code> functions
20187 shall not introduce a data race (17.6.4.7).
20188 <del>exitnote</del>
20189 <ins>&#8212;<i>end note</i>]</ins>
20190 <ins>
20191 [<i>Note:</i>
20192 The order of registration may be indeterminate
20193 if <code>at_quick_exit</code> was called from more than one thread.
20194 &#8212;<i>end note</i>]
20195 </ins>
20196 [<i>Note:</i> The <code>at_quick_exit</code> registrations
20197 are distinct from the <code>atexit</code> registrations,
20198 and applications may need to call both registration functions
20199 with the same argument.
20200 &#8212;<i>end note</i>]
20201 </p></blockquote>
20202
20203 <p>
20204 For the following function.
20205 </p>
20206 <blockquote><pre><code>
20207 void quick_exit [[noreturn]] (int status)
20208 </code></pre></blockquote>
20209
20210 <p>
20211 Edit paragraph 13 as follows.
20212 The intent is to note that thread-local variables may be different.
20213 </p>
20214 <blockquote><p>
20215 <i>Effects:</i>
20216 Functions registered by calls to <code>at_quick_exit</code>
20217 are called in the reverse order of their registration,
20218 except that a function shall be called
20219 after any previously registered functions
20220 that had already been called at the time it was registered.
20221 Objects shall not be destroyed as a result of calling <code>quick_exit</code>.
20222 If control leaves a registered function called by <code>quick_exit</code>
20223 because the function does not provide a handler for a thrown exception,
20224 <code>terminate()</code> shall be called.
20225 <ins>
20226 [<i>Note:</i>
20227 Functions registered by one thread may be called by any thread,
20228 and hence should not rely on the identity of thread-storage-duration objects.
20229 &#8212;<i>end note</i>]
20230 </ins>
20231 After calling registered functions,
20232 <code>quick_exit</code> shall call <code>_Exit(status)</code>.
20233 [<i>Note:</i>
20234 The standard file buffers are not flushed.
20235 See: ISO C 7.20.4.4.
20236 &#8212;<i>end note</i>]
20237 </p></blockquote>
20238
20239
20240
20241
20242
20243 <hr>
20244 <h3><a name="1151"></a>1151. Behavior of the library in the presence of threads is incompletely specified</h3>
20245 <p><b>Section:</b> 17 [library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
20246  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-20</p>
20247 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
20248 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
20249 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
20250 <p><b>Discussion:</b></p>
20251 <p><b>Addresses US 63</b></p>
20252
20253    <p><b>Description</b></p>
20254         <p>The behavior of the library in the presence of threads
20255         is incompletely specified.</p>
20256         <p>For example, if thread 1 assigns to <tt>X</tt>, then writes data
20257         to file <tt>f</tt>, which is read by thread 2, and then accesses
20258         variable <tt>X</tt>, is thread 2 guaranteed to be able to see the
20259         value assigned to <tt>X</tt> by thread 1? In other words, does the
20260         write of the data "happen before" the read?</p>
20261         <p>Another example: does simultaneous access using <tt>operator
20262         at()</tt> to different characters in the same non-const string
20263         really introduce a data race?</p>
20264 <p><b>Suggestion</b></p>
20265 <p><b>Notes</b></p><p>17 SG: should go to threads group; misclassified in document
20266 </p>
20267
20268     <p>Concurrency SG: Create an issue. Hans will look into it.</p>
20269
20270 <p><i>[
20271 2009 Santa Cruz:
20272 ]</i></p>
20273
20274
20275 <blockquote>
20276 Move to "Open". Hans and the rest of the concurrency working group will
20277 study this. We can't make progress without a thorough review and a
20278 paper.
20279 </blockquote>
20280
20281
20282
20283 <p><b>Proposed resolution:</b></p>
20284
20285
20286
20287
20288
20289 <hr>
20290 <h3><a name="1152"></a>1152. expressions parsed differently than intended</h3>
20291 <p><b>Section:</b> 22.4.2.2.2 [facet.num.put.virtuals] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
20292  <b>Submitter:</b> Seungbeom Kim <b>Opened:</b> 2009-06-27  <b>Last modified:</b> 2009-10-28</p>
20293 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#facet.num.put.virtuals">active issues</a> in [facet.num.put.virtuals].</p>
20294 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#facet.num.put.virtuals">issues</a> in [facet.num.put.virtuals].</p>
20295 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
20296 <p><b>Discussion:</b></p>
20297 <p>
20298 In Table 73 -- Floating-point conversions, 22.4.2.2.2 [facet.num.put.virtuals],
20299 in
20300 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>,
20301 we have the following entries:
20302 </p>
20303 <table border="1">
20304 <caption>Table 73 &#8212; Floating-point conversions</caption>
20305 <tbody><tr>
20306 <th>State</th> <th><tt>stdio</tt> equivalent</th>
20307 </tr>
20308 <tr>
20309 <td><tt>floatfield == ios_base::fixed | ios_base::scientific &amp;&amp; !uppercase</tt></td>
20310 <td align="center"><tt>%a</tt></td>
20311 </tr>
20312
20313 <tr>
20314 <td><tt>floatfield == ios_base::fixed | ios_base::scientific</tt></td>
20315 <td align="center"><tt>%A</tt></td>
20316 </tr>
20317 </tbody></table>
20318
20319 <p>
20320 These expressions are supposed to mean:
20321 </p>
20322
20323 <blockquote><pre>floatfield == (ios_base::fixed | ios_base::scientific) &amp;&amp; !uppercase
20324 floatfield == (ios_base::fixed | ios_base::scientific)
20325 </pre></blockquote>
20326 <p>
20327 but technically parsed as:
20328 </p>
20329 <blockquote><pre>((floatfield == ios_base::fixed) | ios_base::scientific) &amp;&amp; (!uppercase)
20330 ((floatfield == ios_base::fixed) | ios_base::scientific)
20331 </pre></blockquote>
20332 <p>
20333 and should be corrected with additional parentheses, as shown above.
20334 </p>
20335
20336 <p><i>[
20337 2009-10-28 Howard:
20338 ]</i></p>
20339
20340
20341 <blockquote>
20342 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
20343 </blockquote>
20344
20345
20346
20347 <p><b>Proposed resolution:</b></p>
20348 <p>
20349 Change Table 83 &#8212; Floating-point conversions in  22.4.2.2.2 [facet.num.put.virtuals]:
20350 </p>
20351
20352 <table border="1">
20353 <caption>Table 83 &#8212; Floating-point conversions</caption>
20354 <tbody><tr>
20355 <th>State</th> <th><tt>stdio</tt> equivalent</th>
20356 </tr>
20357 <tr>
20358 <td><tt>floatfield == <ins>(</ins>ios_base::fixed | ios_base::scientific<ins>)</ins> &amp;&amp; !uppercase</tt></td>
20359 <td align="center"><tt>%a</tt></td>
20360 </tr>
20361
20362 <tr>
20363 <td><tt>floatfield == <ins>(</ins>ios_base::fixed | ios_base::scientific<ins>)</ins></tt></td>
20364 <td align="center"><tt>%A</tt></td>
20365 </tr>
20366 </tbody></table>
20367
20368
20369
20370
20371
20372 <hr>
20373 <h3><a name="1153"></a>1153. Standard library needs review for constructors to be
20374 explicit to avoid treatment as initializer-list constructor</h3>
20375 <p><b>Section:</b> 17 [library], 30 [thread], D [depr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
20376  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-20</p>
20377 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
20378 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
20379 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
20380 <p><b>Discussion:</b></p>
20381
20382 <p><b>Addresses DE 2</b></p>
20383
20384 <p><b>Description</b></p>
20385         <p>Marking a constructor with <tt>explicit</tt> has semantics
20386         even for a constructor with zero or several parameters:
20387         Such a constructor cannot be used with list-initialization
20388         in a copy-initialization context, see 13.3.1.7 [over.match.list]. The
20389         standard library apparently has not been reviewed for
20390         marking non-single-parameter constructors as <tt>explicit</tt>.</p>
20391 <p><b>Suggestion</b></p>
20392         <p>Consider marking zero-parameter and multi-parameter
20393         constructors <tt>explicit</tt> in classes that have at least one
20394         constructor marked <tt>explicit</tt> and that do not have an
20395         initializer-list constructor.</p>
20396
20397 <p><b>Notes</b></p>
20398         <p>Robert Klarer to address this one.</p>
20399
20400 <p><i>[
20401 2009 Santa Cruz:
20402 ]</i></p>
20403
20404
20405 <blockquote>
20406 Move to "Open". Robert Klarer has promised to provide wording.
20407 </blockquote>
20408
20409
20410
20411 <p><b>Proposed resolution:</b></p>
20412
20413
20414
20415
20416
20417 <hr>
20418 <h3><a name="1154"></a>1154. <tt>complex</tt> should accept integral types</h3>
20419 <p><b>Section:</b> 26.4 [complex.numbers] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
20420  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-26</p>
20421 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#complex.numbers">issues</a> in [complex.numbers].</p>
20422 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
20423 <p><b>Discussion:</b></p>
20424
20425 <p><b>Addresses FR 35</b></p>
20426
20427 <p><b>Description</b></p>
20428         <p>Instantiations of the class
20429         template <tt>complex&lt;&gt;</tt> have to be allowed for integral
20430         types, to reflect existing practice and ISO standards
20431         (LIA-III).</p>
20432
20433 <p><b>Suggestion</b></p>
20434
20435 <p><i>[
20436 2009-10-26 Proposed wording in
20437 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n3002.pdf">N3002</a>.
20438 ]</i></p>
20439
20440
20441
20442
20443 <p><b>Proposed resolution:</b></p>
20444 Adopt
20445 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n3002.pdf">N3002</a>.
20446
20447
20448
20449
20450
20451 <hr>
20452 <h3><a name="1156"></a>1156. Constraints on bitmask and enumeration types to be tightened</h3>
20453 <p><b>Section:</b> 17.5.2.1.2 [enumerated.types], 17.5.2.1.3 [bitmask.types] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
20454  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-27</p>
20455 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
20456 <p><b>Discussion:</b></p>
20457
20458 <p><b>Addresses UK 165</b></p>
20459
20460 <p><b>Description</b></p>
20461         <p>Constraints on
20462         bitmask and enumeration types were supposed to be tightened
20463         up as part of the motivation for the <tt>constexpr</tt> feature -
20464         see paper
20465         <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2235.pdf">N2235</a>
20466         for details</p>
20467 <p><b>Suggestion</b></p>
20468         <p>Adopt wording in line with the motivation
20469         described in
20470         <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2235.pdf">N2235</a></p>
20471 <p><b>Notes</b></p>
20472         <p>Robert Klarer to review</p>
20473
20474 <p><i>[
20475 2009 Santa Cruz:
20476 ]</i></p>
20477
20478
20479 <blockquote>
20480 Move to Open. Ping Robert Klarer to provide wording, using N2235 as
20481 guidance.
20482 </blockquote>
20483
20484
20485
20486 <p><b>Proposed resolution:</b></p>
20487
20488
20489
20490
20491
20492 <hr>
20493 <h3><a name="1157"></a>1157. Local types can now instantiate templates</h3>
20494 <p><b>Section:</b> 17.6.3.2.1 [namespace.std] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
20495  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-21</p>
20496 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
20497 <p><b>Discussion:</b></p>
20498
20499 <p><b>Addresses UK 175</b></p>
20500
20501 <p><b>Description</b></p>
20502         <p>Local types can
20503         now be used to instantiate templates, but don't have
20504         external linkage.</p>
20505 <p><b>Suggestion</b></p>
20506         <p>Remove the reference to external linkage.</p>
20507
20508 <p><b>Notes</b></p>
20509 <p>We accept the proposed solution. Martin will draft an issue.</p>
20510
20511 <p><i>[
20512 2009-07-28 Alisdair provided wording.
20513 ]</i></p>
20514
20515
20516 <p><i>[
20517 2009-10 Santa Cruz:
20518 ]</i></p>
20519
20520
20521 <blockquote>
20522 Moved to Ready.
20523 </blockquote>
20524
20525
20526
20527 <p><b>Proposed resolution:</b></p>
20528 <p>
20529 17.6.3.2.1 [namespace.std]
20530 </p>
20531 <p>
20532 Strike "of external linkage" in p1 and p2:
20533 </p>
20534
20535 <blockquote>
20536 <p>
20537 -1- The behavior of a C++ program is undefined if it adds declarations or
20538 definitions to namespace <tt>std</tt> or to a namespace within namespace <tt>std</tt>
20539 unless otherwise specified. A program may add a concept map for any
20540 standard library concept or a template specialization for any standard
20541 library template to namespace <tt>std</tt> only if the declaration depends on a
20542 user-defined type <del>of external linkage</del> and the specialization meets the
20543 standard library requirements for the original template and is not
20544 explicitly prohibited.<sup>179</sup>
20545 </p>
20546
20547 <p>
20548 -2- The behavior of a C++ program is undefined if it declares
20549 </p>
20550 <ul>
20551 <li>
20552 an explicit specialization of any member function of a standard library
20553 class template, or
20554 </li>
20555 <li>
20556 an explicit specialization of any member function template of a standard
20557 library class or class template, or
20558 </li>
20559 <li>
20560 an explicit or partial specialization of any member class template of a
20561 standard library class or class template.
20562 </li>
20563 </ul>
20564 <p>
20565 A program may explicitly instantiate a template defined in the standard
20566 library only if the declaration depends on the name of a user-defined
20567 type <del>of external linkage</del> and the instantiation meets the standard
20568 library requirements for the original template.
20569 </p>
20570 </blockquote>
20571
20572
20573
20574
20575
20576
20577 <hr>
20578 <h3><a name="1158"></a>1158. Encouragement to use monotonic clock</h3>
20579 <p><b>Section:</b> 30.2.4 [thread.req.timing] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
20580  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-10-31</p>
20581 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
20582 <p><b>Discussion:</b></p>
20583
20584 <p><b>Addresses UK 322, US 96</b></p>
20585
20586 <p><b>Description</b></p>
20587         <p>Not all systems
20588         can provide a monotonic clock. How are they expected to
20589         treat a _for function?</p>
20590 <p><b>Suggestion</b></p>
20591         <p>Add at least a note explaining the intent
20592         for systems that do not support a monotonic clock.</p>
20593
20594 <p><b>Notes</b></p>
20595 <p>Create an issue, together with UK 96. Note that the specification as is
20596     already allows a non-monotonic clock due to the word &#8220;should&#8221; rather than
20597     &#8220;shall&#8221;. If this wording is kept, a footnote should be added to make the
20598     meaning clear.</p>
20599
20600 <p><i>[ 2009-06-29 Beman provided a proposed resolution. ] </i></p>
20601
20602 <p><i>[
20603 2009-10-31 Howard adds:
20604 ]</i></p>
20605
20606
20607 <blockquote>
20608 Set to Tentatively Ready after 5 positive votes on c++std-lib.
20609 </blockquote>
20610
20611
20612
20613 <p><b>Proposed resolution:</b></p>
20614
20615 <p><i>Change Timing specifications 30.2.4 [thread.req.timing] as indicated:</i></p>
20616
20617 <p>The member functions whose names end in <tt>_for</tt> take an argument that
20618 specifies a relative time. Implementations
20619 <del>should</del> <ins>are encouraged but not required to</ins> use a
20620 monotonic clock to measure time for these functions.</p>
20621
20622
20623
20624
20625
20626
20627 <hr>
20628 <h3><a name="1159"></a>1159. Unclear spec for <tt>resource_deadlock_would_occur</tt></h3>
20629 <p><b>Section:</b> 30.4.3.2.2 [thread.lock.unique.locking] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
20630  <b>Submitter:</b> LWG <b>Opened:</b> 2009-06-28  <b>Last modified:</b> 2009-07-04</p>
20631 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.lock.unique.locking">active issues</a> in [thread.lock.unique.locking].</p>
20632 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.lock.unique.locking">issues</a> in [thread.lock.unique.locking].</p>
20633 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
20634 <p><b>Discussion:</b></p>
20635
20636 <p><b>Addresses UK 327, UK 328</b></p>
20637
20638 <p><b>UK 327 Description</b></p>
20639         <p>Not clear what
20640         the specification for error condition
20641         <tt>resource_deadlock_would_occur</tt> means. It is perfectly
20642         possible for this thread to own the mutex without setting
20643         owns to true on this specific lock object. It is also
20644         possible for lock operations to succeed even if the thread
20645         does own the mutex, if the mutex is recursive. Likewise, if
20646         the mutex is not recursive and the mutex has been locked
20647         externally, it is not always possible to know that this
20648         error condition should be raised, depending on the host
20649         operating system facilities. It is possible that 'i.e.' was
20650         supposed to be 'e.g.' and that suggests that recursive
20651         locks are not allowed. That makes sense, as the
20652         exposition-only member owns is boolean and not a integer to
20653         count recursive locks.</p>
20654
20655 <p><b>UK 327 Suggestion</b></p>
20656         <p>Add a precondition <tt>!owns</tt>. Change the 'i.e.'
20657         in the error condition to be 'e.g.' to allow for this
20658         condition to propogate deadlock detection by the host OS.</p>
20659 <p><b>UK 327 Notes</b></p>
20660 <p>Create an issue. Assigned to Lawrence Crowl. Note: not sure what try_lock
20661     means for recursive locks when you are the owner. POSIX has language on
20662     this, which should ideally be followed. Proposed fix is not quite right, for
20663     example, try_lock should have different wording from lock.</p>
20664
20665 <p><b>UK 328 Description</b></p>
20666
20667         <p>There is a missing precondition that <tt>owns</tt>
20668         is true, or an <tt>if(owns)</tt> test is missing from the effect
20669         clause</p>
20670 <p><b>UK 328 Suggestion</b></p>
20671         <p>Add a
20672         precondition that <tt>owns == true</tt>. Add an error condition to
20673         detect a violation, rather than yield undefined behaviour.</p>
20674 <p><b>UK 328 Notes</b></p>
20675 <p>Handle in same issue as UK 327. Also uncertain that the proposed resolution
20676     is the correct one.</p>
20677
20678
20679
20680
20681 <p><b>Proposed resolution:</b></p>
20682
20683
20684
20685
20686
20687 <hr>
20688 <h3><a name="1169"></a>1169. <tt>num_get</tt> not fully compatible with <tt>strto*</tt></h3>
20689 <p><b>Section:</b> 22.4.2.1.2 [facet.num.get.virtuals] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
20690  <b>Submitter:</b> Cosmin Truta <b>Opened:</b> 2009-07-04  <b>Last modified:</b> 2009-07-07</p>
20691 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#facet.num.get.virtuals">active issues</a> in [facet.num.get.virtuals].</p>
20692 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#facet.num.get.virtuals">issues</a> in [facet.num.get.virtuals].</p>
20693 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
20694 <p><b>Discussion:</b></p>
20695 <p>
20696 As specified in the latest draft,
20697 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>,
20698 <code>num_get</code> is still not fully compatible with the following C
20699 functions: <code>strtoul</code>, <code>strtoull</code>,
20700 <code>strtof</code> and
20701 <code>strtod</code>.
20702 </p>
20703 <p>
20704 In C, when conversion of a string to an unsigned integer type falls
20705 outside the
20706 representable range, <code>strtoul</code> and <code>strtoull</code> return
20707 <code>ULONG_MAX</code> and <code>ULLONG_MAX</code>, respectively,
20708 regardless
20709 whether the input field represents a positive or a negative value.
20710 On the other hand, the result of <code>num_get</code> conversion of
20711 negative
20712 values to unsigned integer types is zero. This raises a compatibility
20713 issue.
20714 </p>
20715 <p>
20716 Moreover, in C, when conversion of a string to a floating-point type falls
20717 outside the representable range, <code>strtof</code>, <code>strtod</code>
20718 and
20719 <code>strtold</code> return <code>±HUGE_VALF</code>,
20720 <code>±HUGE_VAL</code> and <code>±HUGE_VALL</code>, respectively.
20721 On the other hand, the result of <code>num_get</code> conversion of such
20722 out-of-range floating-point values results in the most positive/negative
20723 representable value.
20724 Although many C library implementations do implement <code>HUGE_VAL</code>
20725 (etc.) as the highest representable (which is, usually, the infinity),
20726 this
20727 isn't required by the C standard. The C library specification makes no
20728 statement regarding the value of <code>HUGE_VAL</code> and friends, which
20729 potentially raises the same compatibility issue as in the above case of
20730 unsigned integers.
20731 In addition, neither C nor C++ define symbolic constants for the maximum
20732 representable floating-point values (they only do so only for the maximum
20733 representable <i>finite</i> floating-point values), which raises a
20734 usability
20735 issue (it would be hard for the programmer to check the result of
20736 <code>num_get</code> against overflow).
20737 </p>
20738 <p>
20739 As such, we propose to adjust the specification of <code>num_get</code> to
20740 closely follow the behavior of all of its underlying C functions.
20741 </p>
20742
20743
20744 <p><b>Proposed resolution:</b></p>
20745
20746 <p>
20747 Change 22.4.2.1.2 [facet.num.get.virtuals] as follows:
20748 </p>
20749 <blockquote>
20750 <p>
20751 <b>Stage 3:</b>
20752 The sequence of <code>char</code>s accumulated in stage 2 (the field) is
20753 converted to a numeric value by the rules of one of the functions declared in
20754 the header <code>&lt;cstdlib&gt;</code>:
20755 </p>
20756 <ul>
20757 <li>For a signed integer value, the function <code>strtoll</code>.</li>
20758 <li>For an unsigned integer value, the function <code>strtoull</code>.</li>
20759 <li><ins>For a <code>float</code> value, the function
20760     <code>strtof</code>.</ins></li>
20761 <li><ins>For a <code>double</code> value, the function
20762     <code>strtod</code>.</ins></li>
20763 <li>For a <del>floating-point</del> <ins><code>long double</code></ins>
20764     value, the function <code>strtold</code>.</li>
20765 </ul>
20766 <p>
20767 The numeric value to be stored can be one of:
20768 </p>
20769 <ul>
20770 <li>zero, if the conversion function fails to convert the entire field.
20771     <del><code>ios_base::failbit</code> is assigned to <code>err</code>.</del></li>
20772 <li>the most positive <ins>(or negative)</ins> representable value, if
20773     the field <ins>to be converted to a signed integer type</ins> represents a
20774     value too large positive <ins>(or negative)</ins> to be represented in
20775     <code>val</code>.
20776     <del><code>ios_base::failbit</code> is assigned to <code>err</code>.</del></li>
20777 <li><del>the most negative representable value or zero for an unsigned integer
20778     type, if the field represents a value too large negative to be represented
20779     in <code>val</code>.
20780     <code>ios_base::failbit</code> is assigned to <code>err</code>.</del></li>
20781 <li><ins>the most positive representable value, if the field to be converted to
20782     an unsigned integer type represents a value that cannot be represented in
20783     <code>val</code>.</ins></li>
20784 <li>the converted value, otherwise.</li>
20785 </ul>
20786 <p>
20787 The resultant numeric value is stored in <code>val</code>.
20788 <ins>If the conversion function fails to convert the entire field, or if the
20789 field represents a value outside the range of representable values,
20790 <code>ios_base::failbit</code> is assigned to <code>err</code>.</ins>
20791 </p>
20792 </blockquote>
20793
20794
20795
20796
20797
20798
20799 <hr>
20800 <h3><a name="1170"></a>1170. String <i>char-like types</i> no longer PODs</h3>
20801 <p><b>Section:</b> 21.1 [strings.general] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
20802  <b>Submitter:</b> Beman Dawes <b>Opened:</b> 2009-06-22  <b>Last modified:</b> 2009-11-04</p>
20803 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
20804 <p><b>Discussion:</b></p>
20805
20806 <p><b>Addresses UK 218</b></p>
20807
20808 <p>Prior to the introduction of constant expressions into the library,
20809 <tt>basic_string</tt> elements had to be POD types, and thus had to be both trivially
20810 copyable and standard-layout. This ensured that they could be memcpy'ed and
20811 would be compatible with other libraries and languages, particularly the C
20812 language and its library.</p>
20813 <p>
20814 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2349.pdf">N2349</a>,
20815 Constant Expressions in the Standard Library Revision 2, changed the
20816 requirement in 21/1 from "POD type" to "literal type". That change had the
20817 effect of removing the trivially copyable and standard-layout requirements from
20818 <tt>basic_string</tt> elements.</p>
20819 <p>This means that <tt>basic_string</tt> elements no longer are guaranteed to be
20820 memcpy'able, and are no longer guaranteed to be standard-layout types:</p>
20821 <blockquote>
20822   <p>3.9/p2 and 3.9/p3 both make it clear that a "trivially copyable type" is
20823   required for memcpy to be guaranteed to work.</p>
20824   <p>Literal types (3.9p12) may have a non-trivial copy assignment operator, and
20825   that violates the trivially copyable requirements given in 9/p 6, bullet item
20826   2. </p>
20827   <p>Literal types (3.9p12) have no standard-layout requirement, either.</p>
20828 </blockquote>
20829 <p>This situation probably arose because the wording for "Constant Expressions
20830 in the Standard Library" was in process at the same time the C++ POD
20831 deconstruction wording was in process. </p>
20832 <p>Since trivially copyable types meet the C++0x requirements for literal types,
20833 and thus work with constant expressions, it seems an easy fix to revert the
20834 <tt>basic_string</tt> element wording to its original state.</p>
20835
20836  <p><i>[
20837  2009-07-28 Alisdair adds:
20838  ]</i></p>
20839
20840
20841 <blockquote>
20842 When looking for any resolution for this issue, consider the definition of
20843 "character container type" in 17.3.4 [defns.character.container].  This
20844 does require the character type to be a POD, and this term is used in a
20845 number of places through clause 21 and 28. This suggests the PODness
20846 constraint remains, but is much more subtle than before.  Meanwhile, I
20847 suspect the change from POD type to literal type was intentional with
20848 the assumption that trivially copyable types with
20849 non-trivial-but-constexpr constructors should serve as well.  I don't
20850 believe the current wording offers the right guarantees for either of
20851 the above designs.
20852 </blockquote>
20853
20854 <p><i>[
20855 2009-11-04 Howard modifies proposed wording to disallow array types as
20856 char-like types.
20857 ]</i></p>
20858
20859
20860
20861
20862 <p><b>Proposed resolution:</b></p>
20863
20864 <p><i>Change General 21.1 [strings.general] as indicated:</i></p>
20865 <blockquote>
20866 <p>This Clause describes components for manipulating sequences of any
20867 <del>literal</del> <ins>non-array POD</ins> (3.9) type. In this Clause
20868 such types are called <i>char-like types</i>, and objects of char-like
20869 types are called <i>char-like objects</i> or simply
20870 <i>characters</i>.</p>
20871 </blockquote>
20872
20873
20874
20875
20876
20877
20878 <hr>
20879 <h3><a name="1171"></a>1171. duration types should be literal</h3>
20880 <p><b>Section:</b> 20.9.3 [time.duration] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
20881  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-07-06  <b>Last modified:</b> 2009-10-31</p>
20882 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#time.duration">active issues</a> in [time.duration].</p>
20883 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.duration">issues</a> in [time.duration].</p>
20884 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
20885 <p><b>Discussion:</b></p>
20886 <p>
20887 The <tt>duration</tt> types in 20.9.3 [time.duration] are exactly the sort of type
20888 that should be "literal types" in the new standard.  Likewise,
20889 arithmetic operations on <tt>duration</tt>s should be declared <tt>constexpr</tt>.
20890 </p>
20891
20892 <p><i>[
20893 2009-09-21 Daniel adds:
20894 ]</i></p>
20895
20896
20897 <blockquote>
20898 An alternative (and possibly preferable solution for potentially
20899 heap-allocating big_int representation types) would be to ask the core
20900 language to allow references to <tt>const</tt> literal types as feasible
20901 arguments for <tt>constexpr</tt> functions.
20902 </blockquote>
20903
20904 <p><i>[
20905 2009-10-30 Alisdair adds:
20906 ]</i></p>
20907
20908
20909 <blockquote>
20910 <p>
20911 I suggest this issue moves from New to Open.
20912 </p>
20913
20914 <p>
20915 Half of this issue was dealt with in paper
20916 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2994.html">n2994</a>
20917 on constexpr constructors.
20918 </p>
20919
20920 <p>
20921 The other half (duration arithmetic) is on hold pending Core support for
20922 <tt>const &amp;</tt> in <tt>constexpr</tt> functions.
20923 </p>
20924
20925 </blockquote>
20926
20927
20928
20929 <p><b>Proposed resolution:</b></p>
20930 <p>
20931 Add <tt>constexpr</tt> to declaration of following functions and constructors:
20932 </p>
20933 <p>
20934 p1 20.9 [time]
20935 </p>
20936
20937 <blockquote>
20938 <p>
20939 <b>Header <tt>&lt;chrono&gt;</tt> synopsis</b>
20940 </p>
20941
20942 <p><i>[Draughting note - observe switch to pass-by-value to support constexpr]</i></p>
20943
20944
20945 <pre><i>// duration arithmetic</i>
20946 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20947    typename common_type&lt;duration&lt;Rep1, Period1&gt;, duration&lt;Rep2, Period2&gt;&gt;::type
20948    <ins>constexpr</ins> operator+(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20949 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20950    typename common_type&lt;duration&lt;Rep1, Period1&gt;, duration&lt;Rep2, Period2&gt;&gt;::type
20951    <ins>constexpr</ins> operator-(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20952 template &lt;class Rep1, class Period, class Rep2&gt;
20953    duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
20954    <ins>constexpr</ins> operator*(<del>const</del> duration&lt;Rep1, Period&gt;<del>&amp;</del> d, <del>const</del> Rep2<del>&amp;</del> s);
20955 template &lt;class Rep1, class Period, class Rep2&gt;
20956    duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
20957    <ins>constexpr</ins> operator*(<del>const</del> Rep1<del>&amp;</del> s, <del>const</del> duration&lt;Rep2, Period&gt;<del>&amp;</del> d);
20958 template &lt;class Rep1, class Period, class Rep2&gt;
20959    duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
20960    <ins>constexpr</ins> operator/(<del>const</del> duration&lt;Rep1, Period&gt;<del>&amp;</del> d, <del>const</del> Rep2<del>&amp;</del> s);
20961 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20962    typename common_type&lt;Rep1, Rep2&gt;::type
20963    <ins>constexpr</ins> operator/(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20964
20965 <i>// duration comparisons</i>
20966 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20967    <ins>constexpr</ins> bool operator==(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20968 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20969    <ins>constexpr</ins> bool operator!=(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20970 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20971    <ins>constexpr</ins> bool operator&lt; (<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20972 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20973    <ins>constexpr</ins> bool operator&lt;=(<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20974 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20975    <ins>constexpr</ins> bool operator&gt; (<del>const</del> duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20976 template &lt;class Rep1, class Period1, class Rep2, class Period2&gt;
20977    <ins>constexpr</ins> bool operator&gt;=(<del>const</del>  duration&lt;Rep1, Period1&gt;<del>&amp;</del> lhs, <del>const</del> duration&lt;Rep2, Period2&gt;<del>&amp;</del> rhs);
20978
20979 <i>// duration_cast</i>
20980 template &lt;class ToDuration, class Rep, class Period&gt;
20981    <ins>constexpr</ins> ToDuration duration_cast(<del>const</del> duration&lt;Rep, Period&gt;<del>&amp;</del> d);
20982 </pre>
20983
20984 <p>
20985 <b>20.9.3 [time.duration]</b>
20986 </p>
20987
20988 <pre>template &lt;class Rep, class Period = ratio&lt;1&gt;&gt;
20989 class duration {
20990   ....
20991 public:
20992   <i>// 20.9.3.1, construct/copy/destroy:</i>
20993  <ins>constexpr</ins> duration() = default;
20994
20995  template &lt;class Rep2&gt;
20996    <ins>constexpr</ins> explicit duration(const Rep2&amp; r);
20997  template &lt;class Rep2, class Period2&gt;
20998    <ins>constexpr</ins> duration(const duration&lt;Rep2, Period2&gt;&amp; d);
20999
21000   <ins>constexpr</ins> duration(const duration&amp;) = default;
21001
21002   <i>// 20.9.3.2, observer:</i>
21003   <ins>constexpr</ins> rep count() const;
21004
21005   <i>// 20.9.3.3, arithmetic:</i>
21006   <ins>constexpr</ins> duration operator+() const;
21007   <ins>constexpr</ins> duration operator-() const;
21008   ...
21009
21010 };
21011 </pre>
21012 </blockquote>
21013 <p><i>[
21014 Note - this edit already seems assumed by definition of the duration static members <tt>zero/min/max</tt>.
21015 They cannot meaningfully be <tt>constexpr</tt> without this change.
21016 ]</i></p>
21017
21018
21019
21020
21021
21022
21023 <hr>
21024 <h3><a name="1173"></a>1173. "Equivalence" wishy-washiness</h3>
21025 <p><b>Section:</b> 17 [library] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
21026  <b>Submitter:</b> David Abrahams <b>Opened:</b> 2009-07-14  <b>Last modified:</b> 2009-10-20</p>
21027 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#library">active issues</a> in [library].</p>
21028 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#library">issues</a> in [library].</p>
21029 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
21030 <p><b>Discussion:</b></p>
21031 <p>
21032 Issue: The <tt>CopyConstructible</tt> requirements are wishy-washy.  It requires
21033 that the copy is "equivalent" to the original, but "equivalent" is never
21034 defined.
21035 </p>
21036 <p>
21037 I believe this to be an example of a more general lack of rigor around
21038 copy and assignment, although I haven't done the research to dig up all
21039 the instances.
21040 </p>
21041 <p>
21042 It's a problem because if you don't know what <tt>CopyConstructible</tt> means,
21043 you also don't know what it means to copy a pair of <tt>CopyConstructible</tt>
21044 types.  It doesn't prevent us from writing code, but it is a hole in our
21045 ability to understand the meaning of copy.
21046 </p>
21047 <p>
21048 Furthermore, I'm pretty sure that vector's copy constructor doesn't
21049 require the elements to be <tt>EqualityComparable</tt>, so that table is actually
21050 referring to some ill-defined notion of equivalence when it uses ==.
21051 </p>
21052
21053 <p><i>[
21054 2009 Santa Cruz:
21055 ]</i></p>
21056
21057
21058 <blockquote>
21059 Move to "Open". Dave is right that this is a big issue. Paper D2987
21060 ("Defining Move Special Member Functions", Bjarne Stroustrup and
21061 Lawrence Crowl) touches on this but does not solve it. This issue is
21062 discussed in Elements of Programming.
21063 </blockquote>
21064
21065
21066
21067 <p><b>Proposed resolution:</b></p>
21068
21069
21070
21071
21072
21073 <hr>
21074 <h3><a name="1175"></a>1175. <tt>unordered</tt> complexity</h3>
21075 <p><b>Section:</b> 23.2.5 [unord.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
21076  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2009-07-17  <b>Last modified:</b> 2009-07-19</p>
21077 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
21078 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
21079 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
21080 <p><b>Discussion:</b></p>
21081 <p>
21082 When I look at the <tt>unordered_*</tt> constructors, I think the complexity is poorly
21083 described and does not follow the style of the rest of the standard.
21084 </p>
21085
21086 <p>
21087 The complexity for the default constructor is specified as constant.
21088  Actually, it is proportional to <tt>n</tt>, but there are no invocations of
21089 <tt>value_type</tt> constructors or other <tt>value_type</tt> operations.
21090 </p>
21091
21092 <p>
21093 For the iterator-based constructor the complexity should be:
21094 </p>
21095
21096 <blockquote>
21097 <i>Complexity:</i> exactly <tt>n</tt> calls to construct <tt>value_type</tt>
21098 from <tt>InputIterator::value_type</tt> (where <tt>n = distance(f,l)</tt>).
21099 The number of calls to <tt>key_equal::operator()</tt> is proportional to
21100 <tt>n</tt> in the average case and <tt>n*n</tt> in the worst case.
21101 </blockquote>
21102
21103
21104
21105 <p><b>Proposed resolution:</b></p>
21106 <p>
21107 </p>
21108
21109
21110
21111
21112
21113 <hr>
21114 <h3><a name="1176"></a>1176. Make <tt>thread</tt> constructor non-variadic</h3>
21115 <p><b>Section:</b> 30.3.1.2 [thread.thread.constr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
21116  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-07-18  <b>Last modified:</b> 2009-07-18</p>
21117 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.thread.constr">active issues</a> in [thread.thread.constr].</p>
21118 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.thread.constr">issues</a> in [thread.thread.constr].</p>
21119 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
21120 <p><b>Discussion:</b></p>
21121 <p>
21122 The variadic <tt>thread</tt> constructor is causing controversy, e.g.
21123 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2901.pdf">N2901</a>.
21124 This issue has been created as a placeholder for this course of action.
21125 </p>
21126
21127 <blockquote><pre>template &lt;class F<del>, class ...Args</del>&gt; thread(F&amp;&amp; f<del>, Args&amp;&amp;... args</del>);
21128 </pre></blockquote>
21129
21130 <p>
21131 See <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#929">929</a> for wording which specifies an rvalue-ref signature but
21132 with "decay behavior", but using variadics.
21133 </p>
21134
21135
21136 <p><b>Proposed resolution:</b></p>
21137 <p>
21138 </p>
21139
21140
21141
21142
21143
21144 <hr>
21145 <h3><a name="1177"></a>1177. Improve "diagnostic required" wording</h3>
21146 <p><b>Section:</b> 20.9.3 [time.duration] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
21147  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-07-18  <b>Last modified:</b> 2009-10-26</p>
21148 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#time.duration">active issues</a> in [time.duration].</p>
21149 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#time.duration">issues</a> in [time.duration].</p>
21150 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
21151 <p><b>Discussion:</b></p>
21152 <p>
21153 "diagnostic required" has been used (by me) for code words meaning "use
21154 <tt>enable_if</tt> to constrain templated functions.  This needs to be
21155 improved by referring to the function signature as not participating in
21156 the overload set, and moving this wording to a <i>Remarks</i> paragraph.
21157 </p>
21158
21159 <p><i>[
21160 2009-10 Santa Cruz:
21161 ]</i></p>
21162
21163
21164 <blockquote>
21165 Moved to Ready.
21166 </blockquote>
21167
21168
21169
21170 <p><b>Proposed resolution:</b></p>
21171 <p><i>[
21172 This proposed resolution addresses <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#947">947</a> and <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#974">974</a>.
21173 ]</i></p>
21174
21175
21176 <ol>
21177 <li>
21178 <p>
21179 Change 20.9.3.1 [time.duration.cons]:
21180 </p>
21181
21182 <blockquote>
21183 <pre>template &lt;class Rep2&gt;
21184   explicit duration(const Rep2&amp; r);
21185 </pre>
21186 <blockquote>
21187 <p>
21188 <i><del>Requires:</del> <ins>Remarks:</ins></i>
21189 <tt>Rep2</tt> shall be implicitly convertible to <tt>rep</tt> and
21190 </p>
21191 <ul>
21192 <li>
21193 <tt>treat_as_floating_point&lt;rep&gt;::value</tt> shall be <tt>true</tt> or
21194 </li>
21195 <li>
21196 <tt>treat_as_floating_point&lt;Rep2&gt;::value</tt> shall be <tt>false</tt>.
21197 </li>
21198 </ul>
21199 <p>
21200 <del>Diagnostic required</del> <ins>If these constraints are not met, this
21201 constructor shall not participate in overload resolution</ins>. [<i>Example:</i>
21202 </p>
21203 <blockquote><pre>duration&lt;int, milli&gt; d(3); // OK
21204 duration&lt;int, milli&gt; d(3.5); // error
21205 </pre></blockquote>
21206
21207 <p>
21208 &#8212; <i>end example</i>]
21209 </p>
21210
21211 <p>
21212 <i>Effects:</i> Constructs an object of type <tt>duration</tt>.
21213 </p>
21214
21215 <p>
21216 <i>Postcondition:</i> <tt>count() == static_cast&lt;rep&gt;(r)</tt>.
21217 </p>
21218
21219 </blockquote>
21220
21221 <pre>template &lt;class Rep2, class Period2&gt;
21222   duration(const duration&lt;Rep2, Period2&gt;&amp; d);
21223 </pre>
21224 <blockquote>
21225 <p>
21226 <i><del>Requires:</del> <ins>Remarks:</ins></i> <tt>treat_as_floating_point&lt;rep&gt;::value</tt> shall be <tt>true</tt> or
21227 <tt>ratio_divide&lt;Period2, period&gt;::type::den</tt> shall be 1<del>. Diagnostic
21228 required</del><ins>, else this constructor shall not participate in overload
21229 resolution</ins>. [<i>Note:</i> This requirement prevents implicit truncation error
21230 when converting between integral-based duration types. Such a
21231 construction could easily lead to confusion about the value of the
21232 duration. &#8212; <i>end note</i>] [<i>Example:</i>
21233 </p>
21234
21235 <blockquote><pre>duration&lt;int, milli&gt; ms(3);
21236 duration&lt;int, micro&gt; us = ms; // OK
21237 duration&lt;int, milli&gt; ms2 = us; // error
21238 </pre></blockquote>
21239
21240 <p>
21241 &#8212; <i>end example</i>]
21242 </p>
21243
21244 <p>
21245 <i>Effects:</i> Constructs an object of type <tt>duration</tt>, constructing
21246 <tt>rep_</tt> from
21247 <tt>duration_cast&lt;duration&gt;(d).count()</tt>.
21248 </p>
21249
21250 </blockquote>
21251
21252
21253 </blockquote>
21254 </li>
21255
21256 <li>
21257 <p>
21258 Change the following paragraphs in 20.9.3.5 [time.duration.nonmember]:
21259 </p>
21260
21261 <blockquote>
21262 <pre>template &lt;class Rep1, class Period, class Rep2&gt;
21263   duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
21264   operator*(const duration&lt;Rep1, Period&gt;&amp; d, const Rep2&amp; s);
21265 </pre>
21266 <blockquote>
21267 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>Rep2</tt> shall be implicitly convertible to
21268 <tt>CR(Rep1, Rep2)</tt><ins>, else this signature shall not participate in
21269 overload resolution</ins>. <del>Diagnostic required.</del>
21270 </blockquote>
21271
21272 <pre>template &lt;class Rep1, class Period, class Rep2&gt;
21273   duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
21274   operator*(const Rep1&amp; s, const duration&lt;Rep2, Period&gt;&amp; d);
21275 </pre>
21276 <blockquote>
21277 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>Rep1</tt> shall be implicitly convertible to
21278 <tt>CR(Rep1, Rep2)</tt><ins>, else this signature shall not participate in
21279 overload resolution</ins>. <del>Diagnostic required.</del>
21280 </blockquote>
21281
21282 <pre>template &lt;class Rep1, class Period, class Rep2&gt;
21283   duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
21284   operator/(const duration&lt;Rep1, Period&gt;&amp; d, const Rep2&amp; s);
21285 </pre>
21286 <blockquote>
21287 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>Rep2</tt> shall be implicitly convertible to
21288 <tt>CR(Rep1, Rep2)</tt> and <tt>Rep2</tt> shall not be an instantiation of
21289 <tt>duration</tt><ins>, else this signature shall not participate in
21290 overload resolution</ins>. <del>Diagnostic required.</del>
21291 </blockquote>
21292
21293 <pre>template &lt;class Rep1, class Period, class Rep2&gt;
21294   duration&lt;typename common_type&lt;Rep1, Rep2&gt;::type, Period&gt;
21295   operator%(const duration&lt;Rep1, Period&gt;&amp; d, const Rep2&amp; s);
21296 </pre>
21297 <blockquote>
21298 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>Rep2</tt> shall be implicitly convertible to
21299 <tt>CR(Rep1, Rep2)</tt> and <tt>Rep2</tt> shall not be an instantiation of
21300 <tt>duration</tt><ins>, else this signature shall not participate in
21301 overload resolution</ins>. <del>Diagnostic required.</del>
21302 </blockquote>
21303
21304 </blockquote>
21305 </li>
21306
21307 <li>
21308 <p>
21309 Change the following paragraphs in 20.9.3.7 [time.duration.cast]:
21310 </p>
21311
21312 <blockquote><pre>template &lt;class ToDuration, class Rep, class Period&gt;
21313   ToDuration duration_cast(const duration&lt;Rep, Period&gt;&amp; d);
21314 </pre>
21315
21316 <blockquote>
21317 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>ToDuration</tt> shall be an instantiation of
21318 <tt>duration</tt><ins>, else this signature shall not participate in
21319 overload resolution</ins>. <del>Diagnostic required.</del>
21320 </blockquote>
21321 </blockquote>
21322 </li>
21323
21324 <li>
21325 <p>
21326 Change the following paragraphs in 20.9.4.7 [time.point.cast]:
21327 </p>
21328
21329 <blockquote><pre>template &lt;class ToDuration, class Clock, class Duration&gt;
21330   time_point&lt;Clock, ToDuration&gt; time_point_cast(const time_point&lt;Clock, Duration&gt;&amp; t);
21331 </pre>
21332
21333 <blockquote>
21334 <i><del>Requires</del> <ins>Remarks</ins>:</i> <tt>ToDuration</tt> shall be an instantiation of
21335 <tt>duration</tt><ins>, else this signature shall not participate in
21336 overload resolution</ins>. <del>Diagnostic required.</del>
21337 </blockquote>
21338 </blockquote>
21339 </li>
21340 </ol>
21341
21342
21343
21344
21345
21346
21347 <hr>
21348 <h3><a name="1180"></a>1180. Missing string_type member typedef in class <tt>sub_match</tt></h3>
21349 <p><b>Section:</b> 28.9.1 [re.submatch.members] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
21350  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-07-25  <b>Last modified:</b> 2009-07-26</p>
21351 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
21352 <p><b>Discussion:</b></p>
21353 <p>
21354 The definition of class template <tt>sub_match</tt> is strongly dependent
21355 on the type <tt>basic_string&lt;value_type&gt;</tt>, both in interface and effects,
21356 but does not provide a corresponding typedef <tt>string_type</tt>, as e.g.
21357 class <tt>match_results</tt> does, which looks like an oversight to me that
21358 should be fixed.
21359 </p>
21360
21361
21362 <p><b>Proposed resolution:</b></p>
21363
21364 <ol>
21365 <li>
21366 <p>
21367 In the class template <tt>sub_match</tt> synopsis 28.9 [re.submatch]/1
21368 change as indicated:
21369 </p>
21370
21371 <blockquote><pre>template &lt;class BidirectionalIterator&gt;
21372 class sub_match : public std::pair&lt;BidirectionalIterator, BidirectionalIterator&gt; {
21373 public:
21374   typedef typename iterator_traits&lt;BidirectionalIterator&gt;::value_type value_type;
21375   typedef typename iterator_traits&lt;BidirectionalIterator&gt;::difference_type difference_type;
21376   typedef BidirectionalIterator iterator;
21377   <ins>typedef basic_string&lt;value_type&gt; string_type;</ins>
21378
21379   bool matched;
21380
21381   difference_type length() const;
21382   operator <del>basic_string&lt;value_type&gt;</del><ins>string_type</ins>() const;
21383   <del>basic_string&lt;value_type&gt;</del><ins>string_type</ins> str() const;
21384   int compare(const sub_match&amp; s) const;
21385   int compare(const <del>basic_string&lt;value_type&gt;</del><ins>string_type</ins>&amp; s) const;
21386   int compare(const value_type* s) const;
21387 };
21388 </pre></blockquote>
21389 </li>
21390
21391 <li>
21392 <p>
21393 In 28.9.1 [re.submatch.members]/2 change as indicated:
21394 </p>
21395
21396 <blockquote><pre>operator <del>basic_string&lt;value_type&gt;</del><ins>string_type</ins>() const;
21397 </pre>
21398
21399 <blockquote>
21400 <i>Returns:</i> <tt>matched ? <del>basic_string&lt;value_type&gt;</del>
21401 <ins>string_type</ins>(first, second) : <del>basic_string&lt;value_type&gt;</del>
21402 <ins>string_type</ins>()</tt>.
21403 </blockquote>
21404 </blockquote>
21405 </li>
21406
21407 <li>
21408 <p>
21409 In 28.9.1 [re.submatch.members]/3 change as indicated:
21410 </p>
21411
21412 <blockquote><pre><del>basic_string&lt;value_type&gt;</del><ins>string_type</ins> str() const;
21413 </pre>
21414
21415 <blockquote>
21416 <i>Returns:</i> <tt>matched ? <del>basic_string&lt;value_type&gt;</del>
21417 <ins>string_type</ins>(first, second) : <del>basic_string&lt;value_type&gt;</del>
21418 <ins>string_type</ins>()</tt>.
21419 </blockquote>
21420 </blockquote>
21421 </li>
21422 </ol>
21423
21424
21425
21426
21427
21428
21429 <hr>
21430 <h3><a name="1181"></a>1181. Invalid <tt>sub_match</tt> comparison operators</h3>
21431 <p><b>Section:</b> 28.9.2 [re.submatch.op] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
21432  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-07-25  <b>Last modified:</b> 2009-07-28</p>
21433 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#re.submatch.op">issues</a> in [re.submatch.op].</p>
21434 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
21435 <p><b>Discussion:</b></p>
21436 <p>
21437 Several heterogeneous comparison operators of class template
21438 <tt>sub_match</tt> are specified by return clauses that are not valid
21439 in general. E.g. 28.9.2 [re.submatch.op]/7:
21440 </p>
21441
21442 <blockquote><pre>template &lt;class BiIter, class ST, class SA&gt;
21443 bool operator==(
21444   const basic_string&lt;
21445     typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21446   const sub_match&lt;BiIter&gt;&amp; rhs);
21447 </pre>
21448 <blockquote>
21449 <i>Returns:</i> <tt>lhs == rhs.str()</tt>.
21450 </blockquote>
21451 </blockquote>
21452
21453 <p>
21454 The returns clause would be ill-formed for all cases where
21455 <tt>ST != std::char_traits&lt;iterator_traits&lt;BiIter&gt;::value_type&gt;</tt>
21456 or <tt>SA != std::allocator&lt;iterator_traits&lt;BiIter&gt;::value_type&gt;</tt>.
21457 </p>
21458 <p>
21459 The generic character of the comparison was intended, so
21460 there are basically two approaches to fix the problem: The
21461 first one would define the semantics of the comparison
21462 using the traits class <tt>ST</tt> (The semantic of <tt>basic_string::compare</tt>
21463 is defined in terms of the compare function of the corresponding
21464 traits class), the second one would define the semantics of the
21465 comparison using the traits class
21466 </p>
21467
21468 <blockquote><pre>std::char_traits&lt;iterator_traits&lt;BiIter&gt;::value_type&gt;
21469 </pre></blockquote>
21470
21471 <p>
21472 which is essentially identical to
21473 </p>
21474
21475 <blockquote><pre>std::char_traits&lt;sub_match&lt;BiIter&gt;::value_type&gt;
21476 </pre></blockquote>
21477
21478 <p>
21479 I suggest to follow the second approach, because
21480 this emphasizes the central role of the <tt>sub_match</tt>
21481 object as part of the comparison and would also
21482 make sure that a <tt>sub_match</tt> comparison using some
21483 <tt>basic_string&lt;char_t, ..&gt;</tt> always is equivalent to
21484 a corresponding comparison with a string literal
21485 because of the existence of further overloads (beginning
21486 from 28.9.2 [re.submatch.op]/19). If users really want to
21487 take advantage of their own <tt>traits::compare</tt>, they can
21488 simply write a corresponding compare function that
21489 does so.
21490 </p>
21491
21492
21493 <p><b>Proposed resolution:</b></p>
21494 <ol>
21495 <li>
21496 <p>
21497 In 28.9.2 [re.submatch.op] change as indicated:
21498 </p>
21499
21500 <ol type="a">
21501 <li>
21502
21503 <p>
21504 If <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1180">1180</a> is accepted:
21505 </p>
21506
21507 <blockquote>
21508 <pre>template &lt;class BiIter, class ST, class SA&gt;
21509   bool operator==(
21510     const basic_string&lt;
21511       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21512     const sub_match&lt;BiIter&gt;&amp; rhs);
21513 </pre>
21514 <blockquote>
21515 7 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21516 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> ==
21517 rhs.str()</tt>.
21518 </blockquote>
21519
21520 <pre>template &lt;class BiIter, class ST, class SA&gt;
21521   bool operator!=(
21522     const basic_string&lt;
21523       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21524     const sub_match&lt;BiIter&gt;&amp; rhs);
21525 </pre>
21526
21527 <blockquote>
21528 8 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21529 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> !=
21530 rhs.str()</tt>.
21531 </blockquote>
21532
21533 <pre>template &lt;class BiIter, class ST, class SA&gt;
21534   bool operator&lt;(
21535     const basic_string&lt;
21536       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21537     const sub_match&lt;BiIter&gt;&amp; rhs);
21538 </pre>
21539
21540 <blockquote>
21541 9 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21542 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> &lt;
21543 rhs.str()</tt>.
21544 </blockquote>
21545
21546 <pre>template &lt;class BiIter, class ST, class SA&gt;
21547   bool operator&gt;(
21548     const basic_string&lt;
21549       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21550     const sub_match&lt;BiIter&gt;&amp; rhs);
21551 </pre>
21552
21553 <blockquote>
21554 10 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21555 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> &gt;
21556 rhs.str()</tt>.
21557 </blockquote>
21558
21559 <pre>template &lt;class BiIter, class ST, class SA&gt;
21560   bool operator&gt;=(
21561     const basic_string&lt;
21562       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21563     const sub_match&lt;BiIter&gt;&amp; rhs);
21564 </pre>
21565
21566 <blockquote>
21567 11 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21568 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> &gt;=
21569 rhs.str()</tt>.
21570 </blockquote>
21571
21572 <pre>template &lt;class BiIter, class ST, class SA&gt;
21573   bool operator&lt;=(
21574     const basic_string&lt;
21575       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21576     const sub_match&lt;BiIter&gt;&amp; rhs);
21577 </pre>
21578
21579 <blockquote>
21580 12 <i>Returns:</i> <tt><del>lhs</del><ins>typename
21581 sub_match&lt;BiIter&gt;::string_type(lhs.begin(), lhs.end())</ins> &lt;=
21582 rhs.str()</tt>.
21583 </blockquote>
21584
21585 <pre>template &lt;class BiIter, class ST, class SA&gt;
21586   bool operator==(const sub_match&lt;BiIter&gt;&amp; lhs,
21587     const basic_string&lt;
21588       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21589 </pre>
21590
21591 <blockquote>
21592 13 <i>Returns:</i> <tt>lhs.str() == <del>rhs</del><ins>typename
21593 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21594 </blockquote>
21595
21596 <pre>template &lt;class BiIter, class ST, class SA&gt;
21597   bool operator!=(const sub_match&lt;BiIter&gt;&amp; lhs,
21598     const basic_string&lt;
21599       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21600 </pre>
21601
21602 <blockquote>
21603 14 <i>Returns:</i> <tt>lhs.str() != <del>rhs</del><ins>typename
21604 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21605 </blockquote>
21606
21607 <pre>template &lt;class BiIter, class ST, class SA&gt;
21608   bool operator&lt;(const sub_match&lt;BiIter&gt;&amp; lhs,
21609     const basic_string&lt;
21610       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21611 </pre>
21612
21613 <blockquote>
21614 15 <i>Returns:</i> <tt>lhs.str() &lt; <del>rhs</del><ins>typename
21615 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21616 </blockquote>
21617
21618 <pre>template &lt;class BiIter, class ST, class SA&gt;
21619   bool operator&gt;(const sub_match&lt;BiIter&gt;&amp; lhs,
21620     const basic_string&lt;
21621       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21622 </pre>
21623
21624 <blockquote>
21625 16 <i>Returns:</i> <tt>lhs.str() &gt; <del>rhs</del><ins>typename
21626 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21627 </blockquote>
21628
21629 <pre>template &lt;class BiIter, class ST, class SA&gt;
21630   bool operator&gt;=(const sub_match&lt;BiIter&gt;&amp; lhs,
21631     const basic_string&lt;
21632       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21633 </pre>
21634
21635 <blockquote>
21636 17 <i>Returns:</i> <tt>lhs.str() &gt;= <del>rhs</del><ins>typename
21637 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21638 </blockquote>
21639
21640 <pre>template &lt;class BiIter, class ST, class SA&gt;
21641   bool operator&lt;=(const sub_match&lt;BiIter&gt;&amp; lhs,
21642     const basic_string&lt;
21643       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21644 </pre>
21645
21646 <blockquote>
21647 18 <i>Returns:</i> <tt>lhs.str() &lt;= <del>rhs</del><ins>typename
21648 sub_match&lt;BiIter&gt;::string_type(rhs.begin(), rhs.end())</ins></tt>.
21649 </blockquote>
21650 </blockquote>
21651
21652 </li>
21653
21654 <li>
21655
21656 <p>
21657 If <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1180">1180</a> is <em>not</em> accepted:
21658 </p>
21659
21660 <blockquote>
21661 <pre>template &lt;class BiIter, class ST, class SA&gt;
21662   bool operator==(
21663     const basic_string&lt;
21664       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21665     const sub_match&lt;BiIter&gt;&amp; rhs);
21666 </pre>
21667 <blockquote>
21668 7 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21669 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> ==
21670 rhs.str()</tt>.
21671 </blockquote>
21672
21673 <pre>template &lt;class BiIter, class ST, class SA&gt;
21674   bool operator!=(
21675     const basic_string&lt;
21676       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21677     const sub_match&lt;BiIter&gt;&amp; rhs);
21678 </pre>
21679
21680 <blockquote>
21681 8 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21682 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> !=
21683 rhs.str()</tt>.
21684 </blockquote>
21685
21686 <pre>template &lt;class BiIter, class ST, class SA&gt;
21687   bool operator&lt;(
21688     const basic_string&lt;
21689       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21690     const sub_match&lt;BiIter&gt;&amp; rhs);
21691 </pre>
21692
21693 <blockquote>
21694 9 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21695 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> &lt;
21696 rhs.str()</tt>.
21697 </blockquote>
21698
21699 <pre>template &lt;class BiIter, class ST, class SA&gt;
21700   bool operator&gt;(
21701     const basic_string&lt;
21702       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21703     const sub_match&lt;BiIter&gt;&amp; rhs);
21704 </pre>
21705
21706 <blockquote>
21707 10 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21708 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> &gt;
21709 rhs.str()</tt>.
21710 </blockquote>
21711
21712 <pre>template &lt;class BiIter, class ST, class SA&gt;
21713   bool operator&gt;=(
21714     const basic_string&lt;
21715       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21716     const sub_match&lt;BiIter&gt;&amp; rhs);
21717 </pre>
21718
21719 <blockquote>
21720 11 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21721 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> &gt;=
21722 rhs.str()</tt>.
21723 </blockquote>
21724
21725 <pre>template &lt;class BiIter, class ST, class SA&gt;
21726   bool operator&lt;=(
21727     const basic_string&lt;
21728       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; lhs,
21729     const sub_match&lt;BiIter&gt;&amp; rhs);
21730 </pre>
21731
21732 <blockquote>
21733 12 <i>Returns:</i> <tt><del>lhs</del><ins>basic_string&lt;typename
21734 sub_match&lt;BiIter&gt;::value_type&gt;(lhs.begin(), lhs.end())</ins> &lt;=
21735 rhs.str()</tt>.
21736 </blockquote>
21737
21738 <pre>template &lt;class BiIter, class ST, class SA&gt;
21739   bool operator==(const sub_match&lt;BiIter&gt;&amp; lhs,
21740     const basic_string&lt;
21741       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21742 </pre>
21743
21744 <blockquote>
21745 13 <i>Returns:</i> <tt>lhs.str() == <del>rhs</del><ins>basic_string&lt;typename
21746 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21747 </blockquote>
21748
21749 <pre>template &lt;class BiIter, class ST, class SA&gt;
21750   bool operator!=(const sub_match&lt;BiIter&gt;&amp; lhs,
21751     const basic_string&lt;
21752       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21753 </pre>
21754
21755 <blockquote>
21756 14 <i>Returns:</i> <tt>lhs.str() != <del>rhs</del><ins>basic_string&lt;typename
21757 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21758 </blockquote>
21759
21760 <pre>template &lt;class BiIter, class ST, class SA&gt;
21761   bool operator&lt;(const sub_match&lt;BiIter&gt;&amp; lhs,
21762     const basic_string&lt;
21763       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21764 </pre>
21765
21766 <blockquote>
21767 15 <i>Returns:</i> <tt>lhs.str() &lt; <del>rhs</del><ins>basic_string&lt;typename
21768 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21769 </blockquote>
21770
21771 <pre>template &lt;class BiIter, class ST, class SA&gt;
21772   bool operator&gt;(const sub_match&lt;BiIter&gt;&amp; lhs,
21773     const basic_string&lt;
21774       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21775 </pre>
21776
21777 <blockquote>
21778 16 <i>Returns:</i> <tt>lhs.str() &gt; <del>rhs</del><ins>basic_string&lt;typename
21779 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21780 </blockquote>
21781
21782 <pre>template &lt;class BiIter, class ST, class SA&gt;
21783   bool operator&gt;=(const sub_match&lt;BiIter&gt;&amp; lhs,
21784     const basic_string&lt;
21785       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21786 </pre>
21787
21788 <blockquote>
21789 17 <i>Returns:</i> <tt>lhs.str() &gt;= <del>rhs</del><ins>basic_string&lt;typename
21790 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21791 </blockquote>
21792
21793 <pre>template &lt;class BiIter, class ST, class SA&gt;
21794   bool operator&lt;=(const sub_match&lt;BiIter&gt;&amp; lhs,
21795     const basic_string&lt;
21796       typename iterator_traits&lt;BiIter&gt;::value_type, ST, SA&gt;&amp; rhs);
21797 </pre>
21798
21799 <blockquote>
21800 18 <i>Returns:</i> <tt>lhs.str() &lt;= <del>rhs</del><ins>basic_string&lt;typename
21801 sub_match&lt;BiIter&gt;::value_type&gt;(rhs.begin(), rhs.end())</ins></tt>.
21802 </blockquote>
21803 </blockquote>
21804
21805 </li>
21806
21807 </ol>
21808
21809 </li>
21810 </ol>
21811
21812
21813
21814
21815
21816 <hr>
21817 <h3><a name="1182"></a>1182. Unfortunate hash dependencies</h3>
21818 <p><b>Section:</b> 20.7.16 [unord.hash] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
21819  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-07-28  <b>Last modified:</b> 2009-09-21</p>
21820 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.hash">active issues</a> in [unord.hash].</p>
21821 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.hash">issues</a> in [unord.hash].</p>
21822 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
21823 <p><b>Discussion:</b></p>
21824 <p>
21825 The implied library dependencies created by spelling out all the <tt>hash</tt>
21826 template specializations in the <tt>&lt;functional&gt;</tt> synopsis are unfortunate.
21827 The potential coupling is greatly reduced if the <tt>hash</tt> specialization is
21828 declared in the appropriate header for each library type, as it is much
21829 simpler to forward declare the primary template and provide a single
21830 specialization than it is to implement a <tt>hash</tt> function for a <tt>string</tt> or
21831 <tt>vector</tt> without providing a definition for the whole <tt>string/vector</tt>
21832 template in order to access the necessary bits.
21833 </p>
21834
21835 <p>
21836 Note that the proposed resolution purely involves moving the
21837 declarations of a few specializations, it specifically does not make any
21838 changes to 20.7.16 [unord.hash].
21839 </p>
21840
21841 <p><i>[
21842 2009-09-15 Daniel adds:
21843 ]</i></p>
21844
21845
21846 <blockquote>
21847 </blockquote>
21848 <p>
21849 I suggest to add to the current existing
21850 proposed resolution the following items.
21851 </p>
21852
21853 <ul>
21854 <li>
21855 <p>
21856 Add to the very first strike-list of the currently suggested resolution
21857 the following lines:
21858 </p>
21859
21860 <blockquote><pre><del>template &lt;&gt; struct hash&lt;std::error_code&gt;;</del>
21861 <del>template &lt;&gt; struct hash&lt;std::thread::id&gt;;</del>
21862 </pre></blockquote>
21863 </li>
21864
21865 <li>
21866 <p>
21867 Add the following declarations to 19.5 [syserr], header
21868 <tt>&lt;system_error&gt;</tt> synopsis after // 19.5.4:
21869 </p>
21870
21871 <blockquote><pre><ins>
21872 // 19.5.x hash support
21873 template &lt;class T&gt; struct hash;
21874 template &lt;&gt; struct hash&lt;error_code&gt;;
21875 </ins>
21876 </pre></blockquote>
21877 </li>
21878
21879 <li>
21880 <p>
21881 Add a new clause 19.5.X (probably after 19.5.4):
21882 </p>
21883
21884 <blockquote>
21885 <p><ins>
21886 19.5.X Hash support [syserr.hash]
21887 </ins></p>
21888
21889 <pre><ins>
21890 template &lt;&gt; struct hash&lt;error_code&gt;;
21891 </ins></pre>
21892
21893 <blockquote><ins>
21894 An explicit specialization of the class template hash (20.7.16 [unord.hash])
21895 shall be provided
21896 for the type <tt>error_code</tt> suitable for using this type as key in
21897 unordered associative
21898 containers (23.5 [unord]).
21899 </ins></blockquote>
21900 </blockquote>
21901 </li>
21902
21903 <li>
21904 <p>
21905 Add the following declarations to 30.3.1.1 [thread.thread.id] just after the
21906 declaration of
21907 the comparison operators:
21908 </p>
21909
21910 <blockquote><pre><ins>
21911 template &lt;class T&gt; struct hash;
21912 template &lt;&gt; struct hash&lt;thread::id&gt;;
21913 </ins></pre></blockquote>
21914 </li>
21915
21916 <li>
21917 <p>
21918 Add a new paragraph at the end of 30.3.1.1 [thread.thread.id]:
21919 </p>
21920
21921 <blockquote>
21922 <pre><ins>
21923 template &lt;&gt; struct hash&lt;thread::id&gt;;
21924 </ins></pre>
21925
21926 <blockquote><ins>
21927 An explicit specialization of the class template hash (20.7.16 [unord.hash])
21928 shall be provided
21929 for the type <tt>thread::id</tt> suitable for using this type as key in
21930 unordered associative
21931 containers (23.5 [unord]).
21932 </ins></blockquote>
21933 </blockquote>
21934 </li>
21935
21936 <li>
21937 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#889">889</a> independently suggests moving the specialization
21938 <tt>std::hash&lt;std::thread::id&gt;</tt> to header <tt>&lt;thread&gt;</tt>.
21939 </li>
21940 </ul>
21941
21942
21943
21944 <p><b>Proposed resolution:</b></p>
21945 <p>
21946 Strike the following specializations declared in the <tt>&lt;functional&gt;</tt>
21947 synopsis p2 20.7 [function.objects]
21948 </p>
21949
21950 <blockquote><pre><del>template &lt;&gt; struct hash&lt;std::string&gt;;</del>
21951 <del>template &lt;&gt; struct hash&lt;std::u16string&gt;;</del>
21952 <del>template &lt;&gt; struct hash&lt;std::u32string&gt;;</del>
21953 <del>template &lt;&gt; struct hash&lt;std::wstring&gt;;</del>
21954
21955 <del>template &lt;class Allocator&gt; struct hash&lt;std::vector&lt;bool, Allocator&gt; &gt;;</del>
21956 <del>template &lt;std::size_t N&gt; struct hash&lt;std::bitset&lt;N&gt; &gt;;</del>
21957 </pre></blockquote>
21958
21959 <p>
21960 Add the following declarations to the synopsis of <tt>&lt;string&gt;</tt> in
21961 21.3 [string.classes]
21962 </p>
21963
21964 <blockquote><pre><ins>// 21.4.x hash support
21965 template &lt;class T&gt; struct hash;
21966 template &lt;&gt; struct hash&lt;string&gt;;
21967 template &lt;&gt; struct hash&lt;u16string&gt;;
21968 template &lt;&gt; struct hash&lt;u32string&gt;;
21969 template &lt;&gt; struct hash&lt;wstring&gt;;</ins>
21970 </pre></blockquote>
21971
21972 <p>
21973 Add a new clause 21.4.X
21974 </p>
21975
21976 <blockquote>
21977 <p>
21978 21.4.X Hash support [basic.string.hash]
21979 </p>
21980
21981 <pre>template &lt;&gt; struct hash&lt;string&gt;;
21982 template &lt;&gt; struct hash&lt;u16string&gt;;
21983 template &lt;&gt; struct hash&lt;u32string&gt;;
21984 template &lt;&gt; struct hash&lt;wstring&gt;;
21985 </pre>
21986
21987 <blockquote>
21988 Explicit specializations of the class template hash (20.7.16 [unord.hash])
21989 shall be provided for the types <tt>string</tt>, <tt>u16string</tt>,
21990 <tt>u32string</tt> and <tt>wstring</tt> suitable for using these types as keys in
21991 unordered associative containers (23.5 [unord]).
21992 </blockquote>
21993 </blockquote>
21994
21995 <p>
21996 Add the following declarations to the synopsis of <tt>&lt;vector&gt;</tt> in
21997 23.3 [sequences]
21998 </p>
21999
22000 <blockquote><pre><ins>
22001 // 21.4.x hash support
22002 template &lt;class T&gt; struct hash;
22003 template &lt;class Allocator&gt; struct hash&lt;vector&lt;bool, Allocator&gt;&gt;;
22004 </ins></pre></blockquote>
22005
22006 <p>
22007 Add a new paragraph to the end of 23.3.7 [vector.bool]
22008 </p>
22009
22010 <blockquote><pre>template &lt;class Allocator&gt; struct hash&lt;vector&lt;bool, Allocator&gt;&gt;;
22011 </pre>
22012 <blockquote>
22013 A partial specialization of the class template hash (20.7.16 [unord.hash])
22014 shall be provided for vectors of boolean values suitable for use as a key
22015 in unordered associative containers (23.5 [unord]).
22016 </blockquote>
22017 </blockquote>
22018
22019 <p>
22020 Add the following declarations to the synopsis of <tt>&lt;bitset&gt;</tt>
22021 in 20.3.7 [template.bitset]
22022 </p>
22023
22024 <blockquote><pre><ins>
22025 // 20.3.6.X hash support
22026 template &lt;class T&gt; struct hash;
22027 template &lt;size_t N&gt; struct hash&lt;bitset&lt;N&gt; &gt;;
22028 </ins></pre></blockquote>
22029
22030 <p>
22031 Add a new subclause 20.3.6.X [bitset.hash]
22032 </p>
22033
22034 <blockquote>
22035 <p>
22036 20.3.6.X bitset hash support [bitset.hash]
22037 </p>
22038
22039 <pre>template &lt;size_t N&gt; struct hash&lt;bitset&lt;N&gt; &gt;;
22040 </pre>
22041
22042 <blockquote>
22043 A partial specialization of the class template hash
22044 (20.7.16 [unord.hash]) shall be provided for bitsets suitable for use as a key in
22045 unordered associative containers (23.5 [unord]).
22046 </blockquote>
22047 </blockquote>
22048
22049
22050
22051
22052
22053
22054 <hr>
22055 <h3><a name="1183"></a>1183. <tt>basic_ios::set_rdbuf</tt> may break class invariants</h3>
22056 <p><b>Section:</b> 27.5.4.2 [basic.ios.members] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
22057  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-07-28  <b>Last modified:</b> 2009-10-22</p>
22058 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#basic.ios.members">active issues</a> in [basic.ios.members].</p>
22059 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#basic.ios.members">issues</a> in [basic.ios.members].</p>
22060 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
22061 <p><b>Discussion:</b></p>
22062 <p>
22063 The protected member function <tt>set_rdbuf</tt> had been added during the
22064 process of adding move and swap semantics to IO classes. A relevant
22065 property of this function is described by it's effects in
22066 27.5.4.2 [basic.ios.members]/19:
22067 </p>
22068
22069 <blockquote>
22070 <i>Effects:</i> Associates the <tt>basic_streambuf</tt> object pointed to by sb with
22071 this stream without calling <tt>clear()</tt>.
22072 </blockquote>
22073
22074 <p>
22075 This means that implementors of or those who derive from existing IO classes
22076 could cause an internal state where the stream buffer could be 0, but the
22077 IO class has the state <tt>good()</tt>. This would break several currently existing
22078 implementations which rely on the fact that setting a stream buffer via the
22079 currently only ways, i.e. either by calling
22080 </p>
22081
22082 <blockquote><pre>void init(basic_streambuf&lt;charT,traits&gt;* sb);
22083 </pre></blockquote>
22084
22085 <p>
22086 or by calling
22087 </p>
22088
22089 <blockquote><pre>basic_streambuf&lt;charT,traits&gt;* rdbuf(basic_streambuf&lt;charT,traits&gt;* sb);
22090 </pre></blockquote>
22091
22092 <p>
22093 to set <tt>rdstate()</tt> to <tt>badbit</tt>, if the buffer is 0. This has the effect that many
22094 internal functions can simply check <tt>rdstate()</tt> instead of <tt>rdbuf()</tt> for being 0.
22095 </p>
22096
22097 <p>
22098 I therefore suggest that a requirement is added for callers of <tt>set_rdbuf</tt> to
22099 set a non-0 value.
22100 </p>
22101
22102 <p><i>[
22103 2009-10 Santa Cruz:
22104 ]</i></p>
22105
22106
22107 <blockquote>
22108 Moved to Open.  Martin volunteers to provide new wording, where
22109 <tt>set_rdbuf()</tt> sets the <tt>badbit</tt> but does not cause an
22110 exception to be thrown like a call to <tt>clear()</tt> would.
22111 </blockquote>
22112
22113 <p><i>[
22114 2009-10-20 Martin provides wording:
22115 ]</i></p>
22116
22117
22118
22119
22120 <p><b>Proposed resolution:</b></p>
22121 <p>
22122 Change 27.5.4.2 [basic.ios.members] around p. 19 as indicated:
22123 </p>
22124
22125 <blockquote><pre>void set_rdbuf(basic_streambuf&lt;charT, traits&gt;* sb);
22126 </pre>
22127
22128 <blockquote>
22129 <p><del>
22130 <i>Effects:</i> Associates the <tt>basic_streambuf</tt> object pointed
22131 to by <tt>sb</tt> with this stream without calling <tt>clear()</tt>.
22132 <i>Postconditions:</i> <tt>rdbuf() == sb</tt>.
22133 </del></p>
22134
22135 <p><ins>
22136 <i>Effects:</i> As if:
22137 </ins></p>
22138
22139 <blockquote><pre><ins>
22140 iostate state = rdstate();
22141 try { rdbuf(sb); }
22142 catch(ios_base::failure) {
22143    if (0 == (state &amp; ios_base::badbit))
22144        unsetf(badbit);
22145 }
22146 </ins></pre></blockquote>
22147
22148 <p>
22149 <i>Throws:</i> Nothing.
22150 </p>
22151
22152 </blockquote>
22153 </blockquote>
22154
22155
22156 <p><b>Rationale:</b></p>
22157 We need to be able to call <tt>set_rdbuf()</tt> on stream objects
22158 for which (<tt>rdbuf() == 0</tt>) holds without causing <tt>ios_base::failure</tt> to
22159 be thrown. We also don't want <tt>badbit</tt> to be set as a result of
22160 setting <tt>rdbuf()</tt> to 0 if it wasn't set before the call. This changed
22161 Effects clause maintains the current behavior (as of N2914) without
22162 requiring that <tt>sb</tt> be non-null.
22163
22164
22165
22166
22167
22168 <hr>
22169 <h3><a name="1185"></a>1185. iterator categories and output iterators</h3>
22170 <p><b>Section:</b> 24.2 [iterator.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
22171  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-07-31  <b>Last modified:</b> 2009-07-31</p>
22172 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
22173 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
22174 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
22175 <p><b>Discussion:</b></p>
22176 <p>
22177 (wording relative to
22178 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>
22179 pending new working paper)
22180 </p>
22181
22182 <p>
22183 According to p3 24.2 [iterator.requirements], Forward iterators,
22184 Bidirectional iterators and Random Access iterators all satisfy the
22185 requirements for an Output iterator:
22186 </p>
22187
22188 <blockquote>
22189 XXX iterators satisfy all the requirements of the input and output iterators
22190 and can be used whenever either kind is specified ...
22191 </blockquote>
22192
22193 <p>
22194 Meanwhile, p4 goes on to contradict this:
22195 </p>
22196
22197 <blockquote>
22198 Besides its category, a forward, bidirectional, or random access
22199 iterator can also be mutable or constant...
22200 </blockquote>
22201
22202 <blockquote>
22203 ... Constant iterators do not satisfy the requirements for output iterators
22204 </blockquote>
22205
22206 <p>
22207 The latter seems to be the overriding concern, as the iterator tag
22208 hierarchy does not define <tt>forward_iterator_tag</tt> as multiply derived from
22209 both <tt>input_iterator_tag</tt> and <tt>output_iterator_tag</tt>.
22210 </p>
22211
22212 <p>
22213 The work on concepts for iterators showed us that output iterator really
22214 is fundamentally a second dimension to the iterator categories, rather
22215 than part of the linear input -&gt; forward -&gt; bidirectional -&gt;
22216 random-access sequence.  It would be good to clear up these words to
22217 reflect that, and separately list output iterator requirements in the
22218 requires clauses for the appropriate algorithms and operations.
22219 </p>
22220
22221
22222 <p><b>Proposed resolution:</b></p>
22223
22224
22225
22226
22227
22228 <hr>
22229 <h3><a name="1186"></a>1186. Forward list could model a stack</h3>
22230 <p><b>Section:</b> 23.3.5.3 [stack] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Concepts">Tentatively NAD Concepts</a>
22231  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-07-31  <b>Last modified:</b> 2009-11-02</p>
22232 <p><b>Discussion:</b></p>
22233 <p>
22234 The library template <tt>forward_list</tt> could easily model the idea of a
22235 <tt>stack</tt>, where the operations work on the front of the list rather than
22236 the back.  However, the standard library <tt>stack</tt> adaptor cannot support
22237 this.
22238 </p>
22239
22240 <p>
22241 It would be relatively easy to write a partial specialization for <tt>stack</tt>
22242 to support <tt>forward_list</tt>, but that opens the question of which header to
22243 place it in.  A much better solution would be to add a <tt>concept_map</tt> for
22244 the <tt>StackLikeContainer</tt> concept to the <tt>&lt;forward_list&gt;</tt> header and then
22245 everything just works, including a user's own further uses in a
22246 stack-like context.
22247 </p>
22248
22249 <p>
22250 Therefore while I am submitting the issue now so that it is on record, I
22251 <em>strongly recommend</em> we resolve as "NAD Concepts" as any non-concepts
22252 based solution will be inferior to the final goal, and the feature is
22253 not so compelling it must be supported ahead of the concepts-based
22254 library.
22255 </p>
22256
22257 <p><i>[
22258 2009-11-02 Howard adds:
22259 ]</i></p>
22260
22261
22262 <blockquote>
22263 Moved to Tentatively NAD Concepts after 5 positive votes on c++std-lib.
22264 </blockquote>
22265
22266
22267 <p><b>Proposed resolution:</b></p>
22268
22269
22270
22271
22272
22273 <hr>
22274 <h3><a name="1187"></a>1187. std::decay</h3>
22275 <p><b>Section:</b> 20.6.7 [meta.trans.other] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
22276  <b>Submitter:</b> Jason Merrill <b>Opened:</b> 2009-08-07  <b>Last modified:</b> 2009-08-22</p>
22277 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.trans.other">issues</a> in [meta.trans.other].</p>
22278 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
22279 <p><b>Discussion:</b></p>
22280 <p>
22281 I notice that <tt>std::decay</tt> is specified to strip the cv-quals from
22282 anything but an array or pointer.  This seems incorrect for values of
22283 class type, since class rvalues can have cv-qualified type (3.10 [basic.lval]/9).
22284 </p>
22285
22286 <p><i>[
22287 2009-08-09 Howard adds:
22288 ]</i></p>
22289
22290
22291 <blockquote>
22292 See the thread starting with c++std-lib-24568 for further discussion.  And
22293 here is a convenience link to the
22294 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2006/n2069.html">original proposal</a>.
22295 Also see the closely related issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#705">705</a>.
22296 </blockquote>
22297
22298
22299 <p><b>Proposed resolution:</b></p>
22300
22301 <p>
22302 Add a note to <tt>decay</tt> in 20.6.7 [meta.trans.other]:
22303 </p>
22304
22305 <blockquote>
22306 [<i>Note:</i> This behavior is similar to the lvalue-to-rvalue (4.1),
22307 array-to-pointer (4.2), and function-to-pointer (4.3) conversions
22308 applied when an lvalue expression is used as an rvalue, but also strips
22309 cv-qualifiers from class types in order to more closely model by-value
22310 argument passing. &#8212; <i>end note</i>]
22311 </blockquote>
22312
22313
22314
22315
22316
22317
22318
22319
22320 <hr>
22321 <h3><a name="1188"></a>1188. Unordered containers should have a minimum load factor as well as a maximum</h3>
22322 <p><b>Section:</b> 23.2.5 [unord.req], 23.5 [unord] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
22323  <b>Submitter:</b> Matt Austern <b>Opened:</b> 2009-08-10  <b>Last modified:</b> 2009-08-11</p>
22324 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
22325 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
22326 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
22327 <p><b>Discussion:</b></p>
22328 <p>
22329 Unordered associative containers have a notion of a maximum load factor:
22330 when the number of elements grows large enough, the containers
22331 automatically perform a rehash so that the number of elements per bucket
22332 stays below a user-specified bound. This ensures that the hash table's
22333 performance characteristics don't change dramatically as the size
22334 increases.
22335 </p>
22336
22337 <p>
22338 For similar reasons, Google has found it useful to specify a minimum
22339 load factor: when the number of elements shrinks by a large enough, the
22340 containers automatically perform a rehash so that the number of elements
22341 per bucket stays above a user-specified bound. This is useful for two
22342 reasons. First, it prevents wasting a lot of memory when an unordered
22343 associative container grows temporarily. Second, it prevents amortized
22344 iteration time from being arbitrarily large; consider the case of a hash
22345 table with a billion buckets and only one element. (This was discussed
22346 even before TR1 was published; it was TR issue 6.13, which the LWG
22347 closed as NAD on the grounds that it was a known design feature.
22348 However, the LWG did not consider the approach of a minimum load
22349 factor.)
22350 </p>
22351
22352 <p>
22353 The only interesting question is when shrinking is allowed. In principle
22354 the cleanest solution would be shrinking on erase, just as we grow on
22355 insert. However, that would be a usability problem; it would break a
22356 number of common idioms involving erase. Instead, Google's hash tables
22357 only shrink on insert and rehash.
22358 </p>
22359
22360 <p>
22361 The proposed resolution allows, but does not require, shrinking in
22362 rehash, mostly because a postcondition for rehash that involves the
22363 minimum load factor would be fairly complicated. (It would probably have
22364 to involve a number of special cases and it would probably have to
22365 mention yet another parameter, a minimum bucket count.)
22366 </p>
22367
22368 <p>
22369 The current behavior is equivalent to a minimum load factor of 0. If we
22370 specify that 0 is the default, this change will have no impact on
22371 backward compatibility.
22372 </p>
22373
22374
22375 <p><b>Proposed resolution:</b></p>
22376 <p>
22377 Add two new rows, and change rehash's postcondition in the unordered
22378 associative container requirements table in 23.2.5 [unord.req]:
22379 </p>
22380
22381 <blockquote>
22382 <table border="1">
22383 <caption>Table 87 &#8212; Unordered associative container requirements
22384 (in addition to container)</caption>
22385
22386 <tbody><tr>
22387 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22388 <th>Complexity</th>
22389 </tr>
22390 <tr>
22391 <td><ins>
22392 <tt>a.min_load_factor()</tt>
22393 </ins></td>
22394 <td><ins>
22395 <tt>float</tt>
22396 </ins></td>
22397 <td><ins>
22398 Returns a non-negative number that the container attempts to keep the
22399 load factor greater than or equal to. The container automatically
22400 decreases the number of buckets as necessary to keep the load factor
22401 above this number.
22402 </ins></td>
22403 <td><ins>
22404 constant
22405 </ins></td>
22406 </tr>
22407
22408 <tr>
22409 <td><ins><tt>a.min_load_factor(z)</tt></ins></td>
22410 <td><ins><tt>void</tt></ins></td>
22411 <td><ins>Pre: <tt>z</tt> shall be non-negative. Changes the container's minimum
22412 load factor, using <tt>z</tt> as a hint. [<i>Footnote:</i> the minimum
22413 load factor should be significantly smaller than the maximum.
22414 If <tt>z</tt> is too large, the implementation may reduce it to a more sensible value.]
22415 </ins></td>
22416 <td><ins>
22417 constant
22418 </ins></td>
22419 </tr>
22420 <tr>
22421 <td><tt>a.rehash(n)</tt></td>
22422 <td><tt>void</tt></td>
22423 <td>
22424 Post: <ins><tt>a.bucket_count() &gt;= n</tt>, and <tt>a.size() &lt;= a.bucket_count()
22425 * a.max_load_factor()</tt>. [<i>Footnote:</i> It is intentional that the
22426 postcondition does not mention the minimum load factor.
22427 This member function is primarily intended for cases where the user knows
22428 that the container's size will increase soon, in which case the container's
22429 load factor will temporarily fall below <tt>a.min_load_factor()</tt>.]</ins>
22430 <del>
22431 <tt>a.bucket_cout &gt; a.size() / a.max_load_factor()</tt> and <tt>a.bucket_count()
22432 &gt;= n</tt>.
22433 </del>
22434 </td>
22435 <td>
22436 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22437 </td>
22438 </tr>
22439 </tbody></table>
22440 </blockquote>
22441
22442 <p>
22443 Add a footnote to 23.2.5 [unord.req] p12:
22444 </p>
22445
22446 <blockquote>
22447 <p>
22448 The insert members shall not affect the validity of references to
22449 container elements, but may invalidate all iterators to the container.
22450 The erase members shall invalidate only iterators and references to the
22451 erased elements.
22452 </p>
22453
22454 <blockquote><ins>
22455 [A consequence of these requirements is that while insert may change the
22456 number of buckets, erase may not. The number of buckets may be reduced
22457 on calls to insert or rehash.]
22458 </ins></blockquote>
22459 </blockquote>
22460
22461 <p>
22462 Change paragraph 13:
22463 </p>
22464
22465 <blockquote>
22466 The insert members shall not affect the validity of iterators if
22467 <del><tt>(N+n) &lt; z * B</tt></del> <ins><tt>zmin * B &lt;= (N+n) &lt;= zmax * B</tt></ins>,
22468 where <tt>N</tt> is the number of elements in
22469 the container prior to the insert operation, <tt>n</tt> is the number of
22470 elements inserted, <tt>B</tt> is the container's bucket count,
22471 <ins><tt>zmin</tt> is the container's minimum load factor,</ins>
22472 and <tt>z<ins>max</ins></tt> is the container's maximum load factor.
22473 </blockquote>
22474
22475 <p>
22476 Add to the <tt>unordered_map</tt> class synopsis in section 23.5.1 [unord.map],
22477 the <tt>unordered_multimap</tt> class synopsis
22478 in 23.5.2 [unord.multimap], the <tt>unordered_set</tt> class synopsis in
22479 23.5.3 [unord.set], and the <tt>unordered_multiset</tt> class synopsis
22480 in 23.5.4 [unord.multiset]:
22481 </p>
22482
22483 <blockquote><pre><ins>
22484 float min_load_factor() const;
22485 void min_load_factor(float z);
22486 </ins></pre></blockquote>
22487
22488 <p>
22489 In 23.5.1.1 [unord.map.cnstr], 23.5.2.1 [unord.multimap.cnstr], 23.5.3.1 [unord.set.cnstr], and
22490 23.5.4.1 [unord.multiset.cnstr], change:
22491 </p>
22492
22493 <blockquote>
22494 ... <tt>max_load_factor()</tt> returns 1.0 <ins>and
22495 <tt>min_load_factor()</tt> returns 0</ins>.
22496 </blockquote>
22497
22498
22499
22500
22501
22502 <hr>
22503 <h3><a name="1189"></a>1189. Awkward interface for changing the number of buckets in an unordered associative container</h3>
22504 <p><b>Section:</b> 23.2.5 [unord.req], 23.5 [unord] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
22505  <b>Submitter:</b> Matt Austern <b>Opened:</b> 2009-08-10  <b>Last modified:</b> 2009-10-28</p>
22506 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
22507 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
22508 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
22509 <p><b>Discussion:</b></p>
22510 <p>
22511 Consider a typical use case: I create an <tt>unordered_map</tt> and then start
22512 adding elements to it one at a time. I know that it will eventually need
22513 to store a few million elements, so, for performance reasons, I would
22514 like to reserve enough capacity that none of the calls to <tt>insert</tt> will
22515 trigger a rehash.
22516 </p>
22517
22518 <p>
22519 Unfortunately, the existing interface makes this awkward. The user
22520 naturally sees the problem in terms of the number of elements, but the
22521 interface presents it as buckets. If <tt>m</tt> is the map and <tt>n</tt> is the expected
22522 number of elements, this operation is written <tt>m.rehash(n /
22523 m.max_load_factor())</tt> &#8212; not very novice friendly.
22524 </p>
22525
22526 <p><i>[
22527 2009-09-30 Daniel adds:
22528 ]</i></p>
22529
22530
22531 <blockquote>
22532 I recommend to replace "<tt>resize</tt>" by a different name like
22533 "<tt>reserve</tt>", because that would better match the intended
22534 use-case. Rational: Any existing resize function has the on-success
22535 post-condition that the provided size is equal to <tt>size()</tt>, which
22536 is not satisfied for the proposal. Reserve seems to fit the purpose of
22537 the actual renaming suggestion.
22538 </blockquote>
22539
22540 <p><i>[
22541 2009-10-28 Ganesh summarizes alternative resolutions and expresses a
22542 strong preference for the second (and opposition to the first):
22543 ]</i></p>
22544
22545
22546 <blockquote>
22547 <ol>
22548 <li>
22549 <p>
22550 In the unordered associative container requirements (23.2.5 [unord.req]),
22551 remove the row for
22552 rehash and replace it with:
22553 </p>
22554
22555 <blockquote>
22556 <table border="1">
22557 <caption>Table 87 &#8212; Unordered associative container requirements
22558 (in addition to container)</caption>
22559
22560 <tbody><tr>
22561 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22562 <th>Complexity</th>
22563 </tr>
22564 <tr>
22565 <td><tt>a.<del>rehash</del><ins>reserve</ins>(n)</tt></td>
22566 <td><tt>void</tt></td>
22567 <td>
22568 Post: <tt>a.bucket_count &gt; <ins>max(</ins>a.size()<ins>, n)</ins>
22569 / a.max_load_factor()</tt><del> and <tt>a.bucket_count()
22570 &gt;= n</tt></del>.
22571 </td>
22572 <td>
22573 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22574 </td>
22575 </tr>
22576 </tbody></table>
22577 </blockquote>
22578
22579 <p>
22580 Make the corresponding change in the class synopses in 23.5.1
22581 [unord.map], 23.5.2 [unord.multimap], 23.5.3 [unord.set], and 23.5.4
22582 [unord.multiset].
22583 </p>
22584 </li>
22585 <li>
22586
22587 <p>
22588 In 23.2.5 [unord.req]/9, table 98, append a new row after the last one:
22589 </p>
22590
22591 <blockquote>
22592 <table border="1">
22593 <caption>Table 87 &#8212; Unordered associative container requirements
22594 (in addition to container)</caption>
22595
22596 <tbody><tr>
22597 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22598 <th>Complexity</th>
22599 </tr>
22600 <tr>
22601 <td><tt>a.rehash(n)</tt></td>
22602 <td><tt>void</tt></td>
22603 <td>
22604 Post: <tt>a.bucket_count &gt; a.size()
22605 / a.max_load_factor()</tt> and <tt>a.bucket_count()
22606 &gt;= n</tt>.
22607 </td>
22608 <td>
22609 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22610 </td>
22611 </tr>
22612 <tr>
22613 <td><ins>
22614 <tt>a.reserve(n)</tt>
22615 </ins></td>
22616 <td><ins>
22617 <tt>void</tt>
22618 </ins></td>
22619 <td><ins>
22620 Same as <tt>a.rehash(ceil(n / a.max_load_factor()))</tt>
22621 </ins></td>
22622 <td><ins>
22623 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22624 </ins></td>
22625 </tr>
22626 </tbody></table>
22627 </blockquote>
22628
22629 <p>
22630 In 23.5.1 [unord.map]/3 in the definition of class template <tt>unordered_map</tt>, in
22631 23.5.2 [unord.multimap]/3 in the definition of class template <tt>unordered_multimap</tt>, in
22632 23.5.3 [unord.set]/3 in the definition of class template <tt>unordered_set</tt> and in
22633 23.5.4 [unord.multiset]/3 in the definition of class template <tt>unordered_multiset</tt>, add the
22634 following line after member function <tt>rehash()</tt>:
22635 </p>
22636
22637 <blockquote><pre>void reserve(size_type n);
22638 </pre></blockquote>
22639
22640 </li>
22641 </ol>
22642 </blockquote>
22643
22644 <p><i>[
22645 2009-10-28 Howard:
22646 ]</i></p>
22647
22648
22649 <blockquote>
22650 <p>
22651 Moved to Tentatively Ready after 5 votes in favor of Ganesh's option 2 above.
22652 The original proposed wording now appears here:
22653 </p>
22654
22655 <blockquote>
22656 <p>
22657 Informally: instead of providing <tt>rehash(n)</tt> provide <tt>resize(n)</tt>, with the
22658 semantics "make the container a good size for <tt>n</tt> elements".
22659 </p>
22660
22661 <p>
22662 In the unordered associative container requirements (23.2.5 [unord.req]),
22663 remove the row for
22664 rehash and replace it with:
22665 </p>
22666
22667 <blockquote>
22668 <table border="1">
22669 <caption>Table 87 &#8212; Unordered associative container requirements
22670 (in addition to container)</caption>
22671
22672 <tbody><tr>
22673 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22674 <th>Complexity</th>
22675 </tr>
22676 <tr>
22677 <td><tt>a.<del>rehash</del><ins>resize</ins>(n)</tt></td>
22678 <td><tt>void</tt></td>
22679 <td>
22680 Post: <tt>a.bucket_count &gt; <ins>max(</ins>a.size()<ins>, n)</ins>
22681 / a.max_load_factor()</tt><del> and <tt>a.bucket_count()
22682 &gt;= n</tt></del>.
22683 </td>
22684 <td>
22685 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22686 </td>
22687 </tr>
22688 </tbody></table>
22689 </blockquote>
22690
22691 <p>Make the corresponding change in the class synopses in 23.5.1
22692 [unord.map], 23.5.2 [unord.multimap], 23.5.3 [unord.set], and 23.5.4
22693 [unord.multiset].
22694 </p>
22695
22696 </blockquote>
22697 </blockquote>
22698
22699
22700 <p><b>Proposed resolution:</b></p>
22701 <p>
22702 In 23.2.5 [unord.req]/9, table 98, append a new row after the last one:
22703 </p>
22704
22705 <blockquote>
22706 <table border="1">
22707 <caption>Table 87 &#8212; Unordered associative container requirements
22708 (in addition to container)</caption>
22709
22710 <tbody><tr>
22711 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22712 <th>Complexity</th>
22713 </tr>
22714 <tr>
22715 <td><tt>a.rehash(n)</tt></td>
22716 <td><tt>void</tt></td>
22717 <td>
22718 Post: <tt>a.bucket_count &gt; a.size()
22719 / a.max_load_factor()</tt> and <tt>a.bucket_count()
22720 &gt;= n</tt>.
22721 </td>
22722 <td>
22723 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22724 </td>
22725 </tr>
22726 <tr>
22727 <td><ins>
22728 <tt>a.reserve(n)</tt>
22729 </ins></td>
22730 <td><ins>
22731 <tt>void</tt>
22732 </ins></td>
22733 <td><ins>
22734 Same as <tt>a.rehash(ceil(n / a.max_load_factor()))</tt>
22735 </ins></td>
22736 <td><ins>
22737 Average case linear in <tt>a.size()</tt>, worst case quadratic.
22738 </ins></td>
22739 </tr>
22740 </tbody></table>
22741 </blockquote>
22742
22743 <p>
22744 In 23.5.1 [unord.map]/3 in the definition of class template <tt>unordered_map</tt>, in
22745 23.5.2 [unord.multimap]/3 in the definition of class template <tt>unordered_multimap</tt>, in
22746 23.5.3 [unord.set]/3 in the definition of class template <tt>unordered_set</tt> and in
22747 23.5.4 [unord.multiset]/3 in the definition of class template <tt>unordered_multiset</tt>, add the
22748 following line after member function <tt>rehash()</tt>:
22749 </p>
22750
22751 <blockquote><pre>void reserve(size_type n);
22752 </pre></blockquote>
22753
22754
22755
22756
22757
22758 <hr>
22759 <h3><a name="1190"></a>1190. Setting the maximum load factor should return the previous value</h3>
22760 <p><b>Section:</b> 23.2.5 [unord.req], 23.5 [unord] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
22761  <b>Submitter:</b> Matt Austern <b>Opened:</b> 2009-08-10  <b>Last modified:</b> 2009-08-11</p>
22762 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
22763 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
22764 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
22765 <p><b>Discussion:</b></p>
22766 <p>
22767 The unordered associative container requirements table specifies that
22768 <tt>a.set_max_load_factor(z)</tt> has return type <tt>void</tt>. However, there is a
22769 useful piece of information to return: the previous value. Users who
22770 don't need it can always ignore it.
22771 </p>
22772
22773
22774 <p><b>Proposed resolution:</b></p>
22775 <p>
22776 In the unordered associative container requirements table, change:
22777 </p>
22778
22779 <blockquote>
22780 <table border="1">
22781 <caption>Table 87 &#8212; Unordered associative container requirements
22782 (in addition to container)</caption>
22783
22784 <tbody><tr>
22785 <th>Expression</th><th>Return type</th><th>Assertion/note pre-/post-condition</th>
22786 <th>Complexity</th>
22787 </tr>
22788
22789 <tr>
22790 <td><tt>a.max_load_factor(z)</tt></td>
22791 <td><tt><del>void</del> <ins>float</ins></tt></td>
22792 <td>Pre: <tt>z</tt> shall be positive. Changes the container's maximum
22793 <del>load</del> load factor, using <tt>z</tt> as a hint.
22794 <ins>Returns: the previous value of
22795 <tt>a.max_load_factor()</tt>.</ins>
22796 </td>
22797 <td>
22798 constant
22799 </td>
22800 </tr>
22801 <tr></tr>
22802 </tbody></table>
22803 </blockquote>
22804
22805 <p>
22806 Change the return type of <tt>set_max_load_factor</tt>
22807 in the class synopses in 23.5.1 [unord.map], 23.5.2 [unord.multimap],  23.5.3 [unord.set],
22808 and 23.5.4 [unord.multiset].
22809 </p>
22810
22811 <p>
22812 If issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1188">1188</a> is also accepted, make the same changes for
22813 <tt>min_load_factor</tt>.
22814 </p>
22815
22816
22817
22818
22819
22820 <hr>
22821 <h3><a name="1191"></a>1191. <tt>tuple get</tt> API should respect rvalues</h3>
22822 <p><b>Section:</b> 20.5.2.6 [tuple.elem] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
22823  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-08-18  <b>Last modified:</b> 2009-10-31</p>
22824 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
22825 <p><b>Discussion:</b></p>
22826 <p>
22827 The <tt>tuple get</tt> API should respect rvalues.  This would allow for moving a
22828 single element out of a <tt>tuple</tt>-like type.
22829 </p>
22830
22831 <p><i>[
22832 2009-10-30 Alisdair adds:
22833 ]</i></p>
22834
22835
22836 <blockquote>
22837 <p>
22838 The issue of rvalue overloads of get for tuple-like types was briefly
22839 discussed in Santa Cruz.
22840 </p>
22841
22842 <p>
22843 The feedback was this would be welcome, but we need full wording for the
22844 other types (<tt>pair</tt> and <tt>array</tt>) before advancing.
22845 </p>
22846
22847 <p>
22848 I suggest the issue moves to Open from New as it has been considered,
22849 feedback given, and it has not (yet) been rejected as NAD.
22850 </p>
22851 </blockquote>
22852
22853
22854
22855 <p><b>Proposed resolution:</b></p>
22856 <p>
22857 Add the following signature to p2 20.5.1 [tuple.general]
22858 </p>
22859
22860 <blockquote><pre><ins>
22861 template &lt;size_t I, class ... Types&gt;
22862 typename tuple_element&lt;I, tuple&lt;Types...&gt; &gt;::type&amp;&amp; get(tuple&lt;Types...&gt; &amp;&amp;);
22863 </ins></pre></blockquote>
22864
22865 <p>
22866 And again to 20.5.2.6 [tuple.elem].
22867 </p>
22868
22869 <blockquote><pre><ins>
22870 template &lt;size_t I, class ... Types&gt;
22871 typename tuple_element&lt;I, tuple&lt;Types...&gt; &gt;::type&amp;&amp; get(tuple&lt;Types...&gt;&amp;&amp; t);
22872 </ins></pre>
22873
22874 <blockquote>
22875 <p><ins>
22876 <i>Effects:</i> Equivalent to <tt>return std::forward&lt;typename tuple_element&lt;I, tuple&lt;Types...&gt; &gt;::type&amp;&amp;&gt;(get&lt;I&gt;(t));</tt>
22877 </ins></p>
22878
22879
22880 <p><ins>
22881 [<i>Note:</i> If a <tt>T</tt> in <tt>Types</tt> is some reference type <tt>X&amp;</tt>,
22882 the return type is <tt>X&amp;</tt>, not <tt>X&amp;&amp;</tt>.
22883 However, if the element type is non-reference type <tt>T</tt>,
22884 the return type is <tt>T&amp;&amp;</tt>. &#8212; <i>end note</i>]
22885 </ins></p>
22886
22887 </blockquote>
22888 </blockquote>
22889
22890 <p>
22891 Add the following signature to p1 20.3 [utility]
22892 </p>
22893
22894 <blockquote><pre><ins>
22895 template &lt;size_t I, class T1, class T2&gt;
22896 typename tuple_element&lt;I, pair&lt;T1,T2&gt; &gt;::type&amp;&amp; get(pair&lt;T1, T2&gt;&amp;&amp;);
22897 </ins></pre></blockquote>
22898
22899 <p>
22900 And to p5 20.3.5 [pair.astuple]
22901 </p>
22902
22903 <blockquote><pre><ins>
22904 template &lt;size_t I, class T1, class T2&gt;
22905 typename tuple_element&lt;I, pair&lt;T1,T2&gt; &gt;::type&amp;&amp; get(pair&lt;T1, T2&gt;&amp;&amp; p);
22906 </ins></pre>
22907
22908 <blockquote>
22909 <p><ins>
22910 <i>Returns:</i> If <tt>I == 0</tt> returns <tt>std::forward&lt;T1&amp;&amp;&gt;(p.first)</tt>;
22911 if <tt>I == 1</tt>
22912 returns <tt>std::forward&lt;T2&amp;&amp;&gt;(p.second)</tt>; otherwise the program is ill-formed.
22913 </ins></p>
22914
22915 <p><ins>
22916 <i>Throws:</i> Nothing.
22917 </ins></p>
22918
22919 </blockquote>
22920
22921 </blockquote>
22922
22923 <p>
22924 Add the following signature to 23.3 [sequences] <tt>&lt;array&gt;</tt> synopsis
22925 </p>
22926
22927 <blockquote><pre><ins>template &lt;size_t I, class T, size_t N&gt;
22928 T&amp;&amp; get(array&lt;T,N&gt; &amp;&amp;);
22929 </ins></pre></blockquote>
22930
22931 <p>
22932 And after p8 23.3.1.7 [array.tuple]
22933 </p>
22934
22935 <blockquote><pre><ins>template &lt;size_t I, class T, size_t N&gt;
22936 T&amp;&amp; get(array&lt;T,N&gt; &amp;&amp; a);
22937 </ins></pre>
22938
22939 <blockquote><ins>
22940 <i>Effects:</i> Equivalent to <tt>return std::move(get&lt;I&gt;(a));</tt>
22941 </ins></blockquote>
22942 </blockquote>
22943
22944
22945
22946
22947
22948
22949 <hr>
22950 <h3><a name="1192"></a>1192. <tt>basic_string</tt> missing definitions for <tt>cbegin</tt> / <tt>cend</tt> / <tt>crbegin</tt> / <tt>crend</tt></h3>
22951 <p><b>Section:</b> 21.4.3 [string.iterators] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
22952  <b>Submitter:</b> Jonathan Wakely <b>Opened:</b> 2009-08-14  <b>Last modified:</b> 2009-10-29</p>
22953 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
22954 <p><b>Discussion:</b></p>
22955 <p>
22956 Unlike the containers in clause 23, <tt>basic_string</tt> has definitions for
22957 <tt>begin()</tt> and <tt>end()</tt>, but these have not been updated to include <tt>cbegin</tt>,
22958 <tt>cend</tt>, <tt>crbegin</tt> and <tt>crend</tt>.
22959 </p>
22960
22961 <p><i>[
22962 2009-10-28 Howard:
22963 ]</i></p>
22964
22965
22966 <blockquote>
22967 Moved to Tentatively NAD after 5 positive votes on c++std-lib.  Added
22968 rationale.
22969 </blockquote>
22970
22971 <p><i>[
22972 2009-10-28 Alisdair disagrees:
22973 ]</i></p>
22974
22975
22976 <blockquote>
22977 <p>
22978 I'm going to have to speak up as the dissenting voice.
22979 </p>
22980
22981 <p>
22982 I agree the issue could be handled editorially, and that would be my
22983 preference if Pete feels this is appropriate. Failing that, I really
22984 think this issue should be accepted and moved to ready.  The other
22985 begin/end functions all have a semantic definition for this template,
22986 and it is confusing if a small few are missing.
22987 </p>
22988
22989 <p>
22990 I agree that an alternative would be to strike <em>all</em> the definitions for
22991 <tt>begin/end/rbegin/rend</tt> and defer completely to the requirements tables in
22992 clause 23.  I think that might be confusing without a forward reference
22993 though, as those tables are defined in a *later* clause than the
22994 basic_string template itself.  If someone wants to pursue this I would
22995 support it, but recommend it as a separate issue.
22996 </p>
22997
22998 <p>
22999 So my preference is strongly to move Ready over NAD, and a stronger
23000 preference for NAD Editorial if Pete is happy to make these changes.
23001 </p>
23002
23003 </blockquote>
23004
23005 <p><i>[
23006 2009-10-29 Howard:
23007 ]</i></p>
23008
23009
23010 <blockquote>
23011 Moved to Tentatively Ready after 5 positive votes on c++std-lib.  Removed
23012 rationale to mark it NAD.  :-)
23013 </blockquote>
23014
23015
23016
23017 <p><b>Proposed resolution:</b></p>
23018 <p>
23019 Add to 21.4.3 [string.iterators]
23020 </p>
23021
23022 <blockquote><pre>iterator       begin();
23023 const_iterator begin() const;
23024 <ins>const_iterator cbegin() const;</ins>
23025 </pre>
23026
23027 <p>...</p>
23028
23029 <pre>iterator       end();
23030 const_iterator end() const;
23031 <ins>const_iterator cend() const;</ins>
23032 </pre>
23033
23034 <p>...</p>
23035
23036 <pre>reverse_iterator       rbegin();
23037 const_reverse_iterator rbegin() const;
23038 <ins>const_reverse_iterator crbegin() const;</ins>
23039 </pre>
23040
23041 <p>...</p>
23042
23043 <pre>reverse_iterator       rend();
23044 const_reverse_iterator rend() const;
23045 <ins>const_reverse_iterator crend() const;</ins>
23046 </pre>
23047
23048 </blockquote>
23049
23050
23051
23052
23053
23054 <hr>
23055 <h3><a name="1193"></a>1193. <tt>default_delete</tt> cannot be instantiated with  incomplete types</h3>
23056 <p><b>Section:</b> 20.8.14.1 [unique.ptr.dltr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
23057  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-08-18  <b>Last modified:</b> 2009-08-22</p>
23058 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
23059 <p><b>Discussion:</b></p>
23060 <p>
23061 According to the general rules of 17.6.3.8 [res.on.functions]/2 b 5 the effects
23062 are undefined, if an incomplete type is used to instantiate a library template. But neither in
23063 20.8.14.1 [unique.ptr.dltr] nor
23064 in any other place of the standard such explicit allowance is given.
23065 Since this template is intended to be instantiated with incomplete
23066 types, this must
23067 be fixed.
23068 </p>
23069
23070
23071 <p><b>Proposed resolution:</b></p>
23072 <p>
23073 Add two new paragraphs directly to 20.8.14.1 [unique.ptr.dltr] (before
23074 20.8.14.1.1 [unique.ptr.dltr.dflt]) with the following
23075 content:
23076 </p>
23077
23078 <blockquote>
23079 <p><ins>
23080 The class template <tt>default_delete</tt> serves as the default deleter (destruction policy) for
23081 the class template <tt>unique_ptr</tt>.
23082 </ins></p>
23083
23084 <p><ins>
23085 The template parameter <tt>T</tt> of <tt>default_delete</tt> may be an incomplete type.
23086 </ins></p>
23087 </blockquote>
23088
23089
23090
23091
23092
23093 <hr>
23094 <h3><a name="1194"></a>1194. Unintended <tt>queue</tt> constructor</h3>
23095 <p><b>Section:</b> 23.3.5 [container.adaptors] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
23096  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-08-20  <b>Last modified:</b> 2009-10-20</p>
23097 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.adaptors">active issues</a> in [container.adaptors].</p>
23098 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.adaptors">issues</a> in [container.adaptors].</p>
23099 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
23100 <p><b>Discussion:</b></p>
23101 <p>
23102 23.3.5.1.1 [queue.defn] has the following <tt>queue</tt> constructor:
23103 </p>
23104
23105 <blockquote><pre>template &lt;class Alloc&gt; explicit queue(const Alloc&amp;);
23106 </pre></blockquote>
23107
23108 <p>
23109 This will be implemented like so:
23110 </p>
23111
23112 <blockquote><pre>template &lt;class Alloc&gt; explicit queue(const Alloc&amp; a) : c(a) {}
23113 </pre></blockquote>
23114
23115 <p>
23116 The issue is that <tt>Alloc</tt> can be anything that a container will construct
23117 from, for example an <tt>int</tt>.  Is this intended to compile?
23118 </p>
23119
23120 <blockquote><pre>queue&lt;int&gt; q(5);
23121 </pre></blockquote>
23122
23123 <p>
23124 Before the addition of this constructor, <tt>queue&lt;int&gt;(5)</tt> would not compile.
23125 I ask, not because this crashes, but because it is new and appears to be
23126 unintended.  We do not want to be in a position of accidently introducing this
23127 "feature" in C++0X and later attempting to remove it.
23128 </p>
23129
23130 <p>
23131 I've picked on <tt>queue</tt>.  <tt>priority_queue</tt> and <tt>stack</tt> have
23132 the same issue.  Is it useful to create a <tt>priority_queue</tt> of 5
23133 identical elements?
23134 </p>
23135
23136 <p><i>[
23137 Daniel, Howard and Pablo collaborated on the proposed wording.
23138 ]</i></p>
23139
23140
23141 <p><i>[
23142 2009-10 Santa Cruz:
23143 ]</i></p>
23144
23145
23146 <blockquote>
23147 Move to Ready.
23148 </blockquote>
23149
23150
23151
23152 <p><b>Proposed resolution:</b></p>
23153
23154 <p><i>[
23155 This resolution includes a semi-editorial clean up, giving definitions to members
23156 which in some cases weren't defined since C++98.
23157 This resolution also offers editorially different wording for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#976">976</a>,
23158 and it also provides wording for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1196">1196</a>.
23159 ]</i></p>
23160
23161
23162 <p>
23163 Change  container.adaptors, p1:
23164 </p>
23165
23166 <blockquote>
23167 The container adaptors each take a <tt>Container</tt> template parameter, and
23168 each constructor takes a <tt>Container</tt> reference argument. This container is
23169 copied into the <tt>Container</tt> member of each adaptor. If the container takes
23170 an allocator, then a compatible allocator may be passed in to the
23171 adaptor's constructor. Otherwise, normal copy or move construction is
23172 used for the container argument. <del>[<i>Note:</i> it is not necessary for an
23173 implementation to distinguish between the one-argument constructor that
23174 takes a <tt>Container</tt> and the one- argument constructor that takes an
23175 allocator_type. Both forms use their argument to construct an instance
23176 of the container. &#8212; <i>end note</i>]</del>
23177 </blockquote>
23178
23179 <p>
23180 Change  queue.defn, p1:
23181 </p>
23182
23183 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23184 class queue {
23185 public:
23186   typedef typename Container::value_type      value_type;
23187   typedef typename Container::reference       reference;
23188   typedef typename Container::const_reference const_reference;
23189   typedef typename Container::size_type       size_type;
23190   typedef Container                           container_type;
23191 protected:
23192   Container c;
23193
23194 public:
23195   explicit queue(const Container&amp;);
23196   explicit queue(Container&amp;&amp; = Container());
23197   queue(queue&amp;&amp; q)<ins>;</ins><del> : c(std::move(q.c)) {}</del>
23198   template &lt;class Alloc&gt; explicit queue(const Alloc&amp;);
23199   template &lt;class Alloc&gt; queue(const Container&amp;, const Alloc&amp;);
23200   template &lt;class Alloc&gt; queue(Container&amp;&amp;, const Alloc&amp;);
23201   template &lt;class Alloc&gt; queue(queue&amp;&amp;, const Alloc&amp;);
23202   queue&amp; operator=(queue&amp;&amp; q)<ins>;</ins><del> { c = std::move(q.c); return *this; }</del>
23203
23204   bool empty() const          { return c.empty(); }
23205   ...
23206 };
23207 </pre></blockquote>
23208
23209 <p>
23210 Add a new section after 23.3.5.1.1 [queue.defn], [queue.cons]:
23211 </p>
23212
23213 <blockquote>
23214 <p><b><tt>queue</tt> constructors [queue.cons]</b></p>
23215
23216 <pre>explicit queue(const Container&amp; cont);
23217 </pre>
23218
23219 <blockquote>
23220
23221 <p>
23222 <i>Effects:</i> Initializes <tt>c</tt> with <tt>cont</tt>.
23223 </p>
23224
23225 </blockquote>
23226
23227 <pre>explicit queue(Container&amp;&amp; cont = Container());
23228 </pre>
23229
23230 <blockquote>
23231
23232 <p>
23233 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(cont)</tt>.
23234 </p>
23235
23236 </blockquote>
23237
23238 <pre>queue(queue&amp;&amp; q)
23239 </pre>
23240
23241 <blockquote>
23242
23243 <p>
23244 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(q.c)</tt>.
23245 </p>
23246
23247 </blockquote>
23248
23249 <p>
23250 For each of the following constructors,
23251 if <tt>uses_allocator&lt;container_type, Alloc&gt;::value</tt> is <tt>false</tt>,
23252 then the constructor shall not participate in overload resolution.
23253 </p>
23254
23255 <pre>template &lt;class Alloc&gt;
23256   explicit queue(const Alloc&amp; a);
23257 </pre>
23258
23259 <blockquote>
23260
23261 <p>
23262 <i>Effects:</i> Initializes <tt>c</tt> with <tt>a</tt>.
23263 </p>
23264
23265 </blockquote>
23266
23267 <pre>template &lt;class Alloc&gt;
23268   queue(const container_type&amp; cont, const Alloc&amp; a);
23269 </pre>
23270
23271 <blockquote>
23272
23273 <p>
23274 <i>Effects:</i> Initializes <tt>c</tt> with <tt>cont</tt> as the first
23275 argument and <tt>a</tt> as the second argument.
23276 </p>
23277
23278 </blockquote>
23279
23280 <pre>template &lt;class Alloc&gt;
23281   queue(container_type&amp;&amp; cont, const Alloc&amp; a);
23282 </pre>
23283
23284 <blockquote>
23285
23286 <p>
23287 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(cont)</tt> as the
23288 first argument and <tt>a</tt> as the second argument.
23289 </p>
23290
23291 </blockquote>
23292
23293 <pre>template &lt;class Alloc&gt;
23294   queue(queue&amp;&amp; q, const Alloc&amp; a);
23295 </pre>
23296
23297 <blockquote>
23298
23299 <p>
23300 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(q.c)</tt> as the
23301 first argument and <tt>a</tt> as the second argument.
23302 </p>
23303
23304 </blockquote>
23305
23306 <pre>queue&amp; operator=(queue&amp;&amp; q);
23307 </pre>
23308
23309 <blockquote>
23310
23311 <p>
23312 <i>Effects:</i> Assigns <tt>c</tt> with <tt>std::move(q.c)</tt>.
23313 </p>
23314
23315 <p>
23316 <i>Returns:</i> <tt>*this</tt>.
23317 </p>
23318
23319 </blockquote>
23320
23321
23322
23323 </blockquote>
23324
23325 <p>
23326 Add to 23.3.5.2.1 [priqueue.cons]:
23327 </p>
23328
23329 <blockquote>
23330
23331 <pre>priority_queue(priority_queue&amp;&amp; q);
23332 </pre>
23333
23334 <blockquote>
23335
23336 <p>
23337 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(q.c)</tt> and
23338 initializes <tt>comp</tt> with <tt>std::move(q.comp)</tt>.
23339 </p>
23340
23341 </blockquote>
23342
23343 <p>
23344 For each of the following constructors,
23345 if <tt>uses_allocator&lt;container_type, Alloc&gt;::value</tt> is <tt>false</tt>,
23346 then the constructor shall not participate in overload resolution.
23347 </p>
23348
23349 <pre>template &lt;class Alloc&gt;
23350   explicit priority_queue(const Alloc&amp; a);
23351 </pre>
23352
23353 <blockquote>
23354
23355 <p>
23356 <i>Effects:</i> Initializes <tt>c</tt> with <tt>a</tt> and value-initializes <tt>comp</tt>.
23357 </p>
23358
23359 </blockquote>
23360
23361 <pre>template &lt;class Alloc&gt;
23362   priority_queue(const Compare&amp; compare, const Alloc&amp; a);
23363 </pre>
23364
23365 <blockquote>
23366
23367 <p>
23368 <i>Effects:</i> Initializes <tt>c</tt> with <tt>a</tt> and initializes <tt>comp</tt>
23369 with <tt>compare</tt>.
23370 </p>
23371
23372 </blockquote>
23373
23374 <pre>template &lt;class Alloc&gt;
23375   priority_queue(const Compare&amp; compare, const Container&amp; cont, const Alloc&amp; a);
23376 </pre>
23377
23378 <blockquote>
23379
23380 <p>
23381 <i>Effects:</i> Initializes <tt>c</tt> with <tt>cont</tt> as the first argument
23382 and <tt>a</tt> as the second argument,
23383 and initializes <tt>comp</tt> with <tt>compare</tt>.
23384 </p>
23385
23386 </blockquote>
23387
23388 <pre>template &lt;class Alloc&gt;
23389   priority_queue(const Compare&amp; compare, Container&amp;&amp; cont, const Alloc&amp; a);
23390 </pre>
23391
23392 <blockquote>
23393
23394 <p>
23395 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(cont)</tt> as
23396 the first argument and <tt>a</tt> as the second argument,
23397 and initializes <tt>comp</tt> with <tt>compare</tt>.
23398 </p>
23399
23400 </blockquote>
23401
23402 <pre>template &lt;class Alloc&gt;
23403   priority_queue(priority_queue&amp;&amp; q, const Alloc&amp; a);
23404 </pre>
23405
23406 <blockquote>
23407
23408 <p>
23409 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(q.c)</tt> as the
23410 first argument and <tt>a</tt> as the second argument,
23411 and initializes <tt>comp</tt> with <tt>std::move(q.comp)</tt>.
23412 </p>
23413
23414 </blockquote>
23415
23416 <pre>priority_queue&amp; operator=(priority_queue&amp;&amp; q);
23417 </pre>
23418
23419 <blockquote>
23420
23421 <p>
23422 <i>Effects:</i> Assigns <tt>c</tt> with <tt>std::move(q.c)</tt> and
23423 assigns <tt>comp</tt> with <tt>std::move(q.comp)</tt>.
23424 </p>
23425
23426 <p>
23427 <i>Returns:</i> <tt>*this</tt>.
23428 </p>
23429
23430 </blockquote>
23431
23432 </blockquote>
23433
23434
23435
23436
23437 <p>
23438 Change 23.3.5.3.1 [stack.defn]:
23439 </p>
23440
23441 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23442 class stack {
23443 public:
23444   typedef typename Container::value_type      value_type;
23445   typedef typename Container::reference       reference;
23446   typedef typename Container::const_reference const_reference;
23447   typedef typename Container::size_type       size_type;
23448   typedef Container                           container_type;
23449 protected:
23450   Container c;
23451
23452 public:
23453   explicit stack(const Container&amp;);
23454   explicit stack(Container&amp;&amp; = Container());
23455   <ins>stack(stack&amp;&amp; s);</ins>
23456   template &lt;class Alloc&gt; explicit stack(const Alloc&amp;);
23457   template &lt;class Alloc&gt; stack(const Container&amp;, const Alloc&amp;);
23458   template &lt;class Alloc&gt; stack(Container&amp;&amp;, const Alloc&amp;);
23459   template &lt;class Alloc&gt; stack(stack&amp;&amp;, const Alloc&amp;);
23460   <ins>stack&amp; operator=(stack&amp;&amp; s);</ins>
23461
23462   bool empty() const          { return c.empty(); }
23463   ...
23464 };
23465 </pre></blockquote>
23466
23467 <p>
23468 Add a new section after 23.3.5.3.1 [stack.defn], [stack.cons]:
23469 </p>
23470
23471 <blockquote>
23472 <p><b><tt>stack</tt> constructors [stack.cons]</b></p>
23473
23474 <pre>stack(stack&amp;&amp; s);
23475 </pre>
23476
23477 <blockquote>
23478
23479 <p>
23480 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(s.c)</tt>.
23481 </p>
23482
23483 </blockquote>
23484
23485 <p>
23486 For each of the following constructors,
23487 if <tt>uses_allocator&lt;container_type, Alloc&gt;::value</tt> is <tt>false</tt>,
23488 then the constructor shall not participate in overload resolution.
23489 </p>
23490
23491 <pre>template &lt;class Alloc&gt;
23492   explicit stack(const Alloc&amp; a);
23493 </pre>
23494
23495 <blockquote>
23496
23497 <p>
23498 <i>Effects:</i> Initializes <tt>c</tt> with <tt>a</tt>.
23499 </p>
23500
23501 </blockquote>
23502
23503 <pre>template &lt;class Alloc&gt;
23504   stack(const container_type&amp; cont, const Alloc&amp; a);
23505 </pre>
23506
23507 <blockquote>
23508
23509 <p>
23510 <i>Effects:</i> Initializes <tt>c</tt> with <tt>cont</tt> as the
23511 first argument and <tt>a</tt> as the second argument.
23512 </p>
23513
23514 </blockquote>
23515
23516 <pre>template &lt;class Alloc&gt;
23517   stack(container_type&amp;&amp; cont, const Alloc&amp; a);
23518 </pre>
23519
23520 <blockquote>
23521
23522 <p>
23523 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(cont)</tt> as the
23524 first argument and <tt>a</tt> as the second argument.
23525 </p>
23526
23527 </blockquote>
23528
23529 <pre>template &lt;class Alloc&gt;
23530   stack(stack&amp;&amp; s, const Alloc&amp; a);
23531 </pre>
23532
23533 <blockquote>
23534
23535 <p>
23536 <i>Effects:</i> Initializes <tt>c</tt> with <tt>std::move(s.c)</tt> as the
23537 first argument and <tt>a</tt> as the second argument.
23538 </p>
23539
23540 </blockquote>
23541
23542 <pre>stack&amp; operator=(stack&amp;&amp; s);
23543 </pre>
23544
23545 <blockquote>
23546
23547 <p>
23548 <i>Effects:</i> Assigns <tt>c</tt> with <tt>std::move(s.c)</tt>.
23549 </p>
23550
23551 <p>
23552 <i>Returns:</i> <tt>*this</tt>.
23553 </p>
23554
23555 </blockquote>
23556
23557 </blockquote>
23558
23559
23560
23561
23562
23563
23564 <hr>
23565 <h3><a name="1197"></a>1197. Can unordered containers have <tt>bucket_count() == 0</tt>?</h3>
23566 <p><b>Section:</b> 23.2.5 [unord.req] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
23567  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-08-24  <b>Last modified:</b> 2009-09-03</p>
23568 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.req">active issues</a> in [unord.req].</p>
23569 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.req">issues</a> in [unord.req].</p>
23570 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
23571 <p><b>Discussion:</b></p>
23572 <p>
23573 Table 97 "Unordered associative container requirements" in
23574 23.2.5 [unord.req] says:
23575 </p>
23576
23577 <blockquote>
23578 <table border="1">
23579 <caption>Table 97 &#8212; Unordered associative container requirements
23580 (in addition to container)</caption>
23581
23582 <tbody><tr>
23583 <th>Expression</th>
23584 <th>Return type</th>
23585 <th>Assertion/note pre-/post-condition</th>
23586 <th>Complexity</th>
23587 </tr>
23588
23589 <tr>
23590 <td><tt>b.bucket(k)</tt></td>
23591 <td><tt>size_type</tt></td>
23592 <td>Returns the index of the bucket in which elements with keys
23593 equivalent to <tt>k</tt> would be found,
23594 if any such element existed.
23595 Post: the return value shall be
23596 in the range <tt>[0,
23597 b.bucket_count())</tt>.</td>
23598 <td>Constant</td>
23599 </tr>
23600
23601 </tbody></table>
23602 </blockquote>
23603
23604 <p>
23605 What should <tt>b.bucket(k)</tt> return if <tt>b.bucket_count() == 0</tt>?
23606 </p>
23607
23608 <p>
23609 I believe allowing <tt>b.bucket_count() == 0</tt> is important.  It is a
23610 very reasonable post-condition of the default constructor, or of a moved-from
23611 container.
23612 </p>
23613
23614 <p>
23615 I can think of several reasonable results from <tt>b.bucket(k)</tt> when
23616 <tt>b.bucket_count() == 0</tt>:
23617 </p>
23618
23619 <ol>
23620 <li>
23621 Return 0.
23622 </li>
23623 <li>
23624 Return <tt>numeric_limits&lt;size_type&gt;::max()</tt>.
23625 </li>
23626 <li>
23627 Throw a <tt>domain_error</tt>.
23628 </li>
23629 <li>
23630 Precondition: <tt>b.bucket_count() != 0</tt>.
23631 </li>
23632 </ol>
23633
23634 <p><i>[
23635 2009-08-26 Daniel adds:
23636 ]</i></p>
23637
23638
23639 <blockquote>
23640 <p>
23641 A forth choice would be to add the pre-condition "<tt>b.bucket_count() != 0</tt>"
23642 and thus imply undefined behavior if this is violated.
23643 </p>
23644
23645 <p><i>[
23646 Howard:  I like this option too, added to the list.
23647 ]</i></p>
23648
23649
23650 <p>
23651 Further on here my own favorite solution (rationale see below):
23652 </p>
23653
23654 <p><b>Suggested resolution:</b></p>
23655
23656 <p>
23657 [Rationale: I suggest to follow choice (1). The main reason is
23658 that all associative container functions which take a key argument,
23659 are basically free of pre-conditions and non-disrupting, therefore
23660 excluding choices (3) and (4). Option (2) seems a bit unexpected
23661 to me. It would be more natural, if several similar functions
23662 would exist which would also justify the existence of a symbolic
23663 constant like npos for this situation. The value 0 is both simple
23664 and consistent, it has exactly the same role as a past-the-end
23665 iterator value. A typical use-case is:
23666 </p>
23667
23668 <blockquote><pre>size_type pos = m.bucket(key);
23669 if (pos != m.bucket_count()) {
23670  ...
23671 } else {
23672  ...
23673 }
23674 </pre></blockquote>
23675
23676 <p>&#8212; end Rationale]</p>
23677
23678 <p>
23679 - Change Table 97 in 23.2.5 [unord.req] as follows (Row b.bucket(k), Column "Assertion/..."):
23680 </p>
23681
23682 <blockquote>
23683 <table border="1">
23684 <caption>Table 97 &#8212; Unordered associative container requirements
23685 (in addition to container)</caption>
23686
23687 <tbody><tr>
23688 <th>Expression</th>
23689 <th>Return type</th>
23690 <th>Assertion/note pre-/post-condition</th>
23691 <th>Complexity</th>
23692 </tr>
23693
23694 <tr>
23695 <td><tt>b.bucket(k)</tt></td>
23696 <td><tt>size_type</tt></td>
23697 <td>Returns the index of the bucket in which elements with keys
23698 equivalent to <tt>k</tt> would be found,
23699 if any such element existed.
23700 Post: <ins>if b.bucket_count() != 0, </ins>the return value shall be
23701 in the range <tt>[0,
23702 b.bucket_count())</tt><ins>, otherwise 0</ins>.</td>
23703 <td>Constant</td>
23704 </tr>
23705
23706 </tbody></table>
23707 </blockquote>
23708
23709 </blockquote>
23710
23711
23712
23713 <p><b>Proposed resolution:</b></p>
23714 <p>
23715 </p>
23716
23717
23718
23719
23720
23721 <hr>
23722 <h3><a name="1198"></a>1198. Container adaptor swap: member or non-member?</h3>
23723 <p><b>Section:</b> 23.3.5 [container.adaptors] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
23724  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2009-08-26  <b>Last modified:</b> 2009-09-30</p>
23725 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.adaptors">active issues</a> in [container.adaptors].</p>
23726 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.adaptors">issues</a> in [container.adaptors].</p>
23727 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
23728 <p><b>Discussion:</b></p>
23729 <p>
23730 Under 23.3.5 [container.adaptors] of
23731 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>
23732 the member function of <tt>swap</tt> of <tt>queue</tt> and <tt>stack</tt> call:
23733 </p>
23734
23735 <blockquote><pre>swap(c, q.c);
23736 </pre></blockquote>
23737
23738 <p>
23739 But under 23.3.5 [container.adaptors] of
23740 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>
23741 these members are specified to call:
23742 </p>
23743
23744 <blockquote><pre>c.swap(q.c);
23745 </pre></blockquote>
23746
23747 <p>
23748 Neither draft specifies the semantics of member <tt>swap</tt> for
23749 <tt>priority_queue</tt> though it is declared.
23750 </p>
23751
23752 <p>
23753 Although the distinction between member <tt>swap</tt> and non-member
23754 <tt>swap</tt> is not important when these adaptors are adapting standard
23755 containers, it may be important for user-defined containers.
23756 </p>
23757 <p>
23758 We (Pablo and Howard) feel that
23759 it is more likely for a user-defined container to support a namespace scope
23760 <tt>swap</tt> than a member <tt>swap</tt>, and therefore these adaptors
23761 should use the container's namespace scope <tt>swap</tt>.
23762 </p>
23763
23764 <p><i>[
23765 2009-09-30 Daniel adds:
23766 ]</i></p>
23767
23768
23769 <blockquote>
23770 The outcome of this issue should be considered with the outcome of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#774">774</a> both in style and in content (e.g. <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#774">774</a> bullet 9
23771 suggests to define the semantic of <tt>void
23772 priority_queue::swap(priority_queue&amp;)</tt> in terms of the member
23773 <tt>swap</tt> of the container).
23774 </blockquote>
23775
23776
23777
23778 <p><b>Proposed resolution:</b></p>
23779 <p><i>[
23780 Changes written with respect to
23781 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>.
23782 ]</i></p>
23783
23784
23785 <p>
23786 Change 23.3.5.1.1 [queue.defn]:
23787 </p>
23788
23789 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23790 class queue {
23791    ...
23792    void swap(queue&amp;<del>&amp;</del> q) { <ins>using std::swap;</ins>
23793                           <del>c.</del>swap(<ins>c, </ins>q.c); }
23794    ...
23795 };
23796 </pre></blockquote>
23797
23798 <p>
23799 Change 23.3.5.2 [priority.queue]:
23800 </p>
23801
23802 <blockquote><pre>template &lt;class T, class Container = vector&lt;T&gt;,
23803           class Compare = less&lt;typename Container::value_type&gt; &gt;
23804 class priority_queue {
23805     ...
23806     void swap(priority_queue&amp;<del>&amp;</del> <ins>q</ins>)<del>;</del> <ins>{ using std::swap;</ins>
23807                                      <ins>swap(c, q.c);</ins>
23808                                      <ins>swap(comp, q.comp); }</ins>
23809     ...
23810 };
23811 </pre></blockquote>
23812
23813 <p>
23814 Change 23.3.5.3.1 [stack.defn]:
23815 </p>
23816
23817 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23818 class stack {
23819    ...
23820    void swap(stack&amp;<del>&amp;</del> s) { <ins>using std::swap;</ins>
23821                           <del>c.</del>swap(<ins>c, </ins>s.c); }
23822    ...
23823 };
23824 </pre></blockquote>
23825
23826
23827
23828
23829
23830
23831 <hr>
23832 <h3><a name="1199"></a>1199. Missing extended copy constructor in container adaptors</h3>
23833 <p><b>Section:</b> 23.3.5 [container.adaptors] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
23834  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2009-08-26  <b>Last modified:</b> 2009-08-31</p>
23835 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#container.adaptors">active issues</a> in [container.adaptors].</p>
23836 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#container.adaptors">issues</a> in [container.adaptors].</p>
23837 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
23838 <p><b>Discussion:</b></p>
23839 <p>
23840 <tt>queue</tt> has a constructor:
23841 </p>
23842
23843 <blockquote><pre>template &lt;class Alloc&gt;
23844   queue(queue&amp;&amp;, const Alloc&amp;);
23845 </pre></blockquote>
23846
23847 <p>
23848 but it is missing a corresponding constructor:
23849 </p>
23850
23851 <blockquote><pre>template &lt;class Alloc&gt;
23852   queue(const queue&amp;, const Alloc&amp;);
23853 </pre></blockquote>
23854
23855 <p>
23856 The same is true of <tt>priority_queue</tt>, and <tt>stack</tt>.  This
23857 "extended copy constructor" is needed for consistency and to ensure that the
23858 user of a container adaptor can always specify the allocator for his adaptor.
23859 </p>
23860
23861
23862
23863 <p><b>Proposed resolution:</b></p>
23864 <p><i>[
23865 This resolution has been harmonized with the proposed resolution to issue
23866 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1194">1194</a>
23867 ]</i></p>
23868
23869
23870 <p>Change 23.3.5.1.1 [queue.defn], p1:</p>
23871
23872 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23873 class queue {
23874 public:
23875   typedef typename Container::value_type      value_type;
23876   typedef typename Container::reference       reference;
23877   typedef typename Container::const_reference const_reference;
23878   typedef typename Container::size_type       size_type;
23879   typedef Container                           container_type;
23880 protected:
23881   Container c;
23882
23883 public:
23884   explicit queue(const Container&amp;);
23885   explicit queue(Container&amp;&amp; = Container());
23886   queue(queue&amp;&amp; q);
23887
23888   template &lt;class Alloc&gt; explicit queue(const Alloc&amp;);
23889   template &lt;class Alloc&gt; queue(const Container&amp;, const Alloc&amp;);
23890   template &lt;class Alloc&gt; queue(Container&amp;&amp;, const Alloc&amp;);
23891   <ins>template &lt;class Alloc&gt; queue(const queue&amp;, const Alloc&amp;);</ins>
23892   template &lt;class Alloc&gt; queue(queue&amp;&amp;, const Alloc&amp;);
23893   queue&amp; operator=(queue&amp;&amp; q);
23894
23895   bool empty() const          { return c.empty(); }
23896   ...
23897 };
23898 </pre></blockquote>
23899
23900 <p>
23901 To the new section  [queue.cons], introduced
23902 in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1194">1194</a>, add:
23903 </p>
23904
23905 <blockquote>
23906
23907 <pre>template &lt;class Alloc&gt;
23908   queue(const queue&amp; q, const Alloc&amp; a);
23909 </pre>
23910
23911 <blockquote><p>
23912 <i>Effects:</i> Initializes <tt>c</tt> with <tt>q.c</tt> as the
23913 first argument and <tt>a</tt> as the second argument.
23914 </p></blockquote>
23915
23916 </blockquote>
23917
23918 <p>Change 23.3.5.2 [priority.queue] as follows (I've an included an editorial change to
23919   move the poorly-placed move-assignment operator):</p>
23920
23921 <blockquote><pre>template &lt;class T, class Container = vector&lt;T&gt;,
23922           class Compare = less&lt;typename Container::value_type&gt; &gt;
23923 class priority_queue {
23924 public:
23925   typedef typename Container::value_type      value_type;
23926   typedef typename Container::reference       reference;
23927   typedef typename Container::const_reference const_reference;
23928   typedef typename Container::size_type       size_type;
23929   typedef          Container                  container_type;
23930 protected:
23931   Container c;
23932   Compare comp;
23933
23934 public:
23935   priority_queue(const Compare&amp; x, const Container&amp;);
23936   explicit priority_queue(const Compare&amp; x = Compare(), Container&amp;&amp; = Container());
23937   template &lt;class InputIterator&gt;
23938     priority_queue(InputIterator first, InputIterator last,
23939                    const Compare&amp; x, const Container&amp;);
23940   template &lt;class InputIterator&gt;
23941     priority_queue(InputIterator first, InputIterator last,
23942                    const Compare&amp; x = Compare(), Container&amp;&amp; = Container());
23943   priority_queue(priority_queue&amp;&amp;);
23944   <del>priority_queue&amp; operator=(priority_queue&amp;&amp;);</del>
23945   template &lt;class Alloc&gt; explicit priority_queue(const Alloc&amp;);
23946   template &lt;class Alloc&gt; priority_queue(const Compare&amp;, const Alloc&amp;);
23947   template &lt;class Alloc&gt; priority_queue(const Compare&amp;,
23948                                         const Container&amp;, const Alloc&amp;);
23949   template &lt;class Alloc&gt; priority_queue(const Compare&amp;,
23950                                         Container&amp;&amp;, const Alloc&amp;);
23951   <ins>template &lt;class Alloc&gt; priority_queue(const priority_queue&amp;, const Alloc&amp;);</ins>
23952   template &lt;class Alloc&gt; priority_queue(priority_queue&amp;&amp;, const Alloc&amp;);
23953
23954   <ins>priority_queue&amp; operator=(priority_queue&amp;&amp;);</ins>
23955   ...
23956 };
23957 </pre></blockquote>
23958
23959 <p>
23960 Add to 23.3.5.2.1 [priqueue.cons]:
23961 </p>
23962
23963 <blockquote>
23964
23965 <pre>template &lt;class Alloc&gt;
23966   explicit priority_queue(const priority_queue&amp; q, const Alloc&amp; a);
23967 </pre>
23968
23969 <blockquote><p>
23970 <i>Effects:</i> Initializes <tt>c</tt> with <tt>q.c</tt> as the
23971 first argument and <tt>a</tt> as the second argument,
23972 and initializes <tt>comp</tt> with <tt>q.comp</tt>.
23973 </p></blockquote>
23974
23975 </blockquote>
23976
23977 <p>
23978 Change 23.3.5.3.1 [stack.defn]:
23979 </p>
23980
23981 <blockquote><pre>template &lt;class T, class Container = deque&lt;T&gt; &gt;
23982 class stack {
23983 public:
23984   typedef typename Container::value_type      value_type;
23985   typedef typename Container::reference       reference;
23986   typedef typename Container::const_reference const_reference;
23987   typedef typename Container::size_type       size_type;
23988   typedef Container                           container_type;
23989 protected:
23990   Container c;
23991
23992 public:
23993   explicit stack(const Container&amp;);
23994   explicit stack(Container&amp;&amp; = Container());
23995   stack(stack&amp;&amp; s);
23996
23997   template &lt;class Alloc&gt; explicit stack(const Alloc&amp;);
23998   template &lt;class Alloc&gt; stack(const Container&amp;, const Alloc&amp;);
23999   template &lt;class Alloc&gt; stack(Container&amp;&amp;, const Alloc&amp;);
24000   <ins>template &lt;class Alloc&gt; stack(const stack&amp;, const Alloc&amp;);</ins>
24001   template &lt;class Alloc&gt; stack(stack&amp;&amp;, const Alloc&amp;);
24002   stack&amp; operator=(stack&amp;&amp; s);
24003
24004   bool empty() const          { return c.empty(); }
24005   ...
24006 };
24007 </pre></blockquote>
24008
24009 <p>
24010 To the new section  [stack.cons], introduced
24011 in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1194">1194</a>, add:
24012 </p>
24013
24014 <blockquote>
24015
24016 <pre>template &lt;class Alloc&gt;
24017   stack(const stack&amp; s, const Alloc&amp; a);
24018 </pre>
24019
24020 <blockquote><p>
24021 <i>Effects:</i> Initializes <tt>c</tt> with <tt>s.c</tt> as the
24022 first argument and <tt>a</tt> as the second argument.
24023 </p></blockquote>
24024 </blockquote>
24025
24026
24027
24028
24029
24030
24031 <hr>
24032 <h3><a name="1200"></a>1200. "surprising" <tt>char_traits&lt;T&gt;::int_type</tt> requirements</h3>
24033 <p><b>Section:</b> 21.2.2 [char.traits.typedefs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
24034  <b>Submitter:</b> Sean Hunt <b>Opened:</b> 2009-09-03  <b>Last modified:</b> 2009-10-28</p>
24035 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#char.traits.typedefs">issues</a> in [char.traits.typedefs].</p>
24036 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
24037 <p><b>Discussion:</b></p>
24038 <p>
24039 The footnote for <tt>int_type</tt> in 21.2.2 [char.traits.typedefs] says that
24040 </p>
24041
24042 <blockquote>
24043 If <tt>eof()</tt>
24044 can be held in <tt>char_type</tt> then some iostreams implementations may give
24045 surprising results.
24046 </blockquote>
24047
24048 <p>
24049 This implies that <tt>int_type</tt> should be a superset of
24050 <tt>char_type</tt>. However, the requirements for <tt>char16_t</tt> and <tt>char32_t</tt> define
24051 <tt>int_type</tt> to be equal to <tt>int_least16_t</tt> and <tt>int_least32_t</tt> respectively.
24052 <tt>int_least16_t</tt> is likely to be the same size as <tt>char_16_t</tt>, which may lead
24053 to surprising behavior, even if <tt>eof()</tt> is not a valid UTF-16 code unit.
24054 The standard should not prescribe surprising behavior, especially
24055 without saying what it is (it's apparently not undefined, just
24056 surprising). The same applies for 32-bit types.
24057 </p>
24058
24059 <p>
24060 I personally recommend that behavior be undefined if <tt>eof()</tt> is a member
24061 of <tt>char_type</tt>, and another type be chosen for <tt>int_type</tt> (my personal
24062 favorite has always been a <tt>struct {bool eof; char_type c;}</tt>).
24063 Alternatively, the exact results of such a situation should be defined,
24064 at least so far that I/O could be conducted on these types as long as
24065 the code units remain valid. Note that the argument that no one streams
24066 <tt>char16_t</tt> or <tt>char32_t</tt> is not really valid as it would be perfectly
24067 reasonable to use a <tt>basic_stringstream</tt> in conjunction with UTF character
24068 types.
24069 </p>
24070
24071 <p><i>[
24072 2009-10-28 Ganesh provides two possible resolutions and expresses a preference
24073 for the second:
24074 ]</i></p>
24075
24076
24077 <blockquote>
24078 <ol>
24079 <li>
24080 <p>
24081 Replace 21.2.3.2 [char.traits.specializations.char16_t] para 3 with:
24082 </p>
24083
24084 <blockquote>
24085 The member <tt>eof()</tt> shall return <del>an implementation-defined
24086 constant that cannot appear as a valid UTF-16 code unit</del>
24087 <ins><tt>UINT_LEAST16_MAX</tt> [<i>Note:</i> this value is guaranteed to
24088 be a permanently reserved UCS-2 code position if <tt>UINT_LEAST16_MAX ==
24089 0xFFFF</tt> and it's not a UCS-2 code position otherwise &#8212; <i>end
24090 note</i>]</ins>.
24091 </blockquote>
24092
24093 <p>
24094 Replace 21.2.3.3 [char.traits.specializations.char32_t] para 3 with:
24095 </p>
24096
24097 <blockquote>
24098 The member <tt>eof()</tt> shall return <del>an implementation-defined constant that
24099 cannot appear as a Unicode code point</del>
24100 <ins>
24101 <tt>UINT_LEAST32_MAX</tt> [<i>Note:</i> this value is guaranteed to be a
24102 permanently reserved UCS-4 code position if <tt>UINT_LEAST32_MAX ==
24103 0xFFFFFFFF</tt> and it's not a UCS-4 code position otherwise &#8212; <i>end
24104 note</i>]</ins>.
24105 </blockquote>
24106 </li>
24107 <li>
24108 <p>
24109 In 21.2.3.2 [char.traits.specializations.char16_t], in the
24110 definition of <tt>char_traits&lt;char16_t&gt;</tt> replace the definition of nested
24111 typedef <tt>int_type</tt> with:
24112 </p>
24113
24114 <blockquote><pre>namespace std {
24115   template&lt;&gt; struct char_traits&lt;char16_t&gt; {
24116     typedef char16_t         char_type;
24117     typedef <del>uint_least16_t</del> <ins>uint_fast16_t</ins> int_type;
24118      ...
24119 </pre></blockquote>
24120
24121 <p>
24122 Replace 21.2.3.2 [char.traits.specializations.char16_t] para 3 with:
24123 </p>
24124
24125 <blockquote>
24126 The member <tt>eof()</tt> shall return <del>an implementation-defined
24127 constant that cannot appear as a valid UTF-16 code unit</del>
24128 <ins><tt>UINT_FAST16_MAX</tt> [<i>Note:</i> this value is guaranteed to
24129 be a permanently reserved UCS-2 code position if <tt>UINT_FAST16_MAX ==
24130 0xFFFF</tt> and it's not a UCS-2 code position otherwise &#8212; <i>end
24131 note</i>]</ins>.
24132 </blockquote>
24133
24134 <p>
24135 In 21.2.3.3 [char.traits.specializations.char32_t], in the
24136 definition of <tt>char_traits&lt;char32_t&gt;</tt> replace the definition of nested
24137 typedef <tt>int_type</tt> with:
24138 </p>
24139
24140 <blockquote><pre>namespace std {
24141   template&lt;&gt; struct char_traits&lt;char32_t&gt; {
24142     typedef char32_t         char_type;
24143     typedef <del>uint_least32_t</del> <ins>uint_fast32_t</ins> int_type;
24144      ...
24145 </pre></blockquote>
24146
24147 <p>
24148 Replace 21.2.3.3 [char.traits.specializations.char32_t] para 3 with:
24149 </p>
24150
24151 <blockquote>
24152 The member <tt>eof()</tt> shall return <del>an implementation-defined constant that
24153 cannot appear as a Unicode code point</del>
24154 <ins>
24155 <tt>UINT_FAST32_MAX</tt> [<i>Note:</i> this value is guaranteed to be a
24156 permanently reserved UCS-4 code position if <tt>UINT_FAST32_MAX ==
24157 0xFFFFFFFF</tt> and it's not a UCS-4 code position otherwise &#8212; <i>end
24158 note</i>]</ins>.
24159 </blockquote>
24160 </li>
24161 </ol>
24162 </blockquote>
24163
24164
24165
24166 <p><b>Proposed resolution:</b></p>
24167 <p>
24168 </p>
24169
24170
24171
24172
24173
24174 <hr>
24175 <h3><a name="1201"></a>1201. Do we always want to unwrap <tt>ref</tt>-wrappers in <tt>make_tuple</tt></h3>
24176 <p><b>Section:</b> 20.5.2.4 [tuple.creation], 20.3.4 [pairs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Future">Tentatively NAD Future</a>
24177  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-05  <b>Last modified:</b> 2009-10-26</p>
24178 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Future">Tentatively NAD Future</a> status.</p>
24179 <p><b>Discussion:</b></p>
24180 <p>
24181 Spotting a recent thread on the boost lists regarding collapsing
24182 optional representations in <tt>optional&lt;optional&lt;T&gt;&gt;</tt> instances, I wonder if
24183 we have some of the same issues with <tt>make_tuple</tt>, and now <tt>make_pair</tt>?
24184 </p>
24185
24186 <p>
24187 Essentially, if my generic code in my own library is handed a
24188 <tt>reference_wrapper</tt> by a user, and my library in turn delegates some logic
24189 to <tt>make_pair</tt> or <tt>make_tuple</tt>, then I am going to end up with a <tt>pair</tt>/<tt>tuple</tt>
24190 holding a real reference rather than the intended reference wrapper.
24191 </p>
24192
24193 <p>
24194 There are two things as a library author I can do at this point:
24195 </p>
24196
24197 <ol type="i">
24198 <li>
24199 document my library also has the same reference-wrapper behaviour as
24200 <tt>std::make_tuple</tt>
24201 </li>
24202 <li>
24203 roll my own <tt>make_tuple</tt> that does not unwrap rereferences, a lost
24204 opportunity to re-use the standard library.
24205 </li>
24206 </ol>
24207
24208 <p>
24209 (There may be some metaprogramming approaches my library can use to wrap
24210 the <tt>make_tuple</tt> call, but all will be significantly more complex than
24211 simply implementing a simplified <tt>make_tuple</tt>.)
24212 </p>
24213
24214 <p>
24215 Now I don't propose we lose this library facility, I think unwrapping
24216 references will be the common behaviour.  However, we might want to
24217 consider adding another overload that does nothing special with
24218 <tt>ref</tt>-wrappers.  Note that we already have a second overload of <tt>make_tuple</tt>
24219 in the library, called <tt>tie</tt>.
24220 </p>
24221
24222 <p><i>[
24223 2009-09-30 Daniel adds:
24224 ]</i></p>
24225
24226
24227 <blockquote>
24228 <p>
24229 I suggest to change the currently proposed paragraph for
24230 <tt>make_simple_pair</tt>
24231 </p>
24232
24233 <blockquote><pre>template&lt;typename... Types&gt;
24234   pair&lt;typename decay&lt;Types&gt;::type...&gt; make_simple_pair(Types&amp;&amp;... t);
24235 </pre>
24236 <blockquote>
24237 <p>
24238 <del><i>Type requirements:</i> <tt>sizeof...(Types) == 2</tt>.</del>
24239 <ins><i>Remarks:</i> The program shall be ill-formed, if
24240 <tt>sizeof...(Types) != 2</tt>.</ins>
24241 </p>
24242 <p>
24243 ...
24244 </p>
24245 </blockquote>
24246 </blockquote>
24247
24248 <p>
24249 or alternatively (but with a slightly different semantic):
24250 </p>
24251
24252 <blockquote>
24253 <blockquote>
24254 <i>Remarks:</i> If <tt>sizeof...(Types) != 2</tt>, this function shall not
24255 participate in overload resolution.
24256 </blockquote>
24257 </blockquote>
24258
24259 <p>
24260 to follow a currently introduced style and because the library does
24261 not have yet a specific "<i>Type requirements</i>" element. If such thing
24262 would be considered as useful this should be done as a separate
24263 issue. Given the increasing complexity of either of these wordings
24264 it might be preferable to use the normal two-argument-declaration
24265 style again in either of the following ways:
24266 </p>
24267
24268 <ol type="A">
24269 <li>
24270 <pre>template&lt;class T1, class T2&gt;
24271 pair&lt;typename decay&lt;T1&gt;::type, typename decay&lt;T2&gt;::type&gt;
24272 make_simple_pair(T1&amp;&amp; t1, T2&amp;&amp; t2);
24273 </pre>
24274 </li>
24275 <li>
24276 <pre>template&lt;class T1, class T2&gt;
24277 pair&lt;V1, V2&gt; make_simple_pair(T1&amp;&amp; t1, T2&amp;&amp; t2);
24278 </pre>
24279 <blockquote>
24280 Let <tt>V1</tt> be <tt>typename decay&lt;T1&gt;::type</tt> and <tt>V2</tt> be
24281 <tt>typename decay&lt;T2&gt;::type</tt>.
24282 </blockquote>
24283 </li>
24284 </ol>
24285
24286 </blockquote>
24287
24288 <p><i>[
24289 2009-10 post-Santa Cruz:
24290 ]</i></p>
24291
24292
24293 <blockquote>
24294 Mark as Tentatively NAD Future.
24295 </blockquote>
24296
24297
24298
24299 <p><b>Proposed resolution:</b></p>
24300 <p>
24301 Add the following function to 20.3.4 [pairs] and signature in
24302 appropriate synopses:
24303 </p>
24304
24305 <blockquote><pre>template&lt;typename... Types&gt;
24306   pair&lt;typename decay&lt;Types&gt;::type...&gt; make_simple_pair(Types&amp;&amp;... t);
24307 </pre>
24308 <blockquote>
24309 <p>
24310 <i>Type requirements:</i> <tt>sizeof...(Types) == 2</tt>.
24311 </p>
24312 <p>
24313 <i>Returns:</i> <tt>pair&lt;typename decay&lt;Types&gt;::type...&gt;(std::forward&lt;Types&gt;(t)...)</tt>.
24314 </p>
24315 </blockquote>
24316 </blockquote>
24317
24318 <p><i>[
24319 Draughting note: I chose a variadic representation similar to <tt>make_tuple</tt>
24320 rather than naming both types as it is easier to read through the
24321 clutter of metaprogramming this way.  Given there are exactly two
24322 elements, the committee may prefer to draught with two explicit template
24323 type parameters instead
24324 ]</i></p>
24325
24326
24327 <p>
24328 Add the following function to 20.5.2.4 [tuple.creation] and
24329 signature in appropriate synopses:
24330 </p>
24331
24332 <blockquote><pre>template&lt;typename... Types&gt;
24333   tuple&lt;typename decay&lt;Types&gt;::type...&gt; make_simple_tuple(Types&amp;&amp;... t);
24334 </pre>
24335 <blockquote>
24336 <p>
24337 <i>Returns:</i> <tt>tuple&lt;typename decay&lt;Types&gt;::type...&gt;(std::forward&lt;Types&gt;(t)...)</tt>.
24338 </p>
24339 </blockquote>
24340 </blockquote>
24341
24342
24343
24344
24345
24346 <hr>
24347 <h3><a name="1202"></a>1202. <tt>integral_constant</tt> needs a spring clean</h3>
24348 <p><b>Section:</b> 20.6.3 [meta.help] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
24349  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-05  <b>Last modified:</b> 2009-09-06</p>
24350 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.help">issues</a> in [meta.help].</p>
24351 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
24352 <p><b>Discussion:</b></p>
24353 <p>
24354 The specification of <tt>integral_constant</tt> has been inherited
24355 essentially unchanged from TR1:
24356 </p>
24357
24358 <blockquote><pre>template &lt;class T, T v&gt;
24359 struct integral_constant {
24360   static const T value = v;
24361   typedef T value_type;
24362   typedef integral_constant&lt;T,v&gt; type;
24363 };
24364 </pre></blockquote>
24365
24366 <p>
24367 In light of 0x language changes there are several things we might
24368 consider changing, notably the form of specification for value.
24369 </p>
24370
24371 <p>
24372 The current form requires a static data member have storage allocated
24373 for it, where we could now implement without this using the new enum
24374 syntax:
24375 </p>
24376
24377 <blockquote><pre>template &lt;class T, T v&gt;
24378 struct integral_constant {
24379   <b>enum : T { value = v };</b>
24380   typedef T value_type;
24381   typedef integral_constant type;
24382 };
24383 </pre></blockquote>
24384
24385 <p>
24386 The effective difference between these two implementation is:
24387 </p>
24388
24389 <ol type="i">
24390 <li>
24391 No requirement to allocate storage for data member (which we hope but do
24392 not guarantee compilers strip today)
24393 </li>
24394
24395 <li>
24396 You can no longer take the address of the constant as
24397 <tt>&amp;integral_constant&lt;T,v&gt;::value;</tt>
24398 </li>
24399 </ol>
24400
24401 <p>
24402 Also note the editorial change to drop the explicit qualification of
24403 <tt>integral_constant</tt> in the <tt>typedef type</tt>.  This makes it quite clear we
24404 mean the current instantiation, and cannot be mistaken for a recursive
24405 metaprogram.
24406 </p>
24407
24408 <p>
24409 Even if we don't mandate this implementation, it would be nice to give
24410 vendors freedom under QoI to choose their preferred representation.
24411 </p>
24412
24413 <p>
24414 The other side of this issue is if we choose to retain the static
24415 constant form.  In that case we should go further and insist on
24416 <tt>constexpr</tt>, much like we did throughout <tt>numeric_limits</tt>:
24417 </p>
24418
24419 <blockquote><pre>template &lt;class T, T v&gt;
24420 struct integral_constant {
24421   static <b>constexpr</b> T value = v;
24422   typedef T value_type;
24423   typedef integral_constant type;
24424 };
24425 </pre></blockquote>
24426
24427 <p>
24428 [Footnote] It turns out <tt>constexpr</tt> is part of the Tentatively Ready
24429 resolution for <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1019">1019</a>.  I don't want to interfere with that issue, but
24430 would like a new issue to consider if the fixed-base enum implementation
24431 should be allowed.
24432 </p>
24433
24434 <p><i>[
24435 2009-09-05 Daniel adds:
24436 ]</i></p>
24437
24438
24439 <blockquote>
24440 <p>
24441 I think that the suggested resolution is incomplete and
24442 may have some possible unwanted side-effects. To understand
24443 why, note that <tt>integral_constant</tt> is <em>completely</em> specified
24444 by code in 20.6.3 [meta.help]. While this is usually considered
24445 as a good thing, let me give a possible user-defined
24446 specialization that would break given the suggested changes:
24447 </p>
24448
24449 <blockquote><pre>enum NodeColor { Red, Black };
24450
24451 std::integral_constant&lt;NodeColor, Red&gt; red;
24452 </pre></blockquote>
24453
24454 <p>
24455 The reason why that breaks is due to the fact that
24456 current core language rules does only allow integral
24457 types as enum-bases, see 7.2 [dcl.enum]/2.
24458 </p>
24459
24460 <p>
24461 So, I think that we cannot leave the implementation the
24462 freedom to decide which way they would like to provide
24463 the implementation, because that is easily user-visible
24464 (I don't speak of addresses, but of instantiation errors),
24465 therefore if applied, this should be either specified or
24466 wording must be added that gives a note about this
24467 freedom of implementation.
24468 </p>
24469
24470 <p>
24471 Another possible disadvantage seems to me that user-expectations
24472 are easy to disappoint if they see a failure
24473 of the test
24474 </p>
24475
24476 <blockquote><pre>assert(typeid(std::integral_constant&lt;int, 0&gt;::value) == typeid(int));
24477 </pre></blockquote>
24478
24479 <p>
24480 or of
24481 </p>
24482
24483 <blockquote><pre>static_assert(std::is_same&lt;decltype(std::integral_constant&lt;int, 0&gt;::value), const int&gt;::value, "Bad library");
24484 </pre></blockquote>
24485
24486 </blockquote>
24487
24488
24489
24490 <p><b>Proposed resolution:</b></p>
24491
24492
24493
24494
24495
24496 <hr>
24497 <h3><a name="1204"></a>1204. Global permission to move</h3>
24498 <p><b>Section:</b> 17.6.3.9 [res.on.arguments] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
24499  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-09-12  <b>Last modified:</b> 2009-10-20</p>
24500 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
24501 <p><b>Discussion:</b></p>
24502 <p>
24503 When a library function binds an rvalue reference parameter to an argument, the
24504 library must be able to assume that the bound argument is a temporary, and not
24505 a moved-from lvalue.  The reason for this is that the library function must be
24506 able to modify that argument without concern that such modifications will corrupt
24507 the logic of the calling code.  For example:
24508 </p>
24509
24510 <blockquote><pre>template &lt;class T, class A&gt;
24511 void
24512 vector&lt;T, A&gt;::push_back(value_type&amp;&amp; v)
24513 {
24514     <font color="#c80000">// This function should move from v, potentially modifying</font>
24515     <font color="#c80000">//   the object v is bound to.</font>
24516 }
24517 </pre></blockquote>
24518
24519 <p>
24520 If <tt>v</tt> is truly bound to a temporary, then <tt>push_back</tt> has the
24521 <em>only</em> reference to this temporary in the entire program.  Thus any
24522 modifications will be invisible to the rest of the program.
24523 </p>
24524
24525 <p>
24526 If the client supplies <tt>std::move(x)</tt> to <tt>push_back</tt>, the onus is
24527 on the client to ensure that the value of <tt>x</tt> is no longer important to
24528 the logic of his program after this statement.  I.e. the client is making a statement
24529 that <tt>push_back</tt> may treat <tt>x</tt> as a temporary.
24530 </p>
24531
24532 <blockquote><em>
24533 The above statement is the very foundation upon which move semantics is based.
24534 </em></blockquote>
24535
24536 <p>
24537 The standard is currently lacking a global statement to this effect.  I propose
24538 the following addition to 17.6.3.9 [res.on.arguments]:
24539 </p>
24540
24541 <blockquote>
24542 <p>
24543 Each of the following statements applies to all arguments to functions
24544 defined in the C++ standard library, unless explicitly stated otherwise.
24545 </p>
24546 <ul>
24547 <li>
24548 If an argument to a function has an invalid value (such as a value
24549 outside the domain of the function, or a pointer invalid for its
24550 intended use), the behavior is undefined.
24551 </li>
24552 <li>
24553 If a function argument is described as being an array, the pointer
24554 actually passed to the function shall have a value such that all address
24555 computations and accesses to objects (that would be valid if the pointer
24556 did point to the first element of such an array) are in fact valid.
24557 </li>
24558 <li><ins>
24559 If a function argument binds to an rvalue reference parameter, the C++
24560 standard library may assume that this parameter is a unique reference
24561 to this argument.  If the parameter is a generic parameter of the
24562 form <tt>T&amp;&amp;</tt>, and an lvalue of type <tt>A</tt> is bound,
24563 then the binding is considered to be to an lvalue reference
24564 (14.9.2.1 [temp.deduct.call]) and thus not covered by this clause.
24565 [<i>Note:</i>
24566 If a program casts an lvalue to an rvalue while passing that lvalue to
24567 a library function (e.g. <tt>move(x)</tt>), then the program is effectively
24568 asking the library to treat that lvalue as a temporary.  The library is at
24569 liberty to optimize away aliasing checks which might be needed if the argument
24570 were an lvalue.
24571 &#8212; <i>end note</i>]
24572 </ins></li>
24573 </ul>
24574
24575 </blockquote>
24576
24577 <p>
24578 Such a global statement will eliminate the need for piecemeal statements such as
24579 23.2.1 [container.requirements.general]/13:
24580 </p>
24581
24582 <blockquote>
24583 An object bound to an rvalue reference parameter of a member function of
24584 a container shall not be an element of that container; no diagnostic
24585 required.
24586 </blockquote>
24587
24588 <p>
24589 Additionally this clarifies that move assignment operators need not perform the
24590 traditional <tt>if (this != &amp;rhs)</tt> test commonly found (and needed) in
24591 copy assignment operators.
24592 </p>
24593
24594 <p><i>[
24595 2009-09-13 Niels adds:
24596 ]</i></p>
24597
24598
24599 <blockquote>
24600 Note: This resolution supports the change of 27.9.1.3 [filebuf.assign]/1,
24601 proposed by LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#900">900</a>.
24602 </blockquote>
24603
24604 <p><i>[
24605 2009 Santa Cruz:
24606 ]</i></p>
24607
24608
24609 <blockquote>
24610 Move to Ready.
24611 </blockquote>
24612
24613
24614
24615 <p><b>Proposed resolution:</b></p>
24616 <p>
24617 Add a bullet to 17.6.3.9 [res.on.arguments]:
24618 </p>
24619
24620 <blockquote>
24621 <p>
24622 Each of the following statements applies to all arguments to functions
24623 defined in the C++ standard library, unless explicitly stated otherwise.
24624 </p>
24625 <ul>
24626 <li>
24627 If an argument to a function has an invalid value (such as a value
24628 outside the domain of the function, or a pointer invalid for its
24629 intended use), the behavior is undefined.
24630 </li>
24631 <li>
24632 If a function argument is described as being an array, the pointer
24633 actually passed to the function shall have a value such that all address
24634 computations and accesses to objects (that would be valid if the pointer
24635 did point to the first element of such an array) are in fact valid.
24636 </li>
24637 <li><ins>
24638 If a function argument binds to an rvalue reference parameter, the C++
24639 standard library may assume that this parameter is a unique reference
24640 to this argument.  If the parameter is a generic parameter of the
24641 form <tt>T&amp;&amp;</tt>, and an lvalue of type <tt>A</tt> is bound,
24642 then the binding is considered to be to an lvalue reference
24643 (14.9.2.1 [temp.deduct.call]) and thus not covered by this clause.
24644 [<i>Note:</i>
24645 If a program casts an lvalue to an rvalue while passing that lvalue to
24646 a library function (e.g. <tt>move(x)</tt>), then the program is effectively
24647 asking the library to treat that lvalue as a temporary.  The library is at
24648 liberty to optimize away aliasing checks which might be needed if the argument
24649 were an lvalue.
24650 &#8212; <i>end note</i>]
24651 </ins></li>
24652 </ul>
24653 </blockquote>
24654
24655 <p>
24656 Delete 23.2.1 [container.requirements.general]/13:
24657 </p>
24658
24659 <blockquote><del>
24660 An object bound to an rvalue reference parameter of a member function of
24661 a container shall not be an element of that container; no diagnostic
24662 required.
24663 </del></blockquote>
24664
24665
24666
24667
24668
24669
24670 <hr>
24671 <h3><a name="1205"></a>1205. Some algorithms could more clearly document their handling of empty ranges</h3>
24672 <p><b>Section:</b> 25 [algorithms] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
24673  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-13  <b>Last modified:</b> 2009-09-13</p>
24674 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#algorithms">active issues</a> in [algorithms].</p>
24675 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#algorithms">issues</a> in [algorithms].</p>
24676 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
24677 <p><b>Discussion:</b></p>
24678 <p>
24679 There are a number of algorithms whose result might depend on the
24680 handling of an empty range.  In some cases the result is not clear,
24681 while in others it would help readers to clearly mention the result
24682 rather than require some subtle intuition of the supplied wording.
24683 </p>
24684
24685 <p>
24686 25.2.1 [alg.all_of]
24687 </p>
24688
24689 <blockquote>
24690 <i>Returns:</i> <tt>true</tt> if <tt>pred(*i)</tt> is <tt>true</tt> for every
24691 iterator <tt>i</tt> in the range <tt>[first,last)</tt>, ...
24692 </blockquote>
24693
24694 <p>
24695 What does this mean if the range is empty?
24696 </p>
24697
24698 <p>
24699 I believe that we intend this to be <tt>true</tt> and suggest a
24700 non-normative note to clarify:
24701 </p>
24702
24703 <p>
24704 Add to p1 25.2.1 [alg.all_of]:
24705 </p>
24706
24707 <blockquote>
24708 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
24709 &#8212; <i>end note</i>]
24710 </blockquote>
24711
24712 <p>
24713 25.2.3 [alg.none_of]
24714 </p>
24715
24716 <blockquote>
24717 <i>Returns:</i> <tt>true</tt> if <tt>pred(*i)</tt> is <tt>false</tt> for every
24718 iterator <tt>i</tt> in the range <tt>[first,last)</tt>, ...
24719 </blockquote>
24720
24721 <p>
24722 What does this mean if the range empty?
24723 </p>
24724
24725 <p>
24726 I believe that we intend this to be <tt>true</tt> and suggest a
24727 non-normative note to clarify:
24728 </p>
24729
24730 <p>
24731 Add to p1 25.2.3 [alg.none_of]:
24732 </p>
24733
24734 <blockquote>
24735 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
24736 &#8212; <i>end note</i>]
24737 </blockquote>
24738
24739 <p>
24740 25.2.2 [alg.any_of]
24741 </p>
24742
24743 <p>
24744 The specification for an empty range is actually fairly clear in this
24745 case, but a note wouldn't hurt and would be consistent with proposals
24746 for <tt>all_of</tt>/<tt>none_of</tt> algorithms.
24747 </p>
24748
24749 <p>
24750 Add to p1 25.2.2 [alg.any_of]:
24751 </p>
24752
24753 <blockquote>
24754 [<i>Note:</i> Returns <tt>false</tt> if <tt>[first,last)</tt> is empty.
24755 &#8212; <i>end note</i>]
24756 </blockquote>
24757
24758 <p>
24759 25.2.6 [alg.find.end]
24760 </p>
24761
24762 <p>
24763 what does this mean if <tt>[first2,last2)</tt> is empty?
24764 </p>
24765
24766 <p>
24767 I believe the wording suggests the algorithm should return
24768 <tt>last1</tt> in this case, but am not 100% sure. Is this in fact the
24769 correct result anyway? Surely an empty range should always match and the
24770 naive expected result would be <tt>first1</tt>?
24771 </p>
24772
24773 <p>
24774 My proposed wording is a note to clarify the current semantic:
24775 </p>
24776
24777 <p>
24778 Add to p2 25.2.6 [alg.find.end]:
24779 </p>
24780
24781 <blockquote>
24782 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
24783 empty. &#8212; <i>end note</i>]
24784 </blockquote>
24785
24786 <p>
24787 I would prefer a normative wording treating empty ranges specially, but
24788 do not believe we can change semantics at this point in the process,
24789 unless existing implementations actually yield this result:
24790 </p>
24791
24792 <p>
24793 Alternative wording: (NOT a note)
24794 </p>
24795 <p>
24796 Add to p2 25.2.6 [alg.find.end]:
24797 </p>
24798 <blockquote>
24799 Returns <tt>first1</tt> if <tt>[first2,last2)</tt> is empty.
24800 </blockquote>
24801
24802 <p>
24803 25.2.7 [alg.find.first.of]
24804 </p>
24805
24806 <p>
24807 The phrasing seems precise when <tt>[first2, last2)</tt> is empty, but a small
24808 note to confirm the reader's understanding might still help.
24809 </p>
24810
24811 <p>
24812 Add to p2 25.2.7 [alg.find.first.of]
24813 </p>
24814 <blockquote>
24815 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
24816 empty. &#8212; <i>end note</i>]
24817 </blockquote>
24818
24819 <p>
24820 25.2.12 [alg.search]
24821 </p>
24822
24823 <p>
24824 What is the expected result if <tt>[first2, last2)</tt> is empty?
24825 </p>
24826
24827 <p>
24828 I believe the wording suggests the algorithm should return <tt>last1</tt> in this
24829 case, but am not 100% sure. Is this in fact the correct result anyway?
24830 Surely an empty range should always match and the naive expected result
24831 would be <tt>first1</tt>?
24832 </p>
24833
24834 <p>
24835 My proposed wording is a note to clarify the current semantic:
24836 </p>
24837
24838 <p>
24839 Add to p2 25.2.12 [alg.search]:
24840 </p>
24841
24842 <blockquote>
24843 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
24844 empty. &#8212; <i>end note</i>]
24845 </blockquote>
24846
24847 <p>
24848 Again, I would prefer a normative wording treating empty ranges
24849 specially, but do not believe we can change semantics at this point in
24850 the process, unless existing implementations actually yield this result:
24851 </p>
24852
24853 <p>
24854 Alternative wording: (NOT a note)
24855 </p>
24856 <p>
24857 Add to p2 25.2.12 [alg.search]:
24858 </p>
24859
24860 <blockquote>
24861 Returns <tt>first1</tt> if <tt>[first2,last2)</tt> is empty.
24862 </blockquote>
24863
24864 <p>
24865 25.3.13 [alg.partitions]
24866 </p>
24867
24868 <p>
24869 Is an empty range partitioned or not?
24870 </p>
24871
24872 <p>
24873 Proposed wording:
24874 </p>
24875
24876 <p>
24877 Add to p1 25.3.13 [alg.partitions]:
24878 </p>
24879
24880 <blockquote>
24881 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
24882 &#8212; <i>end note</i>]
24883 </blockquote>
24884
24885 <p>
24886 25.4.5.1 [includes]
24887 </p>
24888
24889 <blockquote>
24890 <i>Returns:</i> <tt>true</tt> if every element in the range
24891 <tt>[first2,last2)</tt> is contained in the range
24892 <tt>[first1,last1)</tt>. ...
24893 </blockquote>
24894
24895 <p>
24896 I really don't know what this means if <tt>[first2,last2)</tt> is empty.
24897 I could loosely guess that this implies empty ranges always match, and
24898 my proposed wording is to clarify exactly that:
24899 </p>
24900
24901 <p>
24902 Add to p1 25.4.5.1 [includes]:
24903 </p>
24904
24905 <blockquote>
24906 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first2,last2)</tt> is empty.
24907 &#8212; <i>end note</i>]
24908 </blockquote>
24909
24910 <p>
24911 25.4.6.2 [pop.heap]
24912 </p>
24913
24914 <p>
24915 The effects clause is invalid if the range <tt>[first,last)</tt> is empty, unlike
24916 all the other heap alogorithms.  The should be called out in the
24917 requirements.
24918 </p>
24919
24920 <p>
24921 Proposed wording:
24922 </p>
24923 <p>
24924 Revise p2 25.4.6.2 [pop.heap]
24925 </p>
24926
24927 <blockquote>
24928 <i>Requires:</i> The range <tt>[first,last)</tt> shall be a valid
24929 <ins>non-empty</ins> heap.
24930 </blockquote>
24931
24932 <p>
24933 [Editorial] Reverse order of 25.4.6.2 [pop.heap] p1 and p2.
24934 </p>
24935
24936 <p>
24937 25.4.7 [alg.min.max]
24938 </p>
24939
24940 <p>
24941 <tt>minmax_element</tt> does not clearly specify behaviour for an empty
24942 range in the same way that <tt>min_element</tt> and <tt>max_element</tt> do.
24943 </p>
24944
24945 <p>
24946 Add to p31 25.4.7 [alg.min.max]:
24947 </p>
24948
24949 <blockquote>
24950 Returns <tt>make_pair(first, first)</tt> if <tt>first == last</tt>.
24951 </blockquote>
24952
24953 <p>
24954 25.4.8 [alg.lex.comparison]
24955 </p>
24956
24957 <p>
24958 The wording here seems quite clear, especially with the sample algorithm
24959 implementation.  A note is recommended purely for consistency with the
24960 rest of these issue resolutions:
24961 </p>
24962
24963 <p>
24964 Add to p1 25.4.8 [alg.lex.comparison]:
24965 </p>
24966
24967 <blockquote>
24968 [<i>Note:</i> An empty sequence is lexicographically less than any other
24969 non-empty sequence, but not to another empty sequence. &#8212; <i>end note</i>]
24970 </blockquote>
24971
24972
24973 <p><b>Proposed resolution:</b></p>
24974 <p>
24975 Add to p1 25.2.1 [alg.all_of]:
24976 </p>
24977 <blockquote>
24978 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
24979 &#8212; <i>end note</i>]
24980 </blockquote>
24981
24982 <p>
24983 Add to p1 25.2.2 [alg.any_of]:
24984 </p>
24985 <blockquote>
24986 [<i>Note:</i> Returns <tt>false</tt> if <tt>[first,last)</tt> is empty.
24987 &#8212; <i>end note</i>]
24988 </blockquote>
24989
24990 <p>
24991 Add to p1 25.2.3 [alg.none_of]:
24992 </p>
24993 <blockquote>
24994 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
24995 &#8212; <i>end note</i>]
24996 </blockquote>
24997
24998 <p>
24999 Add to p2 25.2.6 [alg.find.end]:
25000 </p>
25001 <blockquote>
25002 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
25003 empty.  &#8212; <i>end note</i>]
25004 </blockquote>
25005
25006 <p>
25007 Add to p2 25.2.7 [alg.find.first.of]
25008 </p>
25009 <blockquote>
25010 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
25011 empty.  &#8212; <i>end note</i>]
25012 </blockquote>
25013
25014 <p>
25015 Add to p2 25.2.12 [alg.search]:
25016 </p>
25017 <blockquote>
25018 [<i>Note:</i> Returns <tt>last1</tt> if <tt>[first2,last2)</tt> is
25019 empty.  &#8212; <i>end note</i>]
25020 </blockquote>
25021
25022 <p>
25023 Add to p1 25.3.13 [alg.partitions]:
25024 </p>
25025 <blockquote>
25026 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first,last)</tt> is empty.
25027 &#8212; <i>end note</i>]
25028 </blockquote>
25029
25030 <p>
25031 Add to p1 25.4.5.1 [includes]:
25032 </p>
25033 <blockquote>
25034 [<i>Note:</i> Returns <tt>true</tt> if <tt>[first2,last2)</tt> is empty.
25035 &#8212; <i>end note</i>]
25036 </blockquote>
25037
25038 <p>
25039 Revise p2 25.4.6.2 [pop.heap]
25040 </p>
25041 <blockquote>
25042 Requires: The range <tt>[first,last)</tt> shall be a valid
25043 <ins>non-empty</ins> heap.
25044 </blockquote>
25045
25046 <p>
25047 [Editorial]
25048 </p>
25049 <blockquote>
25050 Reverse order of 25.4.6.2 [pop.heap] p1 and p2.
25051 </blockquote>
25052
25053 <p>
25054 Add to p31 25.4.7 [alg.min.max]:
25055 </p>
25056 <blockquote>
25057 Returns <tt>make_pair(first, first)</tt> if <tt>first == last</tt>.
25058 </blockquote>
25059
25060 <p>
25061 Add to p1 25.4.8 [alg.lex.comparison]:
25062 </p>
25063 <blockquote>
25064 [<i>Note:</i> An empty sequence is lexicographically less than any other
25065 non-empty sequence, but not less than another empty sequence. &#8212;
25066 <i>end note</i>]
25067 </blockquote>
25068
25069
25070
25071
25072
25073
25074 <hr>
25075 <h3><a name="1206"></a>1206. Incorrect requires for <tt>move_backward</tt> and <tt>copy_backward</tt></h3>
25076 <p><b>Section:</b> 25.3.2 [alg.move] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25077  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-09-13  <b>Last modified:</b> 2009-09-13</p>
25078 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25079 <p><b>Discussion:</b></p>
25080 <p>
25081 25.3.2 [alg.move], p6 says:
25082 </p>
25083
25084 <blockquote>
25085 <pre>template&lt;class BidirectionalIterator1, class BidirectionalIterator2&gt;
25086   BidirectionalIterator2
25087     move_backward(BidirectionalIterator1 first,
25088                   BidirectionalIterator1 last,
25089                   BidirectionalIterator2 result);
25090 </pre>
25091 <blockquote>
25092 <p>...</p>
25093 <p>
25094 <i>Requires:</i> <tt>result</tt> shall not be in the range
25095 <tt>[first,last)</tt>.
25096 </p>
25097 </blockquote>
25098 </blockquote>
25099
25100 <p>
25101 This is essentially an "off-by-one" error.
25102 </p>
25103
25104 <p>
25105 When <tt>result == last</tt>, which
25106 <em>is</em> allowed by this specification, then the range <tt>[first, last)</tt>
25107 is being move assigned into the range <tt>[first, last)</tt>.  The <tt>move</tt>
25108 (forward) algorithm doesn't allow self move assignment, and neither should
25109 <tt>move_backward</tt>.  So <tt>last</tt> should be included in the range which
25110 <tt>result</tt> can not be in.
25111 </p>
25112
25113 <p>
25114 Conversely, when <tt>result == first</tt>, which <em>is not</em> allowed by this
25115 specification, then the range <tt>[first, last)</tt>
25116 is being move assigned into the range <tt>[first - (last-first), first)</tt>.
25117 I.e. into a <em>non-overlapping</em> range.  Therefore <tt>first</tt> should
25118 not be included in the range which <tt>result</tt> can not be in.
25119 </p>
25120
25121 <p>
25122 The same argument applies to <tt>copy_backward</tt> though copy assigning elements
25123 to themselves (<tt>result == last</tt>) should be harmless (though is disallowed
25124 by <tt>copy</tt>).
25125 </p>
25126
25127
25128
25129 <p><b>Proposed resolution:</b></p>
25130 <p>
25131 Change 25.3.2 [alg.move], p6:
25132 </p>
25133
25134 <blockquote>
25135 <pre>template&lt;class BidirectionalIterator1, class BidirectionalIterator2&gt;
25136   BidirectionalIterator2
25137     move_backward(BidirectionalIterator1 first,
25138                   BidirectionalIterator1 last,
25139                   BidirectionalIterator2 result);
25140 </pre>
25141 <blockquote>
25142 <p>...</p>
25143 <p>
25144 <i>Requires:</i> <tt>result</tt> shall not be in the range
25145 <tt><del>[</del><ins>(</ins>first,last<ins>]</ins><del>)</del></tt>.
25146 </p>
25147 </blockquote>
25148 </blockquote>
25149
25150 <p>
25151 Change 25.3.1 [alg.copy], p13:
25152 </p>
25153
25154 <blockquote>
25155 <pre>template&lt;class BidirectionalIterator1, class BidirectionalIterator2&gt;
25156   BidirectionalIterator2
25157     copy_backward(BidirectionalIterator1 first,
25158                   BidirectionalIterator1 last,
25159                   BidirectionalIterator2 result);
25160 </pre>
25161 <blockquote>
25162 <p>...</p>
25163 <p>
25164 <i>Requires:</i> <tt>result</tt> shall not be in the range
25165 <tt><del>[</del><ins>(</ins>first,last<ins>]</ins><del>)</del></tt>.
25166 </p>
25167 </blockquote>
25168 </blockquote>
25169
25170
25171
25172
25173
25174 <hr>
25175 <h3><a name="1207"></a>1207. Underspecified std::list operations?</h3>
25176 <p><b>Section:</b> 23.3.4.4 [list.ops] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25177  <b>Submitter:</b> Loïc Joly <b>Opened:</b> 2009-09-13  <b>Last modified:</b> 2009-09-19</p>
25178 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#list.ops">active issues</a> in [list.ops].</p>
25179 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#list.ops">issues</a> in [list.ops].</p>
25180 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25181 <p><b>Discussion:</b></p>
25182 <p>
25183 It looks to me like some operations of <tt>std::list</tt>
25184 (<tt>sort</tt>, <tt>reverse</tt>, <tt>remove</tt>, <tt>unique</tt> &amp;
25185 <tt>merge</tt>) do not specify the validity of iterators, pointers &amp;
25186 references to elements of the list after those operations. Is it implied
25187 by some other text in the standard?
25188 </p>
25189
25190 <p>
25191 I believe <tt>sort</tt> &amp; <tt>reverse</tt> do not invalidating
25192 anything, <tt>remove</tt> &amp; <tt>unique</tt> only invalidates what
25193 refers to erased elements, <tt>merge</tt> does not invalidate anything
25194 (with the same precision as <tt>splice</tt> for elements who changed of
25195 container). Are those assumptions correct ?
25196 </p>
25197
25198
25199 <p><b>Proposed resolution:</b></p>
25200
25201
25202
25203
25204
25205 <hr>
25206 <h3><a name="1208"></a>1208. valarray initializer_list constructor has incorrect effects</h3>
25207 <p><b>Section:</b> 26.6.2.1 [valarray.cons] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
25208  <b>Submitter:</b> Howard Hinnant <b>Opened:</b> 2009-09-23  <b>Last modified:</b> 2009-10-29</p>
25209 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#valarray.cons">issues</a> in [valarray.cons].</p>
25210 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
25211 <p><b>Discussion:</b></p>
25212 <p>
25213 26.6.2.1 [valarray.cons] says:
25214 </p>
25215
25216 <blockquote>
25217 <pre>valarray(initializer_list&lt;T&gt; il);
25218 </pre>
25219 <blockquote>
25220 <i>Effects:</i> Same as <tt>valarray(il.begin(), il.end())</tt>.
25221 </blockquote>
25222 </blockquote>
25223
25224 <p>
25225 But there is no <tt>valarray</tt> constructor taking two <tt>const T*</tt>.
25226 </p>
25227
25228 <p><i>[
25229 2009-10-29 Howard:
25230 ]</i></p>
25231
25232
25233 <blockquote>
25234 Moved to Tentatively Ready after 6 positive votes on c++std-lib.
25235 </blockquote>
25236
25237
25238 <p><b>Proposed resolution:</b></p>
25239 <p>
25240 Change 26.6.2.1 [valarray.cons]:
25241 </p>
25242
25243 <blockquote>
25244 <pre>valarray(initializer_list&lt;T&gt; il);
25245 </pre>
25246 <blockquote>
25247 <i>Effects:</i> Same as <tt>valarray(il.begin(), il.<del>end</del><ins>size</ins>())</tt>.
25248 </blockquote>
25249 </blockquote>
25250
25251
25252
25253
25254
25255 <hr>
25256 <h3><a name="1209"></a>1209. match_results should be moveable</h3>
25257 <p><b>Section:</b> 28.10.1 [re.results.const] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25258  <b>Submitter:</b> Stephan T. Lavavej <b>Opened:</b> 2009-09-15  <b>Last modified:</b> 2009-09-21</p>
25259 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25260 <p><b>Discussion:</b></p>
25261 <p>
25262 In Working Draft
25263 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>,
25264 <tt>match_results</tt> lacks a move constructor and move
25265 assignment operator. Because it owns dynamically allocated memory, it
25266 should be moveable.
25267 </p>
25268
25269 <p>
25270 As far as I can tell, this isn't tracked by an active issue yet; Library
25271 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#723">723</a> doesn't talk about <tt>match_results</tt>.
25272 </p>
25273
25274 <p><i>[
25275 2009-09-21 Daniel provided wording.
25276 ]</i></p>
25277
25278
25279
25280 <p><b>Proposed resolution:</b></p>
25281 <ol>
25282 <li>
25283 <p>
25284 Add the following member declarations to 28.10 [re.results]/3:
25285 </p>
25286
25287 <blockquote><pre>// 28.10.1, construct/copy/destroy:
25288 explicit match_results(const Allocator&amp; a = Allocator());
25289 match_results(const match_results&amp; m);
25290 <ins>match_results(match_results&amp;&amp; m);</ins>
25291 match_results&amp; operator=(const match_results&amp; m);
25292 <ins>match_results&amp; operator=(match_results&amp;&amp; m);</ins>
25293 ~match_results();
25294 </pre></blockquote>
25295 </li>
25296
25297 <li>
25298 <p>
25299 Add the following new prototype descriptions to 28.10.1 [re.results.const]
25300 using the table numbering of
25301 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>:
25302 </p>
25303
25304 <blockquote>
25305 <pre>match_results(const match_results&amp; m);
25306 </pre>
25307
25308 <blockquote>
25309 4 <i>Effects:</i> Constructs an object of class <tt>match_results</tt>, as a
25310 copy of <tt>m</tt>.
25311 </blockquote>
25312
25313 <pre><ins>match_results(match_results&amp;&amp; m);</ins>
25314 </pre>
25315
25316 <blockquote>
25317 <p>
25318 <ins>5 <i>Effects:</i> Move-constructs an object of class <tt>match_results</tt>
25319 from <tt>m</tt> satisfying the same postconditions as Table 132. Additionally
25320 the stored <tt>Allocator</tt> value is move constructed from <tt>m.get_allocator()</tt>.
25321 After the initialization of <tt>*this</tt> sets <tt>m</tt> to an unspecified but valid
25322 state.</ins>
25323 </p>
25324
25325 <p>
25326 <ins>6 <i>Throws:</i> Nothing if the allocator's move constructor throws nothing.</ins>
25327 </p>
25328 </blockquote>
25329
25330 <pre>match_results&amp; operator=(const match_results&amp; m);
25331 </pre>
25332
25333 <blockquote>
25334 7 <i>Effects:</i> Assigns <tt>m</tt> to <tt>*this</tt>. The postconditions of this function are
25335 indicated in Table 132.
25336 </blockquote>
25337
25338 <pre><ins>match_results&amp; operator=(match_results&amp;&amp; m);</ins>
25339 </pre>
25340
25341 <blockquote>
25342 <p>
25343 <ins>8 <i>Effects:</i> Move-assigns <tt>m</tt> to <tt>*this</tt>. The postconditions of this
25344 function are indicated in Table 132. After the assignment, <tt>m</tt> is in
25345 a valid but unspecified state.</ins>
25346 </p>
25347
25348 <p>
25349 <ins>9 <i>Throws:</i> Nothing.</ins>
25350 </p>
25351 </blockquote>
25352 </blockquote>
25353 </li>
25354
25355 </ol>
25356
25357
25358
25359
25360
25361 <hr>
25362 <h3><a name="1210"></a>1210. iterator reachability should not require a container</h3>
25363 <p><b>Section:</b> 24.2 [iterator.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25364  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-18  <b>Last modified:</b> 2009-09-19</p>
25365 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
25366 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
25367 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25368 <p><b>Discussion:</b></p>
25369 <p>
25370 p6 Iterator requirements 24.2 [iterator.requirements]
25371 </p>
25372
25373 <blockquote>
25374 An iterator <tt>j</tt> is called reachable from an iterator <tt>i</tt> if and only if
25375 there is a finite sequence of applications of the expression <tt>++i</tt> that
25376 makes <tt>i == j</tt>. If <tt>j</tt> is reachable from <tt>i</tt>, they refer to the same
25377 container.
25378 </blockquote>
25379
25380 <p>
25381 A good example would be stream iterators, which do not refer to a
25382 container.  Typically, the end iterator from a range of stream iterators
25383 will compare equal for many such ranges.  I suggest striking the second
25384 sentence.
25385 </p>
25386
25387 <p>
25388 An alternative wording might be:
25389 </p>
25390
25391 <blockquote>
25392 If <tt>j</tt> is reachable from <tt>i</tt>, and both <tt>i</tt> and
25393 <tt>j</tt> are dereferencable iterators, then they refer to the same
25394 range.
25395 </blockquote>
25396
25397
25398 <p><b>Proposed resolution:</b></p>
25399 <p>
25400 Change 24.2 [iterator.requirements], p6:
25401 </p>
25402
25403 <blockquote>
25404 An iterator <tt>j</tt> is called <i>reachable</i> from an iterator
25405 <tt>i</tt> if and only if there is a finite sequence of applications of
25406 the expression <tt>++i</tt> that makes <tt>i == j</tt>. <del>If
25407 <tt>j</tt> is reachable from <tt>i</tt>, they refer to the same
25408 container.</del>
25409 </blockquote>
25410
25411
25412
25413
25414
25415 <hr>
25416 <h3><a name="1211"></a>1211. move iterators should be restricted as input iterators</h3>
25417 <p><b>Section:</b> 24.5.3.1 [move.iterator] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
25418  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-18  <b>Last modified:</b> 2009-10-26</p>
25419 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#move.iterator">issues</a> in [move.iterator].</p>
25420 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
25421 <p><b>Discussion:</b></p>
25422 <p>
25423 I contend that while we can support both bidirectional and random access
25424 traversal, the category of a move iterator should never be better than
25425 <tt>input_iterator_tag</tt>.
25426 </p>
25427
25428 <p>
25429 The contentious point is that you cannot truly have a multipass property
25430 when values are moved from a range.  This is contentious if you view a
25431 moved-from object as still holding a valid value within the range.
25432 </p>
25433
25434 <p>
25435 The second reason comes from the Forward Iterator requirements table:
25436 </p>
25437
25438 <blockquote>
25439 <p>
25440 Forward iterators 24.2.3 [forward.iterators]
25441 </p>
25442
25443 <p>
25444 Table 102 -- Forward iterator requirements
25445 </p>
25446
25447 <blockquote>
25448 For expression <tt>*a</tt> the return type is:
25449 "<tt>T&amp;</tt> if <tt>X</tt> is mutable, otherwise <tt>const T&amp;</tt>"
25450 </blockquote>
25451 </blockquote>
25452
25453 <p>
25454 There is a similar constraint on <tt>a-&gt;m</tt>.
25455 </p>
25456
25457 <p>
25458 There is no support for rvalue references, nor do I believe their should
25459 be.  Again, opinions may vary but either this table or the definition of
25460 <tt>move_iterator</tt> need updating.
25461 </p>
25462
25463 <p>
25464 Note: this requirement probably need updating anyway if we wish to
25465 support proxy iterators but I am waiting to see a new working paper
25466 before filing that issue.
25467 </p>
25468
25469 <p><i>[
25470 2009-10 post-Santa Cruz:
25471 ]</i></p>
25472
25473
25474 <blockquote>
25475 Move to Open. Howard to put his rationale mentioned above into the issue
25476 as a note.
25477 </blockquote>
25478
25479 <p><i>[
25480 2009-10-26 Howard adds:
25481 ]</i></p>
25482
25483
25484 <blockquote>
25485 <p>
25486 <tt>vector::insert(pos, iter, iter)</tt> is significantly more effcient when
25487 <tt>iter</tt> is a random access iterator, as compared to when it is an
25488 input iterator.
25489 </p>
25490
25491 <p>
25492 When <tt>iter</tt> is an input iterator, the best algorithm
25493 is to append the inserted range to the end of the <tt>vector</tt> using
25494 <tt>push_back</tt>.  This may involve several reallocations before the input
25495 range is exhausted.  After the append, then one can use <tt>std::rotate</tt>
25496 to place the inserted range into the correct position in the vector.
25497 </p>
25498
25499 <p>
25500 But when <tt>iter</tt> is a random access iterator, the best algorithm
25501 is to first compute the size of the range to be inserted (<tt>last - first</tt>),
25502 do a buffer reallocation if necessary, scoot existing elements in the <tt>vector</tt>
25503 down to make the "hole", and then insert the new elements directly to their correct
25504 place.
25505 </p>
25506
25507 <blockquote><b>
25508 The insert-with-random-access-iterators algorithm is considerably more efficient
25509 than the insert-with-input-iterators algorithm
25510 </b></blockquote>
25511
25512 <p>
25513 Now consider:
25514 </p>
25515
25516 <blockquote><pre>vector&lt;A&gt; v;
25517 <font color="#c80000">//  ... build up a large vector of A ...</font>
25518 vector&lt;A&gt; temp;
25519 <font color="#c80000">//  ... build up a large temporary vector of A to later be inserted ...</font>
25520 typedef move_iterator&lt;vector&lt;A&gt;::iterator&gt; MI;
25521 <font color="#c80000">//  Now insert the temporary elements:</font>
25522 v.insert(v.begin() + N, MI(temp.begin()), MI(temp.end()));
25523 </pre></blockquote>
25524
25525 <p>
25526 A major motivation for using <tt>move_iterator</tt> in the above example is the
25527 expectation that <tt>A</tt> is cheap to move but expensive to copy.  I.e. the
25528 customer is looking for <em>high performance</em>.  If we allow <tt>vector::insert</tt>
25529 to subtract two <tt>MI</tt>'s to get the distance between them, the customer enjoys
25530 substantially better performance, compared to if we say that <tt>vector::insert</tt>
25531 can not subtract two <tt>MI</tt>'s.
25532 </p>
25533
25534 <p>
25535 I can find no rationale for not giving this performance boost to our customers.
25536 Therefore I am strongly against restricting <tt>move_iterator</tt> to the
25537 <tt>input_iterator_tag</tt> category.
25538 </p>
25539
25540 <p>
25541 I believe that the requirement that forward
25542 iterators have a dereference that returns an lvalue reference to cause unacceptable
25543 pessimization.  For example <tt>vector&lt;bool&gt;::iterator</tt> also does not return
25544 a <tt>bool&amp;</tt> on dereference.  Yet I am not aware of a single vendor that
25545 is willing to ship <tt>vector&lt;bool&gt;::iterator</tt> as an input iterator.
25546 Everyone classifies it as a random access iterator.  Not only does this not
25547 cause any problems, it prevents significant performance problems.
25548 </p>
25549
25550 </blockquote>
25551
25552
25553
25554 <p><b>Proposed resolution:</b></p>
25555 <p>
25556 Class template move_iterator 24.5.3.1 [move.iterator]
25557 </p>
25558
25559 <blockquote><pre>namespace std {
25560 template &lt;class Iterator&gt;
25561 class move_iterator {
25562 public:
25563  ...
25564  typedef <del>typename iterator_traits&lt;Iterator&gt;::iterator_category</del> <ins>input_iterator_tag</ins> iterator_category;
25565 </pre></blockquote>
25566
25567
25568
25569
25570
25571 <hr>
25572 <h3><a name="1212"></a>1212. result of post-increment/decrement operator</h3>
25573 <p><b>Section:</b> 24.2 [iterator.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25574  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-09-18  <b>Last modified:</b> 2009-09-19</p>
25575 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
25576 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
25577 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25578 <p><b>Discussion:</b></p>
25579 <p>Forward iterator and bidirectional iterator place different
25580 requirements on the result of post-increment/decrement operator. The
25581 same form should be used in each case.
25582 </p>
25583
25584 <p>
25585 Merging row from:
25586 </p>
25587
25588 <blockquote><pre>Table 102 -- Forward iterator requirements
25589 Table 103 -- Bidirectional iterator requirements
25590
25591     r++ : convertible to const X&amp;
25592     r-- : convertible to const X&amp;
25593
25594     *r++ : T&amp; if X is mutable, otherwise const T&amp;
25595     *r-- : convertible to T
25596 </pre></blockquote>
25597
25598
25599 <p><b>Proposed resolution:</b></p>
25600
25601
25602
25603
25604
25605 <hr>
25606 <h3><a name="1213"></a>1213. Meaning of valid and singular iterator underspecified</h3>
25607 <p><b>Section:</b> 24.2 [iterator.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25608  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-09-19  <b>Last modified:</b> 2009-09-19</p>
25609 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#iterator.requirements">active issues</a> in [iterator.requirements].</p>
25610 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iterator.requirements">issues</a> in [iterator.requirements].</p>
25611 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25612 <p><b>Discussion:</b></p>
25613 <p>
25614 The terms <em>valid</em> iterator and <em>singular</em> aren't
25615 properly defined. The fuzziness of those terms became even worse
25616 after the resolution of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#208">208</a> (including further updates by <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#278">278</a>). In
25617 24.2 [iterator.requirements] as of
25618 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2723.pdf">N2723</a>
25619 the standard says now:
25620 </p>
25621
25622 <blockquote>
25623 <p>
25624 5 - These values are called past-the-end values. Values of an iterator <tt>i</tt> for
25625 which the expression <tt>*i</tt> is defined are called dereferenceable. The library
25626 never assumes that past-the-end values are dereferenceable. Iterators
25627 can also have singular values that are not associated with any
25628 container. [...] Results of most expressions are undefined for singular
25629 values; the only exceptions are destroying an iterator that holds a
25630 singular value and the assignment of a non-singular value to an iterator
25631 that holds a singular value. [...] Dereferenceable values are always
25632 non-singular.
25633 </p>
25634
25635 <p>
25636 10 - An invalid iterator is an iterator that may be singular.
25637 </p>
25638 </blockquote>
25639
25640 <p>
25641 First, issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#208">208</a> intentionally removed the earlier constraint that past-the-end
25642 values are always non-singular. The reason for this was to support null
25643 pointers as past-the-end iterators of e.g. empty sequences. But there
25644 seem to exist different views on what a singular (iterator) value is. E.g.
25645 according to the <a href="http://www.sgi.com/tech/stl/trivial.html">SGI definition</a>
25646 a null pointer is <em>not</em> a singular value:
25647 </p>
25648
25649 <blockquote>
25650 Dereferenceable iterators are always nonsingular, but the converse is
25651 not true.
25652 For example, a null pointer is nonsingular (there are well defined operations
25653 involving null pointers) even thought it is not dereferenceable.
25654 </blockquote>
25655
25656 <p>
25657 and <a href="http://www.sgi.com/tech/stl/InputIterator.html">proceeds</a>:
25658 </p>
25659
25660 <blockquote>
25661 An iterator is valid if it is dereferenceable or past-the-end.
25662 </blockquote>
25663
25664 <p>
25665 Even if the standard prefers a different meaning of singular here, the
25666 change was
25667 incomplete, because by restricting feasible expressions of singular
25668 iterators to
25669 destruction and assignment isn't sufficient for a past-the-end
25670 iterator: Of-course
25671 it must still be equality-comparable and in general be a readable value.
25672 </p>
25673
25674 <p>
25675 Second, the standard doesn't clearly say whether a past-the-end value is
25676 a valid iterator or not. E.g. 20.8.13 [specialized.algorithms]/1 says:
25677 </p>
25678
25679 <blockquote>
25680 In all of the following algorithms, the formal template parameter
25681 <tt>ForwardIterator</tt>
25682 is required to satisfy the requirements of a forward iterator (24.1.3)
25683 [..], and is
25684 required to have the property that no exceptions are thrown from [..], or
25685 dereference of valid iterators.
25686 </blockquote>
25687
25688 <p>
25689 The standard should make better clear what "singular pointer" and "valid
25690 iterator" means. The fact that the meaning of a valid <em>value</em>
25691 has a core language meaning doesn't imply that for an iterator concept
25692 the term "valid iterator" has the same meaning.
25693 </p>
25694
25695 <p>
25696 Let me add a final example: In X [allocator.concepts.members] of
25697 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2914.pdf">N2914</a>
25698 we find:
25699 </p>
25700
25701 <blockquote><pre>pointer X::allocate(size_type n);
25702 </pre>
25703
25704 <blockquote>
25705 11 <i>Returns:</i> a pointer to the allocated memory. [<i>Note:</i> if <tt>n == 0</tt>, the return
25706 value is unspecified. &#8212;<i>end note</i>]
25707 </blockquote>
25708
25709 <p>
25710 [..]
25711 </p>
25712
25713 <pre>void X::deallocate(pointer p, size_type n);
25714 </pre>
25715
25716 <blockquote>
25717 <i>Preconditions:</i> <tt>p</tt> shall be a non-singular pointer value obtained from a call
25718 to <tt>allocate()</tt> on this allocator or one that compares equal to it.
25719 </blockquote>
25720 </blockquote>
25721
25722 <p>
25723 If singular pointer value would include null pointers this make the
25724 preconditions
25725 unclear if the pointer value is a result of <tt>allocate(0)</tt>: Since the return value
25726 is unspecified, it could be a null pointer. Does that mean that programmers
25727 need to check the pointer value for a null value before calling deallocate?
25728 </p>
25729
25730
25731 <p><b>Proposed resolution:</b></p>
25732
25733
25734
25735
25736
25737 <hr>
25738 <h3><a name="1214"></a>1214. Insufficient/inconsistent key immutability requirements for  associative containers</h3>
25739 <p><b>Section:</b> 23.2.4 [associative.reqmts] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25740  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-09-20  <b>Last modified:</b> 2009-09-20</p>
25741 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#associative.reqmts">active issues</a> in [associative.reqmts].</p>
25742 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#associative.reqmts">issues</a> in [associative.reqmts].</p>
25743 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25744 <p><b>Discussion:</b></p>
25745 <p>
25746 Scott Meyers' mentions on a recent posting on <a href="http://groups.google.de/group/comp.std.c++/msg/6f9160fc428bcbea">c.s.c++</a>
25747 some arguments that point to an incomplete resolution
25748 of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a> and to an inconsistency of requirements on keys in ordered and
25749 unordered associative
25750 containers:
25751 </p>
25752
25753 <blockquote>
25754 <p>
25755 1) <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a> introduced the term immutable without defining it in a unique manner in
25756 23.2.4 [associative.reqmts]/5:
25757 </p>
25758
25759 <blockquote>
25760 [..] Keys in an associative container are immutable.
25761 </blockquote>
25762
25763 <p>
25764 According to conventional dictionaries immutable is an unconditional way of
25765 saying that something cannot be changed. So without any further explicit
25766 allowance a user <em>always</em> runs into undefined behavior if (s)he attempts
25767 to modify such a key. IMO this was not the intend of the committee to resolve
25768 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a> in that way because the comments suggest an interpretation that
25769 should give any user the freedom to modify the key in an <em>explicit</em> way
25770 <em>provided</em> it would not affect the sort order in that container.
25771 </p>
25772
25773 <p>
25774 2) Another observation was that surprisingly no similar 'safety guards'
25775 exists against unintentional key changes for the unordered associative
25776 containers, specifically there is no such requirement as in
25777 23.2.4 [associative.reqmts]/6 that "both <tt>iterator</tt> and <tt>const_iterator</tt> are constant
25778 iterators". But the need for such protection against unintentional
25779 changes as well as the constraints in which manner any explicit
25780 changes may be performed are both missing and necessary, because
25781 such changes could potentially change the <em>equivalence</em> of keys that
25782 is measured by the <tt>hasher</tt> and <tt>key_equal</tt>.
25783 </p>
25784
25785 <p>
25786 I suggest to fix the unconditional wording involved with "immutable keys"
25787 by at least adding a hint for the reader that users <em>may</em> perform such
25788 changes in an explicit manner <em>and</em> to perform similar wording changes
25789 as <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#103">103</a> did for the ordered associative containers also for the unordered
25790 containers.
25791 </p>
25792 </blockquote>
25793
25794
25795 <p><b>Proposed resolution:</b></p>
25796
25797
25798
25799
25800
25801 <hr>
25802 <h3><a name="1215"></a>1215. <tt>list::merge</tt> with unequal allocators</h3>
25803 <p><b>Section:</b> 23.3.4.4 [list.ops] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25804  <b>Submitter:</b> Pablo Halpern <b>Opened:</b> 2009-09-24  <b>Last modified:</b> 2009-09-24</p>
25805 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#list.ops">active issues</a> in [list.ops].</p>
25806 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#list.ops">issues</a> in [list.ops].</p>
25807 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25808 <p><b>Discussion:</b></p>
25809 <p>
25810 In Bellevue (I think), we passed
25811 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2525.pdf">N2525</a>,
25812 which, among other things, specifies that the behavior of
25813 <tt>list::splice</tt> is undefined if the allocators of the two lists
25814 being spliced do not compare equal. The same rationale should apply to
25815 <tt>list::merge</tt>. The intent of <tt>list::merge</tt> (AFAIK) is to
25816 move nodes from one sorted <tt>list</tt> into another sorted
25817 <tt>list</tt> without copying the elements. This is possible only if the
25818 allocators compare equal.
25819 </p>
25820
25821
25822 <p><b>Proposed resolution:</b></p>
25823 <p>
25824 Relative to the August 2009 WP,
25825 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2857.pdf">N2857</a>,
25826 change 23.3.4.4 [list.ops],
25827 paragraphs 22-25 as follows:
25828 </p>
25829
25830 <blockquote>
25831 <pre>void merge(list&amp;&amp; x);
25832 template &lt;class Compare&gt; void merge(list&amp;&amp; x, Compare comp);
25833 </pre>
25834 <blockquote>
25835 <p>
25836 <i>Requires</i>: both the list and the argument list shall be sorted
25837 according to operator&lt; or comp.
25838 </p>
25839 <p>
25840 <i>Effects</i>: If <tt>(&amp;x == this)</tt> does nothing; otherwise, merges the
25841 two sorted ranges <tt>[begin(), end())</tt> and <tt>[x.begin(),
25842 x.end())</tt>. The result is a range in which the elements will be
25843 sorted in non-decreasing order according to the ordering defined by
25844 <tt>comp</tt>; that is, for every iterator <tt>i</tt>, in the range other than the
25845 <tt>first</tt>, the condition <tt>comp(*i, *(i - 1)<ins>)</ins></tt> will be
25846 <tt>false</tt>.
25847 </p>
25848 <p>
25849 <i>Remarks</i>: Stable. If <tt>(&amp;x != this)</tt> the range <tt>[x.begin(), x.end())</tt> is
25850 empty after the merge. <ins>No elements are copied by this operation.
25851 The behavior is undefined if <tt>this-&gt;get_allocator() !=
25852 x.get_allocator()</tt>.</ins>
25853 </p>
25854 <p>
25855 <i>Complexity</i>: At most <tt>size() + x.size() - 1</tt> applications of <tt>comp</tt>
25856 if <tt>(&amp;x != this)</tt>; otherwise, no applications of <tt>comp</tt> are performed. If an
25857 exception is thrown other than by a comparison there are no effects.
25858 </p>
25859 </blockquote>
25860 </blockquote>
25861
25862
25863
25864
25865
25866 <hr>
25867 <h3><a name="1216"></a>1216. LWG 1066 Incomplete?</h3>
25868 <p><b>Section:</b> 18.8.6 [except.nested] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
25869  <b>Submitter:</b> Pete Becker <b>Opened:</b> 2009-09-25  <b>Last modified:</b> 2009-10-20</p>
25870 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#except.nested">active issues</a> in [except.nested].</p>
25871 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#except.nested">issues</a> in [except.nested].</p>
25872 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
25873 <p><b>Discussion:</b></p>
25874 <p>
25875 LWG <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1066">1066</a> adds <tt>[[noreturn]]</tt> to a bunch of things.
25876 It doesn't add it to <tt>rethrow_nested()</tt>, which seems like an obvious
25877 candidate. I've made the changes indicated in the issue, and haven't
25878 changed <tt>rethrow_nested()</tt>.
25879 </p>
25880
25881 <p><i>[
25882 2009 Santa Cruz:
25883 ]</i></p>
25884
25885
25886 <blockquote>
25887 Move to Ready.
25888 </blockquote>
25889
25890
25891
25892 <p><b>Proposed resolution:</b></p>
25893 <p>
25894 Add <tt>[[noreturn]]</tt> to <tt>rethrow_nested()</tt> in 18.8.6 [except.nested].
25895 </p>
25896
25897
25898
25899
25900
25901 <hr>
25902 <h3><a name="1218"></a>1218. mutex destructor synchronization</h3>
25903 <p><b>Section:</b> 30.4.1 [thread.mutex.requirements] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25904  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
25905 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.mutex.requirements">active issues</a> in [thread.mutex.requirements].</p>
25906 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.mutex.requirements">issues</a> in [thread.mutex.requirements].</p>
25907 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25908 <p><b>Discussion:</b></p>
25909 <p>
25910 If an object <tt>*o</tt> contains a mutex <tt>mu</tt> and a
25911 correctly-maintained reference count <tt>c</tt>, is the following code
25912 safe?
25913 </p>
25914
25915 <blockquote><pre>o-&gt;mu.lock();
25916 bool del = (--(o-&gt;c) == 0);
25917 o-&gt;mu.unlock();
25918 if (del) { delete o; }
25919 </pre></blockquote>
25920
25921 <p>
25922 If the implementation of <tt>mutex::unlock()</tt> can touch the mutex's
25923 memory after the moment it becomes free, this wouldn't be safe, and
25924 "Construction and destruction of an object of a Mutex type need not be
25925 thread-safe" 30.4.1 [thread.mutex.requirements] may imply that
25926 it's not safe. Still, it's useful to allow mutexes to guard reference
25927 counts, and if it's not allowed, users are likely to write bugs.
25928 </p>
25929
25930
25931 <p><b>Proposed resolution:</b></p>
25932 <p>
25933 </p>
25934
25935
25936
25937
25938
25939 <hr>
25940 <h3><a name="1219"></a>1219. unique_lock::lock and resource_deadlock_would_occur</h3>
25941 <p><b>Section:</b> 30.4.3.2.2 [thread.lock.unique.locking] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
25942  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
25943 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.lock.unique.locking">active issues</a> in [thread.lock.unique.locking].</p>
25944 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.lock.unique.locking">issues</a> in [thread.lock.unique.locking].</p>
25945 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
25946 <p><b>Discussion:</b></p>
25947 <p>
25948 <tt>unique_lock::lock</tt> and friends raise
25949 "<tt>resource_deadlock_would_occur</tt> -- if the current thread already
25950 owns the mutex (i.e., on entry, <tt>owns</tt> is <tt>true</tt>)."  1)
25951 The current thread owning a mutex is not the same as any particular
25952 <tt>unique_lock::owns</tt> being <tt>true</tt>. 2) There's no need to
25953 raise this exception for a <tt>recursive_mutex</tt> if <tt>owns</tt> is
25954 <tt>false</tt>. 3) If <tt>owns</tt> is true, we need to raise some
25955 exception or the unique_lock will lose track of whether to unlock itself
25956 on destruction, but "deadlock" isn't it. For (3), s/bool owns/int
25957 ownership_level/ would fix it.
25958 </p>
25959
25960
25961 <p><b>Proposed resolution:</b></p>
25962
25963
25964
25965
25966
25967 <hr>
25968 <h3><a name="1220"></a>1220. What does condition_variable wait on?</h3>
25969 <p><b>Section:</b> 30.5 [thread.condition] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
25970  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-11-06</p>
25971 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition">active issues</a> in [thread.condition].</p>
25972 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition">issues</a> in [thread.condition].</p>
25973 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
25974 <p><b>Discussion:</b></p>
25975 <p>
25976 "Class <tt>condition_variable</tt> provides a condition variable that can only
25977 wait on an object of type <tt>unique_lock</tt>" should say "...object of type
25978 <tt>unique_lock&lt;mutex&gt;</tt>"
25979 </p>
25980
25981 <p><i>[
25982 2009-11-06 Howard adds:
25983 ]</i></p>
25984
25985
25986 <blockquote>
25987 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
25988 </blockquote>
25989
25990
25991
25992 <p><b>Proposed resolution:</b></p>
25993 <p>
25994 Change 30.5 [thread.condition], p1:
25995 </p>
25996
25997 <blockquote>
25998 Condition variables provide synchronization primitives used to block a
25999 thread until notified by some other thread that some condition is met or
26000 until a system time is reached. Class <tt>condition_variable</tt>
26001 provides a condition variable that can only wait on an object of type
26002 <tt>unique_lock<ins>&lt;mutex&gt;</ins></tt>, allowing maximum
26003 efficiency on some platforms. Class <tt>condition_variable_any</tt>
26004 provides a general condition variable that can wait on objects of
26005 user-supplied lock types.
26006 </blockquote>
26007
26008
26009
26010
26011
26012 <hr>
26013 <h3><a name="1221"></a>1221. condition_variable wording</h3>
26014 <p><b>Section:</b> 30.5.1 [thread.condition.condvar] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26015  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
26016 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvar">active issues</a> in [thread.condition.condvar].</p>
26017 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvar">issues</a> in [thread.condition.condvar].</p>
26018 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26019 <p><b>Discussion:</b></p>
26020 <p>
26021 30.5.1 [thread.condition.condvar] says:
26022 </p>
26023
26024 <blockquote>
26025 <pre>~condition_variable();
26026 </pre>
26027 <blockquote>
26028 <i>Precondition:</i> There shall be no thread blocked on <tt>*this</tt>.
26029 [<i>Note:</i> That is, all threads shall have been notified; they may
26030 subsequently block on the lock specified in the wait. Beware that
26031 destroying a <tt>condition_variable</tt> object while the corresponding
26032 predicate is <tt>false</tt> is likely to lead to undefined behavior.
26033 &#8212; <i>end note</i>]
26034 </blockquote>
26035 </blockquote>
26036
26037 <p>
26038 The text hasn't introduced the notion of a "corresponding predicate"
26039 yet.
26040 </p>
26041
26042
26043 <p><b>Proposed resolution:</b></p>
26044
26045
26046
26047
26048
26049 <hr>
26050 <h3><a name="1222"></a>1222. condition_variable incorrect effects for exception safety</h3>
26051 <p><b>Section:</b> 30.5 [thread.condition] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26052  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
26053 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition">active issues</a> in [thread.condition].</p>
26054 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition">issues</a> in [thread.condition].</p>
26055 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26056 <p><b>Discussion:</b></p>
26057 <p>
26058 30.5.1 [thread.condition.condvar] says:
26059 </p>
26060
26061 <blockquote>
26062 <pre>void wait(unique_lock&lt;mutex&gt;&amp; lock);
26063 </pre>
26064 <blockquote>
26065 <p>...</p>
26066 <p>
26067 <i>Effects:</i>
26068 </p>
26069 <ul>
26070 <li>...</li>
26071 <li>
26072 If the function exits via an exception, <tt>lock.unlock()</tt> shall be
26073 called prior to exiting the function scope.
26074 </li>
26075 </ul>
26076 </blockquote>
26077 </blockquote>
26078
26079 <p>
26080 Should that be <tt>lock.lock()</tt>?
26081 </p>
26082
26083
26084 <p><b>Proposed resolution:</b></p>
26085
26086 <p>
26087 Change 30.5.1 [thread.condition.condvar] p10:
26088 </p>
26089
26090 <blockquote>
26091 <pre>void wait(unique_lock&lt;mutex&gt;&amp; lock);
26092 </pre>
26093 <blockquote>
26094 <p>...</p>
26095 <p>
26096 <i>Effects:</i>
26097 </p>
26098 <ul>
26099 <li>...</li>
26100 <li>
26101 If the function exits via an exception, <tt>lock.<del>un</del>lock()</tt> shall be
26102 called prior to exiting the function scope.
26103 </li>
26104 </ul>
26105 </blockquote>
26106 </blockquote>
26107
26108 <p>
26109 And make a similar change in p16, and in 30.5.2 [thread.condition.condvarany],
26110 p8 and p13.
26111 </p>
26112
26113
26114
26115
26116
26117
26118 <hr>
26119 <h3><a name="1223"></a>1223. condition_variable_any lock matching?</h3>
26120 <p><b>Section:</b> 30.5.2 [thread.condition.condvarany] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26121  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
26122 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvarany">active issues</a> in [thread.condition.condvarany].</p>
26123 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvarany">issues</a> in [thread.condition.condvarany].</p>
26124 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26125 <p><b>Discussion:</b></p>
26126 <p>
26127 For <tt>condition_variable_any</tt>, must all lock arguments to concurrent wait calls
26128 "match" in some way, similar to the requirement in
26129 30.5.1 [thread.condition.condvar] that <tt>lock.mutex()</tt> returns the same
26130 value for each of the lock arguments supplied by all concurrently
26131 waiting threads (via <tt>wait</tt> or <tt>timed_wait</tt>)?
26132 </p>
26133
26134
26135
26136 <p><b>Proposed resolution:</b></p>
26137
26138
26139
26140
26141
26142 <hr>
26143 <h3><a name="1224"></a>1224. condition_variable_any support for recursive mutexes?</h3>
26144 <p><b>Section:</b> 30.5.2 [thread.condition.condvarany] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26145  <b>Submitter:</b> Jeffrey Yasskin <b>Opened:</b> 2009-09-30  <b>Last modified:</b> 2009-09-30</p>
26146 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#thread.condition.condvarany">active issues</a> in [thread.condition.condvarany].</p>
26147 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#thread.condition.condvarany">issues</a> in [thread.condition.condvarany].</p>
26148 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26149 <p><b>Discussion:</b></p>
26150 <p>
26151 For <tt>condition_variable_any</tt>, are recursive mutexes allowed? (I think "no")
26152 </p>
26153
26154
26155
26156 <p><b>Proposed resolution:</b></p>
26157
26158
26159
26160
26161
26162 <hr>
26163 <h3><a name="1225"></a>1225. C++0x result_of issue </h3>
26164 <p><b>Section:</b> 20.7.4 [func.ret] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26165  <b>Submitter:</b> Sebastian Gesemann <b>Opened:</b> 2009-10-05  <b>Last modified:</b> 2009-10-17</p>
26166 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#func.ret">issues</a> in [func.ret].</p>
26167 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26168 <p><b>Discussion:</b></p>
26169 <p>
26170 I think the text about <tt>std::result_of</tt> could be a little more precise.
26171 Quoting from
26172 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>...
26173 </p>
26174
26175 <blockquote>
26176 <p>
26177 20.7.4 [func.ret] Function object return types
26178 </p>
26179
26180 <pre>template&lt;class&gt; class result_of;
26181
26182 template&lt;class Fn, class... ArgTypes&gt;
26183 class result_of&lt;Fn(ArgTypes...)&gt; {
26184 public:
26185   typedef <i>see below</i> type;
26186 };
26187 </pre>
26188
26189 <p>
26190 Given an rvalue <tt>fn</tt> of type <tt>Fn</tt> and values <tt>t1, t2,
26191 ..., tN</tt> of types <tt>T1, T2, ... TN</tt> in <tt>ArgTypes</tt>
26192 respectivly, the <tt>type</tt> member is the result type of the
26193 expression <tt>fn(t1,t2,...,tN)</tt>. the values <tt>ti</tt> are lvalues
26194 when the corresponding type <tt>Ti</tt> is an lvalue-reference type, and
26195 rvalues otherwise.
26196 </p>
26197 </blockquote>
26198
26199 <p>
26200 This text doesn't seem to consider lvalue reference types for <tt>Fn</tt>.
26201 Also, it's not clear whether this class template can be used for
26202 "SFINAE" like <tt>std::enable_if</tt>. Example:
26203 </p>
26204
26205 <blockquote><pre>template&lt;typename Fn, typename... Args&gt;
26206 typename std::result_of&lt;Fn(Args...)&gt;::type
26207 apply(Fn &amp;&amp; fn, Args &amp;&amp; ...args)
26208 {
26209   // Fn may be an lvalue reference, too
26210   return std::forward&lt;Fn&gt;(fn)(std::forward&lt;Args&gt;(args)...);
26211 }
26212 </pre></blockquote>
26213
26214 <p>
26215 Either <tt>std::result_of&lt;...&gt;</tt> can be instantiated and simply may not have
26216 a typedef "<tt>type</tt>" (--&gt;SFINAE) or instantiating the class template for
26217 some type combinations will be a "hard" compile-time error.
26218 </p>
26219
26220
26221 <p><b>Proposed resolution:</b></p>
26222
26223 <p><i>[
26224 These changes will require compiler support
26225 ]</i></p>
26226
26227
26228 <p>
26229 Change 20.7.4 [func.ret]:
26230 </p>
26231
26232 <blockquote><pre>template&lt;class&gt; class result_of; // <i>undefined</i>
26233
26234 template&lt;class Fn, class... ArgTypes&gt;
26235 class result_of&lt;Fn(ArgTypes...)&gt; {
26236 public:
26237   <del>typedef</del> <i>see below</i> <del>type;</del>
26238 };
26239 </pre>
26240
26241 <p><del>
26242 Given an rvalue <tt>fn</tt> of type <tt>Fn</tt> and values <tt>t1, t2,
26243 ..., tN</tt> of types <tt>T1, T2, ... TN</tt> in <tt>ArgTypes</tt>
26244 respectivly, the <tt>type</tt> member is the result type of the
26245 expression <tt>fn(t1,t2,...,tN)</tt>. the values <tt>ti</tt> are lvalues
26246 when the corresponding type <tt>Ti</tt> is an lvalue-reference type, and
26247 rvalues otherwise.
26248 </del></p>
26249
26250 <p>
26251 <ins>The class template <tt>result_of</tt> shall meet the requirements of a
26252 <i>TransformationTrait</i>: Given the types <tt>Fn</tt>, <tt>T1</tt>, <tt>T2</tt>, ..., <tt>TN</tt> every
26253 template specialization <tt>result_of&lt;Fn(T1,T2,...,TN)&gt;</tt> shall define the
26254 member typedef type equivalent to <tt>decltype(<i>RE</i>)</tt> if and only if
26255 the expression <tt><i>RE</i></tt>
26256 </ins></p>
26257
26258 <blockquote><pre><ins>
26259 value&lt;Fn&gt;() ( value&lt;T1&gt;(), value&lt;T2&gt;(), ... value&lt;TN&gt;()  )
26260 </ins></pre></blockquote>
26261
26262 <p><ins>
26263 would be well-formed. Otherwise, there shall be no member typedef
26264 <tt>type</tt> defined.
26265 </ins></p>
26266
26267 </blockquote>
26268
26269 <p><i>[
26270 The <tt>value&lt;&gt;</tt> helper function is a utility Daniel Krügler
26271 proposed in
26272 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2958.html">N2958</a>.
26273 ]</i></p>
26274
26275
26276
26277
26278
26279
26280 <hr>
26281 <h3><a name="1226"></a>1226. Incomplete changes of #890</h3>
26282 <p><b>Section:</b> 30.6.2 [futures.errors] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
26283  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-05  <b>Last modified:</b> 2009-10-27</p>
26284 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
26285 <p><b>Discussion:</b></p>
26286 <p>
26287 Defect issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#890">890</a> overlooked to adapt the <tt>future_category</tt> from
26288 30.6.1 [futures.overview] and 30.6.2 [futures.errors]:
26289 </p>
26290
26291 <blockquote><pre>extern const error_category* const future_category;
26292 </pre></blockquote>
26293
26294 <p>
26295 which should be similarly transformed into function form.
26296 </p>
26297
26298 <p><i>[
26299 2009-10-27 Howard:
26300 ]</i></p>
26301
26302
26303 <blockquote>
26304 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
26305 </blockquote>
26306
26307
26308
26309 <p><b>Proposed resolution:</b></p>
26310 <ol>
26311 <li>
26312 <p>
26313 Change in 30.6.1 [futures.overview], header <tt>&lt;future&gt;</tt> synopsis:
26314 </p>
26315
26316 <blockquote><pre><del>extern</del> const error_category<ins>&amp;</ins><del>* const</del> future_category<ins>()</ins>;
26317 </pre></blockquote>
26318 </li>
26319
26320 <li>
26321 <p>
26322 Change in 30.6.2 [futures.errors]:
26323 </p>
26324
26325 <blockquote><pre><del>extern</del> const error_category<ins>&amp;</ins><del>* const</del> future_category<ins>()</ins>;
26326 </pre>
26327
26328 <blockquote>
26329 <p>
26330 <del>1- <tt>future_category</tt> shall point to a statically initialized object
26331 of a type derived from class <tt>error_category</tt>.</del>
26332 </p>
26333 <p>
26334 <ins>1- <i>Returns:</i> A reference to an object of a type
26335 derived from class <tt>error_category</tt>.</ins>
26336 </p>
26337 </blockquote>
26338
26339 <pre>constexpr error_code make_error_code(future_errc e);
26340 </pre>
26341
26342 <blockquote>
26343 3 <i>Returns:</i> <tt>error_code(static_cast&lt;int&gt;(e),
26344 <del>*</del>future_category<ins>()</ins>)</tt>.
26345 </blockquote>
26346
26347 <pre>constexpr error_code make_error_condition(future_errc e);
26348 </pre>
26349
26350 <blockquote>
26351 4 <i>Returns:</i> <tt>error_condition(static_cast&lt;int&gt;(e),
26352 <del>*</del>future_category<ins>()</ins>)</tt>.
26353 </blockquote>
26354 </blockquote>
26355 </li>
26356 </ol>
26357
26358
26359
26360
26361
26362 <hr>
26363 <h3><a name="1227"></a>1227. <tt>&lt;bitset&gt;</tt> synopsis overspecified</h3>
26364 <p><b>Section:</b> 20.3.7 [template.bitset] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
26365  <b>Submitter:</b> Bo Persson <b>Opened:</b> 2009-10-05  <b>Last modified:</b> 2009-10-26</p>
26366 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#template.bitset">active issues</a> in [template.bitset].</p>
26367 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#template.bitset">issues</a> in [template.bitset].</p>
26368 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
26369 <p><b>Discussion:</b></p>
26370 <p>
26371 The resolutions to some library defect reports, like <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#1178">1178</a>
26372 requires that <tt>#includes</tt> in each synopsis should be taken
26373 literally. This means that the <tt>&lt;bitset&gt;</tt> header now
26374 <em>must</em> include <tt>&lt;stdexcept&gt;</tt>, even though none of the
26375 exceptions are mentioned in the <tt>&lt;bitset&gt;</tt> header.
26376 </p>
26377 <p>
26378 Many other classes are required to throw exceptions like
26379 <tt>invalid_argument</tt> and <tt>out_of_range</tt>, without explicitly
26380 including <tt>&lt;stdexcept&gt;</tt> in their synopsis. It is totally
26381 possible for implementations to throw the needed exceptions from utility
26382 functions, whose implementations are not visible in the headers.
26383 </p>
26384 <p>
26385 I propose that <tt>&lt;stdexcept&gt;</tt> is removed from the
26386 <tt>&lt;bitset&gt;</tt> header.
26387 </p>
26388
26389 <p><i>[
26390 2009-10 Santa Cruz:
26391 ]</i></p>
26392
26393
26394 <blockquote>
26395 Moved to Ready.
26396 </blockquote>
26397
26398
26399
26400 <p><b>Proposed resolution:</b></p>
26401 <p>
26402 Change 20.3.7 [template.bitset]:
26403 </p>
26404
26405 <blockquote><pre>#include &lt;cstddef&gt;        // for size_t
26406 #include &lt;string&gt;
26407 <del>#include &lt;stdexcept&gt;      // for invalid_argument,</del>
26408                           <del>// out_of_range, overflow_error</del>
26409 #include &lt;iosfwd&gt;         // for istream, ostream
26410 namespace std {
26411 ...
26412 </pre></blockquote>
26413
26414
26415
26416
26417
26418 <hr>
26419 <h3><a name="1228"></a>1228. User-specialized nothrow type traits</h3>
26420 <p><b>Section:</b> 20.6.4.3 [meta.unary.prop] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Open">Open</a>
26421  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-10-07  <b>Last modified:</b> 2009-10-26</p>
26422 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#meta.unary.prop">active issues</a> in [meta.unary.prop].</p>
26423 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.unary.prop">issues</a> in [meta.unary.prop].</p>
26424 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Open">Open</a> status.</p>
26425 <p><b>Discussion:</b></p>
26426 <p>
26427 According to p1 20.6.2 [meta.type.synop]:
26428 </p>
26429
26430 <blockquote>
26431 The behavior of a program that adds specializations for any of the class
26432 templates defined in this subclause is undefined unless otherwise
26433 specified.
26434 </blockquote>
26435
26436 <p>
26437 I believe we should 'otherwise specify' for the nothrow traits, are
26438 these are exactly the use cases where the end user actually has more
26439 information than the compiler.
26440 </p>
26441
26442 <p><i>[
26443 2009-10 Santa Cruz:
26444 ]</i></p>
26445
26446
26447 <blockquote>
26448 Moved to Open.  Definitely need to give the users the ability to ensure
26449 that the traits give the right answers. Unsure we want to give them the
26450 ability to say this in more than one way. Believes the noexcept proposal
26451 already gives this.
26452 </blockquote>
26453
26454
26455
26456 <p><b>Proposed resolution:</b></p>
26457 <p>
26458 Add the following comment:
26459 </p>
26460
26461 <blockquote>
26462 user specialization permitted to derive from <tt>std::true_type</tt> when the
26463 operation is known not to throw.
26464 </blockquote>
26465
26466 <p>
26467 to the following traits in 20.6.4.3 [meta.unary.prop] Table 43 Type
26468 property predicates.
26469 </p>
26470
26471 <p><i>[
26472 This may require a new Comments column
26473 ]</i></p>
26474
26475
26476 <blockquote><pre>has_nothrow_default_constructor
26477 has_nothrow_copy_constructor
26478 has_nothrow_assign
26479 </pre></blockquote>
26480
26481
26482
26483
26484
26485 <hr>
26486 <h3><a name="1231"></a>1231. <tt>weak_ptr</tt> comparisons incompletely resolved</h3>
26487 <p><b>Section:</b> 20.8.15.3.5 [util.smartptr.weak.obs] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
26488  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-10  <b>Last modified:</b> 2009-11-06</p>
26489 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
26490 <p><b>Discussion:</b></p>
26491 <p>
26492 The
26493 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2637.pdf">n2637</a>
26494 paper suggested several updates of the ordering semantics of
26495 <tt>shared_ptr</tt>
26496 and <tt>weak_ptr</tt>, among those the explicit comparison operators of <tt>weak_ptr</tt> were
26497 removed/deleted, instead a corresponding functor <tt>owner_less</tt> was added.
26498 The problem
26499 is that
26500 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2637.pdf">n2637</a>
26501 did not clearly enough specify, how the previous wording
26502 parts describing
26503 the comparison semantics of <tt>weak_ptr</tt> should be removed.
26504 </p>
26505
26506 <p><i>[
26507 2009-11-06 Howard adds:
26508 ]</i></p>
26509
26510
26511 <blockquote>
26512 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
26513 </blockquote>
26514
26515
26516
26517 <p><b>Proposed resolution:</b></p>
26518 <ol>
26519 <li>
26520 <p>
26521 Change 20.8.15.3 [util.smartptr.weak]/2 as described, the intention is to fix
26522 the now no longer valid
26523 requirement that <tt>weak_ptr</tt> is <tt>LessComparable</tt> [Note the deleted comma]:
26524 </p>
26525
26526 <blockquote>
26527 Specializations of <tt>weak_ptr</tt> shall be <tt>CopyConstructible</tt><del>,</del>
26528 <ins>and</ins> <tt>CopyAssignable</tt>,
26529 <del>and <tt>LessThanComparable</tt>,</del> allowing their use in standard containers.
26530 </blockquote>
26531 </li>
26532
26533 <li>
26534 <p>
26535 In 20.8.15.3.5 [util.smartptr.weak.obs] remove the paragraphs 9-11 including prototype:
26536 </p>
26537
26538 <blockquote>
26539 <del>template&lt;class T, class U&gt; bool operator&lt;(const weak_ptr&lt;T&gt;&amp; a, const weak_ptr&lt;U&gt;&amp; b);</del>
26540
26541 <p>
26542 <del><i>Returns:</i> an unspecified value such that</del>
26543 </p>
26544 <ul>
26545 <li>
26546 <del><tt>operator&lt;</tt> is a strict weak ordering as described in 25.4;</del>
26547 </li>
26548 <li>
26549 <del>under the equivalence relation defined by <tt>operator&lt;</tt>, <tt>!(a
26550 &lt; b) &amp;&amp; !(b &lt; a)</tt>, two <tt>weak_ptr</tt> instances are
26551 equivalent if and only if they share ownership or are both empty.</del>
26552 </li>
26553 </ul>
26554
26555 <p>
26556 <del><i>Throws:</i> nothing.</del>
26557 </p>
26558
26559 <p>
26560 <del>[<i>Note:</i> Allows <tt>weak_ptr</tt> objects to be used as keys in associative
26561 containers. &#8212; <i>end note</i>]</del>
26562 </p>
26563 </blockquote>
26564 </li>
26565 </ol>
26566
26567
26568
26569
26570
26571 <hr>
26572 <h3><a name="1233"></a>1233. Missing <tt>unique_ptr</tt> signatures in synopsis</h3>
26573 <p><b>Section:</b> 20.8 [memory] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Editorial">Tentatively NAD Editorial</a>
26574  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-11  <b>Last modified:</b> 2009-11-04</p>
26575 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#memory">issues</a> in [memory].</p>
26576 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Editorial">Tentatively NAD Editorial</a> status.</p>
26577 <p><b>Discussion:</b></p>
26578 <p>
26579 Related to <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#296">296</a>.  Some <tt>unique_ptr</tt> signatures are missing
26580 from the synopsis in 20.8 [memory].
26581 </p>
26582
26583 <p><i>[
26584 2009-11-04 Howard adds:
26585 ]</i></p>
26586
26587
26588 <blockquote>
26589 Moved to Tentatively NAD Editorial.  The editor has adopted the fix.
26590 </blockquote>
26591
26592
26593 <p><b>Proposed resolution:</b></p>
26594 <p>
26595 Add in 20.8 [memory], Header <tt>&lt;memory&gt;</tt> synopsis
26596 missing declarations as shown below:
26597 </p>
26598
26599 <blockquote><pre>// 20.8.11 Class unique_ptr:
26600 template &lt;class X&gt; class default_delete;
26601 <ins>template&lt;class T&gt; struct default_delete&lt;T[]&gt;;</ins>
26602 template &lt;class X, class D = default_delete&lt;T&gt;&gt; class unique_ptr;
26603 <ins>template&lt;class T, class D&gt; class unique_ptr&lt;T[], D&gt;;</ins>
26604
26605 <ins>template&lt;class T, class D&gt; void swap(unique_ptr&lt;T, D&gt;&amp; x, unique_ptr&lt;T, D&gt;&amp; y);</ins>
26606
26607 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26608 bool operator==(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26609 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26610 bool operator!=(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26611 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26612 bool operator&lt;(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26613 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26614 bool operator&lt;=(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26615 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26616 bool operator&gt;(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26617 <ins>template&lt;class T1, class D1, class T2, class D2&gt;
26618 bool operator&gt;=(const unique_ptr&lt;T1, D1&gt;&amp; x, const unique_ptr&lt;T2, D2&gt;&amp; y);</ins>
26619 </pre></blockquote>
26620
26621
26622
26623
26624
26625 <hr>
26626 <h3><a name="1234"></a>1234. "Do the right thing" and NULL</h3>
26627 <p><b>Section:</b> 23.2.3 [sequence.reqmts] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
26628  <b>Submitter:</b> Matt Austern <b>Opened:</b> 2009-10-09  <b>Last modified:</b> 2009-10-13</p>
26629 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#sequence.reqmts">active issues</a> in [sequence.reqmts].</p>
26630 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#sequence.reqmts">issues</a> in [sequence.reqmts].</p>
26631 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
26632 <p><b>Discussion:</b></p>
26633 <p>
26634 On g++ 4.2.4 (x86_64-linux-gnu), the following file gives a compile
26635 error:
26636 </p>
26637
26638 <blockquote><pre>#include &lt;vector&gt;
26639 void foo() { std::vector&lt;int*&gt; v(500l, NULL); }
26640 </pre></blockquote>
26641
26642 <p>
26643 Is this supposed to work?
26644 </p>
26645
26646 <p>
26647 The issue: if <tt>NULL</tt> happens to be defined as <tt>0l</tt>, this is an invocation of
26648 the constructor with two arguments of the same integral type.
26649 23.2.3 [sequence.reqmts]/11 says that this will behave as if the the
26650 overloaded constructor
26651 </p>
26652
26653 <blockquote><pre>X(size_type, const value_type&amp; = value_type(),
26654   const allocator_type&amp; = allocator_type())
26655 </pre></blockquote>
26656
26657 <p>
26658 were called instead, with the arguments
26659 <tt>static_cast&lt;size_type&gt;(first)</tt>, <tt>last</tt> and
26660 <tt>alloc</tt>, respectively. However, it does not say whether this
26661 actually means invoking that constructor with the exact textual form of
26662 the arguments as supplied by the user, or whether the standard permits
26663 an implementation to invoke that constructor with variables of the same
26664 type and value as what the user passed in. In most cases this is a
26665 distinction without a difference. In this particular case it does make a
26666 difference, since one of those things is a null pointer constant and the
26667 other is not.
26668 </p>
26669
26670 <p>
26671 Note that an implementation based on forwarding functions will use the
26672 latter interpretation.
26673 </p>
26674
26675
26676 <p><b>Proposed resolution:</b></p>
26677
26678
26679
26680
26681
26682 <hr>
26683 <h3><a name="1237"></a>1237. Constrained error_code/error_condition members</h3>
26684 <p><b>Section:</b> 19.5 [syserr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Ready</a>
26685  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-14  <b>Last modified:</b> 2009-10-26</p>
26686 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#syserr">issues</a> in [syserr].</p>
26687 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Ready">Ready</a> status.</p>
26688 <p><b>Discussion:</b></p>
26689 <p>
26690 I'm just reflecting on the now SFINAE-constrained constructors
26691 and assignment operators of <tt>error_code</tt> and <tt>error_condition</tt>:
26692 </p>
26693 <p>
26694 These are the <em>only</em> library components that are pro-actively
26695 announcing that they are using <tt>std::enable_if</tt> as constraining tool,
26696 which has IMO several disadvantages:
26697 </p>
26698
26699 <ol>
26700 <li>
26701 <p>
26702 With the availability of template default arguments and
26703 decltype, using <tt>enable_if</tt> in C++0x standard library, seems
26704 unnecessary restricting implementation freedom. E.g. there
26705 should be not need for a useless specification of a dummy
26706 default function argument, which only confuses the reader.
26707 A more reasonable implementation could e.g. be
26708 </p>
26709
26710 <blockquote><pre>template &lt;class ErrorCodeEnum
26711  class = typename enable_if&lt;is_error_code_enum&lt;ErrorCodeEnum&gt;::value&gt;::type&gt;
26712 error_code(ErrorCodeEnum e);
26713 </pre></blockquote>
26714
26715 <p>
26716 As currently specified, the function signatures are so unreadable,
26717 that errors quite easily happen, see e.g. <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1229">1229</a>.
26718 </p>
26719 </li>
26720
26721 <li>
26722 <p>
26723 We have a <em>lot</em> of constrained functions in other places, that
26724 now have a standard phrase that is easily understandable:
26725 </p>
26726
26727 <blockquote>
26728 <i>Remarks:</i> This constructor/function shall participate in overload
26729 resolution if and only if X.
26730 </blockquote>
26731
26732 <p>
26733 where X describes the condition. Why should these components deviate?
26734 </p>
26735 </li>
26736
26737 <li>
26738 <p>
26739 If <tt>enable_if</tt> would <em>not</em> be explicitly specified, the standard library
26740 is much better prepared for the future. It would also be possible, that
26741 libraries with partial support for not-yet-standard-concepts could provide
26742 a much better diagnostic as is possible with <tt>enable_if</tt>. This again
26743 would allow for experimental concept implementations in the wild,
26744 which as a result would make concept standardization a much more
26745 natural thing, similar to the way as templates were standardized
26746 in C++.
26747 </p>
26748
26749 <p>
26750 In summary: I consider it as a library defect that <tt>error_code</tt> and
26751 <tt>error_condition</tt> explicitly require a dependency to <tt>enable_if</tt> and
26752 do limit implementation freedom and I volunteer to prepare a
26753 corresponding resolution.
26754 </p>
26755 </li>
26756 </ol>
26757
26758 <p><i>[
26759 2009-10-18 Beman adds:
26760 ]</i></p>
26761
26762
26763 <blockquote>
26764 I support this proposed resolution, and thank Daniel for writing it up.
26765 </blockquote>
26766
26767 <p><i>[
26768 2009-10 Santa Cruz:
26769 ]</i></p>
26770
26771
26772 <blockquote>
26773 Moved to Ready.
26774 </blockquote>
26775
26776
26777
26778 <p><b>Proposed resolution:</b></p>
26779 <p><i>[
26780 Should this resolution be accepted, I recommend to resolve <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1229">1229</a> as NAD
26781 ]</i></p>
26782
26783
26784 <ol>
26785 <li>
26786 <p>
26787 In 19.5.2.1 [syserr.errcode.overview]/1, class <tt>error_code</tt>,
26788 change as indicated:
26789 </p>
26790
26791 <blockquote><pre>// 19.5.2.2 constructors:
26792 error_code();
26793 error_code(int val, const error_category&amp; cat);
26794 template &lt;class ErrorCodeEnum&gt;
26795   error_code(ErrorCodeEnum e<del>,
26796     typename enable_if&lt;is_error_code_enum&lt;ErrorCodeEnum&gt;::value&gt;::type * = 0</del>);
26797
26798 // 19.5.2.3 modifiers:
26799 void assign(int val, const error_category&amp; cat);
26800 template &lt;class ErrorCodeEnum&gt;
26801   <del>typename enable_if&lt;is_error_code_enum&lt;ErrorCodeEnum&gt;::value&gt;::type</del><ins>error_code</ins>&amp;
26802     operator=(ErrorCodeEnum e);
26803 void clear();
26804 </pre></blockquote>
26805 </li>
26806
26807 <li>
26808 <p>
26809 Change 19.5.2.2 [syserr.errcode.constructors] around the prototype before p. 7:
26810 </p>
26811
26812 <blockquote><pre>template &lt;class ErrorCodeEnum&gt;
26813 error_code(ErrorCodeEnum e<del>,
26814   typename enable_if&lt;is_error_code_enum&lt;ErrorCodeEnum&gt;::value&gt;::type * = 0</del>);
26815 </pre>
26816 <blockquote>
26817 <p>
26818 <ins><i>Remarks:</i> This constructor shall not participate in overload
26819 resolution, unless
26820 <tt>is_error_code_enum&lt;ErrorCodeEnum&gt;::value == true</tt>.</ins>
26821 </p>
26822 </blockquote>
26823 </blockquote>
26824 </li>
26825
26826 <li>
26827 <p>
26828 Change 19.5.2.3 [syserr.errcode.modifiers] around the prototype before p. 3:
26829 </p>
26830
26831 <blockquote><pre>template &lt;class ErrorCodeEnum&gt;
26832   <del>typename enable_if&lt;is_error_code_enum&lt;ErrorCodeEnum&gt;::value&gt;::type</del><ins>error_code</ins>&amp;
26833     operator=(ErrorCodeEnum e);
26834 </pre>
26835
26836 <blockquote>
26837 <ins><i>Remarks:</i> This operator shall not participate in overload resolution, unless
26838 <tt>is_error_code_enum&lt;ErrorCodeEnum&gt;::value == true</tt>.</ins>
26839 </blockquote>
26840 </blockquote>
26841 </li>
26842
26843 <li>
26844 <p>
26845 In 19.5.3.1 [syserr.errcondition.overview]/1, class <tt>error_condition</tt>, change
26846 as indicated:
26847 </p>
26848
26849 <blockquote><pre>// 19.5.3.2 constructors:
26850 error_condition();
26851 error_condition(int val, const error_category&amp; cat);
26852 template &lt;class ErrorConditionEnum&gt;
26853   error_condition(ErrorConditionEnum e<del>,
26854     typename enable_if&lt;is_error_condition_enum&lt;ErrorConditionEnum&gt;::type* = 0</del>);
26855
26856 // 19.5.3.3 modifiers:
26857 void assign(int val, const error_category&amp; cat);
26858 template&lt;<del>typename</del><ins>class</ins> ErrorConditionEnum&gt;
26859   <del>typename enable_if&lt;is_error_condition_enum&lt;ErrorConditionEnum&gt;, error_code&gt;::type</del><ins>error_condition</ins> &amp;
26860     operator=( ErrorConditionEnum e );
26861 void clear();
26862 </pre></blockquote>
26863 </li>
26864
26865 <li>
26866 <p>
26867 Change 19.5.3.2 [syserr.errcondition.constructors] around the
26868 prototype before p. 7:
26869 </p>
26870
26871 <blockquote><pre>template &lt;class ErrorConditionEnum&gt;
26872   error_condition(ErrorConditionEnum e<del>,
26873     typename enable_if&lt;is_error_condition_enum&lt;ErrorConditionEnum&gt;::value&gt;::type* = 0</del>);
26874 </pre>
26875 <blockquote>
26876 <ins><i>Remarks:</i> This constructor shall not participate in overload
26877 resolution, unless
26878 <tt>is_error_condition_enum&lt;ErrorConditionEnum&gt;::value == true</tt>.</ins>
26879 </blockquote>
26880 </blockquote>
26881 </li>
26882
26883 <li>
26884 <p>
26885 Change 19.5.3.3 [syserr.errcondition.modifiers] around the
26886 prototype before p. 3:
26887 </p>
26888
26889 <blockquote><pre>template &lt;class ErrorConditionEnum&gt;
26890   <del>typename enable_if&lt;is_error_condition_enum&lt;ErrorConditionEnum&gt;::value&gt;::type</del><ins>error_condition</ins>&amp;
26891     operator=(ErrorConditionEnum e);
26892 </pre>
26893
26894 <blockquote>
26895 <p>
26896 <ins><i>Remarks:</i> This operator shall not participate in overload resolution, unless
26897 <tt>is_error_condition_enum&lt;ErrorConditionEnum&gt;::value == true</tt>.</ins>
26898 </p>
26899
26900 <p>
26901 <i>Postcondition:</i> <tt>*this == make_error_condition(e)</tt>.
26902 </p>
26903
26904 <p>
26905 <ins><i>Returns:</i> <tt>*this</tt></ins>
26906 </p>
26907 </blockquote>
26908 </blockquote>
26909
26910 </li>
26911 </ol>
26912
26913
26914
26915
26916
26917
26918 <hr>
26919 <h3><a name="1238"></a>1238. defining algorithms taking iterator for range</h3>
26920 <p><b>Section:</b> 25 [algorithms] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Future">Tentatively NAD Future</a>
26921  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-10-15  <b>Last modified:</b> 2009-11-03</p>
26922 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#algorithms">active issues</a> in [algorithms].</p>
26923 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#algorithms">issues</a> in [algorithms].</p>
26924 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Future">Tentatively NAD Future</a> status.</p>
26925 <p><b>Discussion:</b></p>
26926 <p>
26927 The library has many algorithms that take a source range represented by
26928 a pair of iterators, and the start of some second sequence given by a
26929 single iterator.  Internally, these algorithms will produce undefined
26930 behaviour if the second 'range' is not as large as the input range, but
26931 none of the algorithms spell this out in Requires clauses, and there is
26932 no catch-all wording to cover this in clause 17 or the front matter of
26933 25.
26934 </p>
26935
26936 <p>
26937 There was an attempt to provide such wording in paper
26938 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2944.pdf">n2944</a>
26939 but this
26940 seems incidental to the focus of the paper, and getting the wording of
26941 this issue right seems substantially more difficult than the simple
26942 approach taken in that paper.  Such wording will be removed from an
26943 updated paper, and hopefully tracked via the LWG issues list instead.
26944 </p>
26945
26946 <p>
26947 It seems there are several classes of problems here and finding wording
26948 to solve all in one paragraph could be too much.  I suspect we need
26949 several overlapping requirements that should cover the desired range of
26950 behaviours.
26951 </p>
26952
26953 <p>
26954 Motivating examples:
26955 </p>
26956
26957 <p>
26958 A good initial example is the <tt>swap_ranges</tt> algorithm.  Here there is a
26959 clear requirement that <tt>first2</tt> refers to the start of a valid range at
26960 least as long as the range <tt>[first1, last1)</tt>.  <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2944.pdf">n2944</a> tries to solve this
26961 by positing a hypothetical <tt>last2</tt> iterator that is implied by the
26962 signature, and requires <tt>distance(first2,last2) &lt; distance(first1,last1)</tt>.
26963  This mostly works, although I am uncomfortable assuming that <tt>last2</tt> is
26964 clearly defined and well known without any description of how to obtain
26965 it (and I have no idea how to write that).
26966 </p>
26967
26968 <p>
26969 A second motivating example might be the <tt>copy</tt> algorithm.  Specifically,
26970 let us image a call like:
26971 </p>
26972
26973 <blockquote><pre>copy(istream_iterator&lt;int&gt;(is),istream_iterator(),ostream_iterator&lt;int&gt;(os));
26974 </pre></blockquote>
26975
26976 <p>
26977 In this case, our input iterators are literally simple <tt>InputIterators</tt>,
26978 and the destination is a simple <tt>OutputIterator</tt>.  In neither case am I
26979 happy referring to <tt>std::distance</tt>, in fact it is not possible for the
26980 <tt>ostream_iterator</tt> at all as it does not meet the requirements.  However,
26981 any wording we provide must cover both cases.  Perhaps we might deduce
26982 <tt>last2 == ostream_iterator&lt;int&gt;{}</tt>, but that might not always be valid for
26983 user-defined iterator types.  I can well imagine an 'infinite range'
26984 that writes to <tt>/dev/null</tt> and has no meaningful <tt>last2</tt>.
26985 </p>
26986
26987 <p>
26988 The motivating example in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2944.pdf">n2944</a> is <tt>std::equal</tt>, and that seems to fall somewhere between the
26989 two.
26990 </p>
26991
26992 <p>
26993 Outlying examples might be <tt>partition_copy</tt> that takes two output
26994 iterators, and the <tt>_n</tt> algorithms where a range is specified by a
26995 specific number of iterations, rather than traditional iterator pair.
26996 We should also <em>not</em> accidentally apply inappropriate constraints to
26997 <tt>std::rotate</tt> which takes a third iterator that is not intended to be a
26998 separate range at all.
26999 </p>
27000
27001 <p>
27002 I suspect we want some wording similar to:
27003 </p>
27004
27005 <blockquote>
27006 For algorithms that operate on ranges where the end iterator of the
27007 second range is not specified, the second range shall contain at least
27008 as many elements as the first.
27009 </blockquote>
27010
27011 <p>
27012 I don't think this quite captures the intent yet though.  I am not sure
27013 if 'range' is the right term here rather than sequence.  More awkwardly,
27014 I am not convinced we can describe an Output sequence such as produce by
27015 an <tt>ostream_iterator</tt> as "containing elements", at least not as a
27016 precondition to the call before they have been written.
27017 </p>
27018
27019 <p>
27020 Another idea was to describe require that the trailing iterator support
27021 at least distance(input range) applications of <tt>operator++</tt> and may be
27022 written through the same number of times if a mutable/output iterator.
27023 </p>
27024
27025 <p>
27026 We might also consider handling the case of an output range vs. an input
27027 range in separate paragraphs, if that simplifies how we describe some of
27028 these constraints.
27029 </p>
27030
27031 <p><i>[
27032 2009-11-03 Howard adds:
27033 ]</i></p>
27034
27035
27036 <blockquote>
27037 Moved to Tentatively NAD Future after 5 positive votes on c++std-lib.
27038 </blockquote>
27039
27040
27041 <p><b>Proposed resolution:</b></p>
27042
27043
27044
27045
27046
27047 <hr>
27048 <h3><a name="1239"></a>1239. Defect report</h3>
27049 <p><b>Section:</b> 20.6.4.3 [meta.unary.prop] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#NAD%20Editorial">Tentatively NAD Editorial</a>
27050  <b>Submitter:</b> David Abrahams <b>Opened:</b> 2009-10-16  <b>Last modified:</b> 2009-10-26</p>
27051 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#meta.unary.prop">active issues</a> in [meta.unary.prop].</p>
27052 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.unary.prop">issues</a> in [meta.unary.prop].</p>
27053 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20NAD%20Editorial">Tentatively NAD Editorial</a> status.</p>
27054 <p><b>Discussion:</b></p>
27055 <p>
27056 Table 43 defines a number of traits that yield true for arrays of class
27057 types with the trait's property, but not arrays of other types with that
27058 property.  For example, <tt>has_trivial_default_constructor</tt>:
27059 </p>
27060
27061 <blockquote>
27062 <tt>T</tt> is a trivial type (3.9) or a class type with a trivial default
27063 constructor (12.1) or an array of such a class type.
27064 </blockquote>
27065
27066 <p><i>[
27067 2009-10 post-Santa Cruz:
27068 ]</i></p>
27069
27070
27071 <blockquote>
27072 <p>
27073 An array of a trivial type is a trivial type.
27074 </p>
27075 <p>
27076 Mark as Tentatively NAD Editorial. The wording is OK as is,
27077 since an array of a trivial type is a trivial type, but the wording as
27078 proposed might be clearer.
27079 </p>
27080 </blockquote>
27081
27082
27083
27084 <p><b>Proposed resolution:</b></p>
27085 <p>
27086 Change all the traits in question following this pattern:
27087 </p>
27088
27089 <blockquote>
27090 <tt>T</tt> is a trivial type (3.9) or a class type with a trivial default
27091  constructor (12.1)<ins>,</ins> or an array of such a <del>class</del> type.
27092 </blockquote>
27093
27094 <p>
27095 i.e., add a comma and delete a "class."
27096 </p>
27097
27098
27099
27100
27101
27102 <hr>
27103 <h3><a name="1240"></a>1240. Deleted comparison functions of std::function not  needed</h3>
27104 <p><b>Section:</b> 20.7.15.2 [func.wrap.func] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27105  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-18  <b>Last modified:</b> 2009-10-19</p>
27106 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#func.wrap.func">issues</a> in [func.wrap.func].</p>
27107 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27108 <p><b>Discussion:</b></p>
27109 <p>
27110 The class template <tt>std::function</tt> contains the following member
27111 declarations:
27112 </p>
27113
27114 <blockquote><pre>// deleted overloads close possible hole in the type system
27115 template&lt;class R2, class... ArgTypes2&gt;
27116   bool operator==(const function&lt;R2(ArgTypes2...)&gt;&amp;) = delete;
27117 template&lt;class R2, class... ArgTypes2&gt;
27118   bool operator!=(const function&lt;R2(ArgTypes2...)&gt;&amp;) = delete;
27119 </pre></blockquote>
27120
27121 <p>
27122 The leading comment here is part of the history of <tt>std::function</tt>, which
27123 was introduced with <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2002/n1402.html#undefined_operators">N1402</a>.
27124 During that time no explicit conversion functions existed, and the
27125 "safe-bool" idiom (based on pointers-to-member) was a popular
27126 technique. The only disadvantage of this idiom was that given two
27127 objects <tt>f1</tt> and <tt>f2</tt> of type <tt>std::function</tt> the expression
27128 </p>
27129
27130 <blockquote><pre>f1 == f2;
27131 </pre></blockquote>
27132
27133 <p>
27134 was well-formed, just because the built-in <tt>operator==</tt> for pointer to member
27135 was considered after a single user-defined conversion. To fix this, an
27136 overload set of <em>undefined</em> comparison functions was added,
27137 such that overload resolution would prefer those ending up in a linkage error.
27138 The new language facility of deleted functions provided a much better
27139 diagnostic mechanism to fix this issue.
27140 </p>
27141
27142 <p>
27143 The central point of this issue is, that with the replacement of the
27144 safe-bool idiom by explicit conversion to bool the original "hole in the
27145 type system" does no longer exist and therefore the comment is wrong and
27146 the superfluous function definitions should be removed as well. An
27147 explicit conversion function is considered in direct-initialization
27148 situations only, which indirectly contain the so-called "contextual
27149 conversion to bool" (4 [conv]/3). These conversions are not considered for
27150 <tt>==</tt> or <tt>!=</tt> as defined by the core language.
27151 </p>
27152
27153
27154 <p><b>Proposed resolution:</b></p>
27155 <p>
27156 In 20.7.15.2 [func.wrap.func]/1, class function change as indicated:
27157 </p>
27158
27159 <blockquote><pre>// 20.7.15.2.3, function capacity:
27160 explicit operator bool() const;
27161
27162 <del>// deleted overloads close possible hole in the type system</del>
27163 <del>template&lt;class R2, class... ArgTypes2&gt;</del>
27164   <del>bool operator==(const function&lt;R2(ArgTypes2...)&gt;&amp;) = delete;</del>
27165 <del>template&lt;class R2, class... ArgTypes2&gt;</del>
27166   <del>bool operator!=(const function&lt;R2(ArgTypes2...)&gt;&amp;) = delete;</del>
27167 </pre></blockquote>
27168
27169
27170
27171
27172
27173 <hr>
27174 <h3><a name="1241"></a>1241. unique_copy needs to require EquivalenceRelation</h3>
27175 <p><b>Section:</b> 25.3.9 [alg.unique] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Ready">Tentatively Ready</a>
27176  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-10-17  <b>Last modified:</b> 2009-10-31</p>
27177 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#alg.unique">issues</a> in [alg.unique].</p>
27178 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Tentatively%20Ready">Tentatively Ready</a> status.</p>
27179 <p><b>Discussion:</b></p>
27180 <p>
27181 A lot of fixes were silently applied during concept-time and we should
27182 not lose them again. The Requires clause of 25.3.9 [alg.unique]/5
27183 doesn't mention that <tt>==</tt> and the predicate need to satisfy an
27184 <tt>EquivalenceRelation</tt>, as it is correctly said for
27185 <tt>unique</tt>. This was intentionally fixed during conceptification,
27186 were we had:
27187 </p>
27188
27189 <blockquote><pre>template&lt;InputIterator InIter, class OutIter&gt;
27190   requires OutputIterator&lt;OutIter, RvalueOf&lt;InIter::value_type&gt;::type&gt;
27191         &amp;&amp; EqualityComparable&lt;InIter::value_type&gt;
27192         &amp;&amp; HasAssign&lt;InIter::value_type, InIter::reference&gt;
27193         &amp;&amp; Constructible&lt;InIter::value_type, InIter::reference&gt;
27194   OutIter unique_copy(InIter first, InIter last, OutIter result);
27195
27196 template&lt;InputIterator InIter, class OutIter,
27197          EquivalenceRelation&lt;auto, InIter::value_type&gt; Pred&gt;
27198   requires OutputIterator&lt;OutIter, RvalueOf&lt;InIter::value_type&gt;::type&gt;
27199         &amp;&amp; HasAssign&lt;InIter::value_type, InIter::reference&gt;
27200         &amp;&amp; Constructible&lt;InIter::value_type, InIter::reference&gt;
27201         &amp;&amp; CopyConstructible&lt;Pred&gt;
27202   OutIter unique_copy(InIter first, InIter last, OutIter result, Pred pred);
27203 </pre></blockquote>
27204
27205 <p>
27206 Note that EqualityComparable implied an equivalence relation.
27207 </p>
27208
27209 <p><i>[
27210 N.B. <tt>adjacent_find</tt> was also specified to require
27211 <tt>EquivalenceRelation</tt>, but that was considered as a defect in
27212 concepts, see <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#1000">1000</a>
27213 ]</i></p>
27214
27215
27216 <p><i>[
27217 2009-10-31 Howard adds:
27218 ]</i></p>
27219
27220
27221 <blockquote>
27222 Moved to Tentatively Ready after 5 positive votes on c++std-lib.
27223 </blockquote>
27224
27225
27226
27227 <p><b>Proposed resolution:</b></p>
27228 <p>
27229 Change 25.3.9 [alg.unique]/5 as indicated:
27230 </p>
27231
27232 <blockquote><pre>template&lt;class InputIterator, class OutputIterator&gt;
27233   OutputIterator
27234     unique_copy(InputIterator first, InputIterator last, OutputIterator result);
27235
27236 template&lt;class InputIterator, class OutputIterator, class BinaryPredicate&gt;
27237   OutputIterator
27238     unique_copy(InputIterator first, InputIterator last,
27239                 OutputIterator result, BinaryPredicate pred);
27240 </pre>
27241 <blockquote>
27242 <i>Requires:</i> <ins>The comparison function shall be an equivalence
27243 relation.</ins> The ranges <tt>[first,last)</tt> and
27244 <tt>[result,result+(last-first))</tt> shall not overlap. The expression
27245 <tt>*result = *first</tt> shall be valid. If neither
27246 <tt>InputIterator</tt> nor <tt>OutputIterator</tt> meets the
27247 requirements of forward iterator then the value type of
27248 <tt>InputIterator</tt> shall be <tt>CopyConstructible</tt> (34) and
27249 <tt>CopyAssignable</tt> (table 36). Otherwise <tt>CopyConstructible</tt>
27250 is not required.
27251 </blockquote>
27252 </blockquote>
27253
27254
27255
27256
27257
27258 <hr>
27259 <h3><a name="1244"></a>1244. wait_*() in *future for synchronous functions</h3>
27260 <p><b>Section:</b> 30.6 [futures] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27261  <b>Submitter:</b> Detlef Vollmann <b>Opened:</b> 2009-10-22  <b>Last modified:</b> 2009-10-23</p>
27262 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#futures">issues</a> in [futures].</p>
27263 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27264 <p><b>Discussion:</b></p>
27265 <p>
27266 With the addition of <tt>async()</tt>, a <tt>future</tt> might be
27267 associated with a function that is not running in a different thread but
27268 is stored to by run synchronously on the <tt>get()</tt> call. It's not
27269 clear what the <tt>wait()</tt> functions should do in this case.
27270 </p>
27271
27272 <p>
27273 Suggested resolution:
27274 </p>
27275
27276 <p>
27277 Throw an exception.
27278 </p>
27279
27280
27281 <p><b>Proposed resolution:</b></p>
27282
27283
27284
27285
27286
27287 <hr>
27288 <h3><a name="1245"></a>1245. <tt>std::hash&lt;string&gt;</tt> &amp; co</h3>
27289 <p><b>Section:</b> 20.7.16 [unord.hash] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27290  <b>Submitter:</b> Paolo Carlini <b>Opened:</b> 2009-10-22  <b>Last modified:</b> 2009-10-25</p>
27291 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#unord.hash">active issues</a> in [unord.hash].</p>
27292 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#unord.hash">issues</a> in [unord.hash].</p>
27293 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27294 <p><b>Discussion:</b></p>
27295 <p>
27296 In 20.7.16 [unord.hash], <tt>operator()</tt> is specified as
27297 taking the argument by value. Moreover, it is said that <tt>operator()</tt> shall
27298 not throw exceptions.
27299 </p>
27300
27301 <p>
27302 However, for the specializations for class types, like <tt>string</tt>, <tt>wstring</tt>,
27303 etc, the former requirement seems suboptimal from the performance point
27304 of view (a specific PR has been filed about this in the GCC Bugzilla)
27305 and, together with the latter requirement, hard if not impossible to
27306 fulfill. It looks like pass by const reference should be allowed in such
27307 cases.
27308 </p>
27309
27310
27311 <p><b>Proposed resolution:</b></p>
27312 <p>
27313 Add to 20.7.16 [unord.hash]/2:
27314 </p>
27315
27316 <blockquote>
27317 <pre>namespace std {
27318   template &lt;class T&gt;
27319   struct hash : public std::unary_function&lt;T, std::size_t&gt; {
27320     std::size_t operator()(T val) const;
27321   };
27322 }
27323 </pre>
27324
27325 <p>
27326 The return value of <tt>operator()</tt> is unspecified, except that
27327 equal arguments shall yield the same result. <tt>operator()</tt> shall
27328 not throw exceptions. <ins>It is also unspecified whether
27329 <tt>operator()</tt> of <tt>std::hash</tt> specializations for class
27330 types takes its argument by value or const reference.</ins>
27331 </p>
27332 </blockquote>
27333
27334
27335
27336
27337
27338 <hr>
27339 <h3><a name="1246"></a>1246. <tt>vector::resize()</tt> missing efficiency guarantee</h3>
27340 <p><b>Section:</b> 23.3.6.2 [vector.capacity] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27341  <b>Submitter:</b> David Abrahams <b>Opened:</b> 2009-10-24  <b>Last modified:</b> 2009-10-25</p>
27342 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#vector.capacity">issues</a> in [vector.capacity].</p>
27343 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27344 <p><b>Discussion:</b></p>
27345 <p>
27346 If <tt>v</tt> is a <tt>vector</tt>, I think repeated calls to
27347 <tt>v.resize( v.size() + 1 )</tt> should be amortized O(1), but it's not
27348 clear that's true from the text of the standard:
27349 </p>
27350
27351 <blockquote><pre>void resize(size_type sz);
27352 </pre>
27353 <blockquote>
27354 <i>Effects:</i> If <tt>sz &lt; size()</tt>, equivalent to <tt>erase(begin() + sz, end());</tt>. If
27355 <tt>size() &lt; sz</tt>, appends <tt>sz - size()</tt> default constructed elements to the
27356 sequence.
27357 </blockquote>
27358 </blockquote>
27359
27360 <p>
27361 Seems to me if we used <tt>push_back</tt> instead of appends, we might be giving
27362 the guarantee I'd like.  Thoughts?
27363 </p>
27364
27365
27366 <p><b>Proposed resolution:</b></p>
27367 <p>
27368 In 23.3.6.2 [vector.capacity]/10, change
27369 </p>
27370
27371 <blockquote><pre>void resize(size_type sz);
27372 </pre>
27373 <blockquote>
27374 <i>Effects:</i> If <tt>sz &lt; size()</tt>, equivalent to <tt>erase(begin() + sz, end());</tt>. If
27375 <tt>size() &lt; sz</tt>, <del>appends <tt>sz - size()</tt> default constructed elements to the
27376 sequence</del>
27377 <ins>equivalent to <tt>sz - size()</tt> consecutive evaluations of <tt>push_back(T())</tt></ins>.
27378 </blockquote>
27379 </blockquote>
27380
27381
27382
27383
27384
27385
27386 <hr>
27387 <h3><a name="1247"></a>1247. <tt>auto_ptr</tt> is overspecified</h3>
27388 <p><b>Section:</b> D.10.1 [auto.ptr] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#Review">Review</a>
27389  <b>Submitter:</b> Alisdair Meredith <b>Opened:</b> 2009-10-24  <b>Last modified:</b> 2009-11-06</p>
27390 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#auto.ptr">issues</a> in [auto.ptr].</p>
27391 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#Review">Review</a> status.</p>
27392 <p><b>Discussion:</b></p>
27393 <p>
27394 This issue is extracted as the ongoing point-of-interest from earlier
27395 issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#463">463</a>.
27396 </p>
27397
27398 <p>
27399 <tt>auto_ptr</tt> is overspecified as the <tt>auto_ptr_ref</tt>
27400 implementation detail is formally specified, and the technique is
27401 observable so workarounds for compiler defects can  cause a working
27402 implementation of the primary <tt>auto_ptr</tt> template become
27403 non-conforming.
27404 </p>
27405
27406 <p>
27407 <tt>auto_ptr_ref</tt> is a documentation aid to describe a possible
27408 mechanism to implement the class.  It should be marked exposition only,
27409 as per similar classes, e.g., <tt>istreambuf_iterator::proxy</tt>
27410 </p>
27411
27412 <p><i>[
27413 2009-10-25 Daniel adds:
27414 ]</i></p>
27415
27416
27417 <blockquote>
27418 <p>
27419 I wonder, whether the revised wording shouldn't be as straight as
27420 for <tt>istream_buf</tt> by adding one further sentence:
27421 </p>
27422
27423 <blockquote>
27424 An implementation is permitted to provide equivalent functionality without
27425 providing a class with this name.
27426 </blockquote>
27427 </blockquote>
27428
27429 <p><i>[
27430 2009-11-06 Alisdair adds Daniel's suggestion to the proposed wording.
27431 ]</i></p>
27432
27433
27434 <p><i>[
27435 2009-11-06 Howard moves issue to Review.
27436 ]</i></p>
27437
27438
27439
27440
27441 <p><b>Proposed resolution:</b></p>
27442 <p>
27443 Add the term "exposition only" in the following two places:
27444 </p>
27445
27446 <p>
27447 Ammend D.10.1 [auto.ptr]p2:
27448 </p>
27449
27450 <blockquote>
27451 <p>
27452 <ins>The exposition only class </ins> <del>T</del><ins>t</ins>emplate <tt>auto_ptr_ref</tt>
27453 holds a reference to an <tt>auto_ptr</tt>. It is used by the
27454 <tt>auto_ptr</tt> conversions to allow <tt>auto_ptr</tt> objects to be
27455 passed to and returned from functions.
27456 <ins>An implementation is permitted to provide equivalent functionality
27457 without providing a class with this name.</ins>
27458 </p>
27459
27460 <blockquote><pre>namespace std {
27461  template &lt;class Y&gt; struct auto_ptr_ref { }; <ins>// exposition only</ins>
27462 </pre></blockquote>
27463 </blockquote>
27464
27465
27466
27467
27468
27469 <hr>
27470 <h3><a name="1249"></a>1249. basic_ios default ctor</h3>
27471 <p><b>Section:</b> 27.5.4.1 [basic.ios.cons] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27472  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2009-10-25  <b>Last modified:</b> 2009-10-26</p>
27473 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#basic.ios.cons">issues</a> in [basic.ios.cons].</p>
27474 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27475 <p><b>Discussion:</b></p>
27476 <p>
27477 The basic_ios default ctor is required to leave the objects members
27478 uninitialized (see below). The paragraph says the object must be
27479 initialized by calling basic_ios::init() before it's destroyed by
27480 I can't find a requirement that it be initialized before calling
27481 any of the class other member functions. Am I not looking in the
27482 right place or that an issue?
27483 </p>
27484
27485 <p><i>[
27486 2009-10-25 Daniel adds:
27487 ]</i></p>
27488
27489
27490 <blockquote>
27491 <p>
27492 I agree, that your wording makes that clearer, but suggest to write
27493 </p>
27494
27495 <blockquote>
27496 ... calling <tt>basic_ios::init<del>()</del></tt> before ...
27497 </blockquote>
27498
27499 <p>
27500 Doing so, I recommend to adapt that of <tt>ios_base();</tt> as well, where
27501 we have:
27502 </p>
27503
27504 <blockquote>
27505 <i>Effects:</i> Each <tt>ios_base</tt> member has an indeterminate value
27506 after construction. These members shall be initialized by calling
27507 <tt>basic_ios::init</tt>. If an <tt>ios_base</tt> object is destroyed
27508 before these initializations have taken place, the behavior is
27509 undefined.
27510 </blockquote>
27511 </blockquote>
27512
27513
27514
27515 <p><b>Proposed resolution:</b></p>
27516 <p>
27517 Change 27.5.2.7 [ios.base.cons] p1:
27518 </p>
27519
27520 <blockquote><pre>ios_base();
27521 </pre>
27522 <blockquote>
27523 <i>Effects:</i> Each <tt>ios_base</tt> member has an indeterminate value
27524 after construction. <del>These</del> <ins>The object's</ins> members shall be initialized by calling
27525 <tt>basic_ios::init</tt> <ins>before the object's first use or before
27526  it is destroyed, whichever comes first; otherwise the behavior
27527  is undefined.</ins>. <del>If an <tt>ios_base</tt> object is destroyed
27528 before these initializations have taken place, the behavior is
27529 undefined.</del>
27530 </blockquote>
27531 </blockquote>
27532
27533 <p>
27534 Change 27.5.4.1 [basic.ios.cons] p2:
27535 </p>
27536
27537 <blockquote><pre>basic_ios();
27538 </pre>
27539 <blockquote>
27540 <i>Effects:</i> Constructs an object of class <tt>basic_ios</tt>
27541 (27.5.2.7) leaving its member objects uninitialized. The object shall be
27542 initialized by calling <del>its</del>
27543 <tt><ins>basic_ios::</ins>init</tt> <ins>before its first
27544 use or before it is destroyed, whichever comes first; otherwise the
27545 behavior is undefined.</ins> <del>member function. If it is destroyed
27546 before it has been initialized the behavior is undefined.</del>
27547 </blockquote>
27548 </blockquote>
27549
27550
27551
27552
27553
27554 <hr>
27555 <h3><a name="1250"></a>1250. <tt>&lt;bitset&gt;</tt> still overspecified</h3>
27556 <p><b>Section:</b> 20.3.7 [template.bitset] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27557  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2009-10-29  <b>Last modified:</b> 2009-10-29</p>
27558 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#template.bitset">active issues</a> in [template.bitset].</p>
27559 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#template.bitset">issues</a> in [template.bitset].</p>
27560 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27561 <p><b>Discussion:</b></p>
27562 <p>
27563 Issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1227">1227</a> &#8212; <tt>&lt;bitset&gt;</tt> synopsis overspecified makes the observation
27564 that <tt>std::bitset</tt>, and in fact the whole library, may be implemented
27565 without needing to <tt>#include &lt;stdexcept&gt;</tt> in any library header. The
27566 proposed resolution removes the <tt>#include &lt;stdexcept&gt;</tt> directive from
27567 the header.
27568 </p>
27569
27570 <p>
27571 I'd like to add that the <tt>&lt;bitset&gt;</tt> header (as well as the rest of
27572 the library) has also been implemented without #including the
27573 <tt>&lt;cstddef&gt;</tt> header in any library header. In the case of <tt>std::bitset</tt>,
27574 the template is fully usable (i.e., it may be instantiated and all
27575 its member functions may be used) without ever mentioning <tt>size_t</tt>.
27576 In addition, just like no library header except for <tt>&lt;bitset&gt;</tt>
27577 <tt>#includes &lt;stdexcept&gt;</tt> in its synopsis, no header but <tt>&lt;bitset&gt;</tt>
27578 <tt>#includes &lt;cstddef&gt;</tt> either.
27579 </p>
27580
27581 <p>
27582 Thus I suggest that the <tt>#include &lt;cstddef&gt;</tt> directive be similarly
27583 removed from the synopsis of <tt>&lt;bitset&gt;</tt>.
27584 </p>
27585
27586
27587 <p><b>Proposed resolution:</b></p>
27588 <p>
27589 Change 20.3.7 [template.bitset]:
27590 </p>
27591
27592 <blockquote><pre><del>#include &lt;cstddef&gt;        // for size_t</del>
27593 #include &lt;string&gt;
27594 #include &lt;iosfwd&gt;         // for istream, ostream
27595 namespace std {
27596 ...
27597 </pre></blockquote>
27598
27599
27600
27601
27602
27603 <hr>
27604 <h3><a name="1251"></a>1251. move constructing <tt>basic_stringbuf</tt></h3>
27605 <p><b>Section:</b> 27.8.1.1 [stringbuf.cons] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27606  <b>Submitter:</b> Martin Sebor <b>Opened:</b> 2009-10-29  <b>Last modified:</b> 2009-10-29</p>
27607 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#stringbuf.cons">issues</a> in [stringbuf.cons].</p>
27608 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27609 <p><b>Discussion:</b></p>
27610 <p>
27611 I just came across issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a> -- Global permission to move, which
27612 seems to address the concern raised by the example in c++std-lib-25030.
27613 </p>
27614 <p>
27615 IIUC, the example violates the permission to assume that arguments
27616 bound to rvalue references are unnamed temporaries granted to
27617 implementations by the resolution of issue <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1204">1204</a> - Global permission
27618 to move.
27619 </p>
27620
27621 <p>
27622 I.e., the <tt>ostringstream(ostringstream &amp;&amp;rhs)</tt> ctor can leave the <tt>rhs</tt>
27623 pointers pointing to the newly constructed object's buffer just as
27624 long as the dtor doesn't change or invalidate the buffer. The caller
27625 may not make any assumptions about rhs after the move beyond it being
27626 safe to destroy or reassign.
27627 </p>
27628
27629 <p>
27630 So unless I misunderstood something, I still think the <tt>basic_stringbuf</tt>
27631 move ctor is overspecified. Specifically, I think the third sentence
27632 in the Effects clause and the last 6 bullets in the Postconditions
27633 clause can, and IMO should, be stricken.
27634 </p>
27635
27636
27637 <p><b>Proposed resolution:</b></p>
27638 <p>
27639 Strike from 27.8.1.1 [stringbuf.cons]:
27640 </p>
27641
27642 <blockquote><pre>basic_stringbuf(basic_stringbuf&amp;&amp; rhs);
27643 </pre>
27644 <blockquote>
27645 <p>
27646 <i>Effects:</i> Move constructs from the rvalue <tt>rhs</tt>. It is
27647 implementation-defined whether the sequence pointers in <tt>*this</tt>
27648 (<tt>eback()</tt>, <tt>gptr()</tt>, <tt>egptr()</tt>, <tt>pbase()</tt>,
27649 <tt>pptr()</tt>, <tt>epptr()</tt>) obtain the values which <tt>rhs</tt>
27650 had. <del>Whether they do or not, <tt>*this</tt> and <tt>rhs</tt> reference
27651 separate buffers (if any at all) after the construction.</del> The openmode,
27652 locale and any other state of <tt>rhs</tt> is also copied.
27653 </p>
27654
27655 <p>
27656 <i>Postconditions:</i> Let <tt>rhs_p</tt> refer to the state of
27657 <tt>rhs</tt> just prior to this construction and let <tt>rhs_a</tt>
27658 referto the state of <tt>rhs</tt> just after this construction.
27659 </p>
27660 <ul>
27661 <li>
27662 <tt>str() == rhs_p.str()</tt>
27663 </li>
27664 <li>
27665 <tt>gptr() - eback() == rhs_p.gptr() - rhs_p.eback()</tt>
27666 </li>
27667 <li>
27668 <tt>egptr() - eback() == rhs_p.egptr() - rhs_p.eback()</tt>
27669 </li>
27670 <li>
27671 <tt>pptr() - pbase() == rhs_p.pptr() - rhs_p.pbase()</tt>
27672 </li>
27673 <li>
27674 <tt>epptr() - pbase() == rhs_p.epptr() - rhs_p.pbase()</tt>
27675 </li>
27676 <li><del>
27677 if <tt>(eback()) eback() != rhs_a.eback()</tt>
27678 </del></li>
27679 <li><del>
27680 if <tt>(gptr()) gptr() != rhs_a.gptr()</tt>
27681 </del></li>
27682 <li><del>
27683 if <tt>(egptr()) egptr() != rhs_a.egptr()</tt>
27684 </del></li>
27685 <li><del>
27686 if <tt>(pbase()) pbase() != rhs_a.pbase()</tt>
27687 </del></li>
27688 <li><del>
27689 if <tt>(pptr()) pptr() != rhs_a.pptr()</tt>
27690 </del></li>
27691 <li><del>
27692 if <tt>(epptr()) epptr() != rhs_a.epptr()</tt>
27693 </del></li>
27694 </ul>
27695 </blockquote>
27696 </blockquote>
27697
27698
27699
27700
27701
27702
27703 <hr>
27704 <h3><a name="1252"></a>1252. <tt>wbuffer_convert::state_type</tt> inconsistency</h3>
27705 <p><b>Section:</b> 22.3.3.2.3 [conversions.buffer] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27706  <b>Submitter:</b> Bo Persson  <b>Opened:</b> 2009-10-21  <b>Last modified:</b> 2009-10-31</p>
27707 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27708 <p><b>Discussion:</b></p>
27709 <p>
27710 The synopisis for <tt>wbuffer_convert</tt> 22.3.3.2.3 [conversions.buffer]/2 contains
27711 </p>
27712
27713 <blockquote><pre>typedef typename Tr::state_type   state_type;
27714 </pre></blockquote>
27715
27716 <p>
27717 making <tt>state_type</tt> a synonym for (possibly) some
27718 <tt>char_traits&lt;x&gt;::state_type</tt>.
27719 </p>
27720
27721 <p>
27722 However, in paragraph 9 of the same section, we have
27723 </p>
27724
27725 <blockquote><pre>typedef typename Codecvt::state_type state_type;
27726 </pre>
27727
27728 <blockquote>
27729 The type shall be a synonym for <tt>Codecvt::state_type</tt>.
27730 </blockquote>
27731 </blockquote>
27732
27733 <p>
27734 From what I can see, it might be hard to implement
27735 <tt>wbuffer_convert</tt> if the types were not both
27736 <tt>std::mbstate_t</tt>, but I cannot find a requirement that they must
27737 be the same type.
27738 </p>
27739
27740
27741 <p><b>Proposed resolution:</b></p>
27742
27743
27744
27745
27746
27747 <hr>
27748 <h3><a name="1253"></a>1253. invalidation of iterators and <tt>emplace</tt> vs. <tt>insert</tt> inconsistence in assoc. containers</h3>
27749 <p><b>Section:</b> 23.2.4 [associative.reqmts] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27750  <b>Submitter:</b> Boris Du&#353;ek <b>Opened:</b> 2009-10-24  <b>Last modified:</b> 2009-10-31</p>
27751 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#associative.reqmts">active issues</a> in [associative.reqmts].</p>
27752 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#associative.reqmts">issues</a> in [associative.reqmts].</p>
27753 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27754 <p><b>Discussion:</b></p>
27755 <p>
27756 In the latest published draft
27757 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2960.pdf">N2960</a>,
27758 section 23.2.4 [associative.reqmts], paragraph 8, it is specifies
27759 that that <tt>insert</tt> does not invalidate any iterators. As per
27760 23.2.1 [container.requirements.general], paragraph 12, this holds
27761 true not only for <tt>insert</tt>, but <tt>emplace</tt> as well. This
27762 gives the <tt>insert</tt> member a special treatment w.r.t.
27763 <tt>emplace</tt> member in 23.2.4 [associative.reqmts], par. 8,
27764 since both modify the container. For the sake of consistency, in 23.2.4 [associative.reqmts], par. 8: either reference to
27765 <tt>insert</tt> should be removed (i.e. count on 23.2.1 [container.requirements.general], par. 12), or reference to
27766 <tt>emplace</tt> be added (i.e. mention all members of assoc. containers
27767 that modify it).
27768 </p>
27769
27770
27771 <p><b>Proposed resolution:</b></p>
27772 <p>
27773 </p>
27774
27775
27776
27777
27778
27779 <hr>
27780 <h3><a name="1254"></a>1254. Misleading sentence in <tt>vector&lt;bool&gt;::flip</tt></h3>
27781 <p><b>Section:</b> 23.3.7 [vector.bool] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27782  <b>Submitter:</b> Christopher Jefferson <b>Opened:</b> 2009-11-01  <b>Last modified:</b> 2009-11-01</p>
27783 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#vector.bool">issues</a> in [vector.bool].</p>
27784 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27785 <p><b>Discussion:</b></p>
27786 <p>
27787 The effects of <tt>vector&lt;bool&gt;::flip</tt> has the line:
27788 </p>
27789
27790 <blockquote>
27791 It is unspecified whether the function has any effect on allocated but
27792 unused bits.
27793 </blockquote>
27794
27795 <p>
27796 While this is technically true, it is misleading, as any member function
27797 in any standard container may change unused but allocated memory. Users
27798 can never observe such changes as it would also be undefined behaviour
27799 to read such memory.
27800 </p>
27801
27802
27803 <p><b>Proposed resolution:</b></p>
27804 <p>
27805 Strike second sentence from the definition of <tt>vector&lt;bool&gt;::flip()</tt>,
27806 23.3.7 [vector.bool], paragraph 5.
27807 </p>
27808
27809 <blockquote>
27810 <i>Effects:</i> Replaces each element in the container with its complement.
27811 <del>It is unspecified whether the function has any effect on allocated
27812 but unused bits.</del>
27813 </blockquote>
27814
27815
27816
27817
27818
27819 <hr>
27820 <h3><a name="1255"></a>1255. <tt>declval</tt> should be added to the library</h3>
27821 <p><b>Section:</b> 20.3 [utility] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
27822  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-11-03  <b>Last modified:</b> 2009-11-04</p>
27823 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
27824 <p><b>Discussion:</b></p>
27825 <p>
27826 During the Santa Cruz meeting it was decided to split off the provision
27827 of the library utility <tt>value()</tt> proposed in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2979.html">N2979</a>
27828 from the concrete request of the
27829 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2939.html#uk300">UK 300</a>
27830 comment.
27831 The provision of a new library component that allows the production of
27832 values in unevaluated expressions is considered as important
27833 to realize constrained templates in C++0x where concepts are not
27834 available.
27835 </p>
27836
27837 <p>
27838 The following proposed resolution is an improvement over that suggested in
27839 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2958.html">N2958</a>,
27840 because the proposed component can now be defined without loss of
27841 general usefulness and any <i>use</i> by user-code will make the program ill-formed.
27842 A possible prototype implementation that satisfies the core language
27843 requirements
27844 can be written as:
27845 </p>
27846
27847 <blockquote><pre>template&lt;class T&gt;
27848   struct declval_protector {
27849     static const bool stop = false;
27850     static typename std::add_rvalue_reference&lt;T&gt;::type delegate(); <font color="#c80000">// undefined</font>
27851   };
27852
27853 template&lt;class T&gt;
27854 typename std::add_rvalue_reference&lt;T&gt;::type declval() {
27855   static_assert(declval_protector&lt;T&gt;::stop, "declval() must not be used!");
27856   return declval_protector&lt;T&gt;::delegate();
27857 }
27858 </pre></blockquote>
27859
27860 <p>
27861 Further-on the earlier suggested name <tt>value()</tt> has been changed to <tt>declval()</tt>
27862 after discussions with committee members.
27863 </p>
27864
27865 <p>
27866 Finally the suggestion shown below demonstrates that it can simplify
27867 existing standard wording by directly using it in the library
27868 specification, and that it also improves an overlooked corner case for
27869 <tt>common_type</tt> by adding support for <tt>cv void</tt>.
27870 </p>
27871
27872
27873 <p><b>Proposed resolution:</b></p>
27874 <p><i>[
27875 The following edit assumes that the earlier component identity
27876 has been removed as part of applying the solution of <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#939">939</a>,
27877 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2951.html">N2951</a>,
27878 and
27879 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2984.html">N2984</a>.
27880 Note that the proposal does not depend on this application,
27881 but it just simplifies the editorial representation
27882 ]</i></p>
27883
27884
27885 <ol>
27886 <li>
27887 <p>
27888 Change 20.3 [utility], header <tt>&lt;utility&gt;</tt> synopsis
27889 as indicated:
27890 </p>
27891
27892 <blockquote><pre>// 20.3.2, forward/move:
27893 template &lt;class T, class U&gt; T&amp;&amp; forward(U&amp;&amp; u);;
27894 template &lt;class T&gt; typename remove_reference&lt;T&gt;::type&amp;&amp; move(T&amp;&amp;);
27895
27896 <ins>// 20.3.3, declval:</ins>
27897 <ins>template &lt;class T&gt; typename add_rvalue_reference&lt;T&gt;::type declval(); // for unused context</ins>
27898 </pre></blockquote>
27899 </li>
27900
27901 <li>
27902 <p>
27903 Immediately after the current section 20.3.3 [forward] insert a
27904 new section:
27905 </p>
27906 <p>
27907 <ins>20.3.3 Function template declval [declval]</ins>
27908 </p>
27909 <p>
27910 <ins>The library provides the function template <tt>declval</tt> to simplify
27911 the definition of expressions in
27912 unevaluated and unused contexts (3.2 [basic.def.odr], 5 [expr]). The
27913 template parameter <tt>T</tt> of <tt>declval</tt> may
27914 be an incomplete type.</ins>
27915 </p>
27916
27917 <pre><ins>template &lt;class T&gt; typename add_rvalue_reference&lt;T&gt;::type declval(); // for unused context</ins>
27918 </pre>
27919
27920 <blockquote>
27921 <p>
27922 <ins><i>Remarks:</i> If this function is used according to 3.2 [basic.def.odr],
27923 the program shall be ill-formed.</ins>
27924 </p>
27925
27926 <p>
27927 <ins>[<i>Example:</i></ins>
27928 </p>
27929
27930 <blockquote><pre><ins>
27931 template&lt;class To, class From&gt;
27932 decltype(static_cast&lt;To&gt;(declval&lt;From&gt;())) convert(From&amp;&amp;);
27933 </ins></pre></blockquote>
27934
27935 <p>
27936 <ins>
27937 declares a function template <tt>convert</tt>, which does only participate in
27938 overloading, if the type <tt>From</tt> can be
27939 explicitly casted to type <tt>To</tt> &#8212; <i>end example</i>]</ins>
27940 </p>
27941 </blockquote>
27942
27943 </li>
27944
27945 <li>
27946 <p>
27947 This bullet just makes clear that after applying <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2984.html">N2984</a>, the changes in 20.6.4.3 [meta.unary.prop], before
27948 table Type property queries should <em>not</em> use <tt>declval</tt>,
27949 because the well-formedness requirement of the specification of
27950 <tt>is_constructible</tt> would become more complicated, because we
27951 would need to make sure that the expression <i>CE</i> is checked in an
27952 unevaluated context.
27953 </p>
27954 </li>
27955
27956 <li>
27957 <p>
27958 Also 20.6.5 [meta.rel]/4 is not modified similar to the previous bullet,
27959 because with
27960 the stricter requirements of not using <tt>declval()</tt> the well-formedness condition
27961 would be harder to specify. The following changes are only editorial ones (e.g.
27962 the removal of the duplicate declaration of <tt>create()</tt>):
27963 </p>
27964
27965 <blockquote>
27966 <p>
27967 Given the following function prototype:
27968 </p>
27969
27970 <blockquote><pre>template &lt;class T&gt;
27971   typename add_rvalue_reference&lt;T&gt;::type create();
27972 </pre></blockquote>
27973
27974 <p>
27975 the predicate condition for a template specialization
27976 <tt>is_convertible&lt;From, To&gt;</tt> shall be satisfied if and only
27977 if the return expression in the following code would be well-formed,
27978 including any
27979 im<del>m</del>plicit conversions to the return type of the function:
27980 </p>
27981
27982 <blockquote><pre><del>template &lt;class T&gt;
27983 typename add_rvalue_reference&lt;T&gt;::type create();</del>
27984 To test() {
27985   return create&lt;From&gt;();
27986 }
27987 </pre></blockquote>
27988 </blockquote>
27989 </li>
27990
27991 <li>
27992 <p>
27993 Change the entry in column "Comments" for <tt>common_type</tt> in Table 51 &#8212;
27994 Other transformations (20.6.7 [meta.trans.other]):
27995 </p>
27996
27997 <p><i>[
27998 NB: This wording change extends the type domain of <tt>common_type</tt> for <tt>cv
27999 void =&gt; cv void</tt> transformations and thus makes <tt>common_type</tt> usable for
28000 all binary type combinations that are supported by <tt>is_convertible</tt>
28001 ]</i></p>
28002
28003
28004 <blockquote>
28005 The member typedef <tt>type</tt> shall be defined as set out below. All
28006 types in the parameter pack <tt>T</tt> shall be complete <ins>or
28007 (possibly cv-qualified) <tt>void</tt></ins>. A program may specialize
28008 this trait if at least one template parameter in the specialization is a
28009 user-defined type. [<i>Note:</i> Such specializations are needed when
28010 only explicit conversions are desired among the template arguments.
28011 &#8212; <i>end note</i>]
28012 </blockquote>
28013 </li>
28014
28015 <li>
28016 <p>
28017 Change 20.6.7 [meta.trans.other]/3 as indicated:
28018 </p>
28019
28020 <p><i>[
28021 NB: This wording change is more than an editorial simplification of
28022 the definition of <tt>common_type</tt>: It also extends its usefulness for <tt>cv
28023 void</tt> types as outlined above
28024 ]</i></p>
28025
28026
28027 <blockquote>
28028 <p>
28029 The nested typedef <tt>common_type::type</tt> shall be defined as follows:
28030 </p>
28031
28032 <blockquote>
28033 <p>
28034 [..]
28035 </p>
28036 <pre>template &lt;class T, class U&gt;
28037 struct common_type&lt;T, U&gt; {
28038 <del>private:
28039   static T&amp;&amp; __t();
28040   static U&amp;&amp; __u();
28041 public:</del>
28042   typedef decltype(true ? <del>__t</del><ins>declval&lt;T&gt;</ins>() : <del>__u</del><ins>declval&lt;U&gt;</ins>()) type;
28043 };
28044 </pre>
28045 </blockquote>
28046 </blockquote>
28047 </li>
28048 </ol>
28049
28050
28051
28052
28053
28054
28055 <hr>
28056 <h3><a name="1256"></a>1256. <tt>weak_ptr</tt> comparison functions should be removed</h3>
28057 <p><b>Section:</b> 20.8.15.3 [util.smartptr.weak] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
28058  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-11-04  <b>Last modified:</b> 2009-11-04</p>
28059 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
28060 <p><b>Discussion:</b></p>
28061 <p>
28062 Additional to the necessary cleanup of the description of the the
28063 <tt>weak_ptr</tt> component from 20.8.15.3 [util.smartptr.weak]
28064 described in <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#1231">1231</a> it turns out that the currently deleted
28065 comparison functions of <tt>weak_ptr</tt> are not needed at all: There
28066 is no safe-bool conversion from <tt>weak_ptr</tt>, and it won't silently
28067 chose a conversion to <tt>shared_ptr</tt>.
28068 </p>
28069
28070
28071 <p><b>Proposed resolution:</b></p>
28072 <p>
28073 Change 20.8.15.3 [util.smartptr.weak]/1 as indicated:
28074 </p>
28075
28076 <blockquote><pre>namespace std {
28077 template&lt;class T&gt; class weak_ptr {
28078 public:
28079 ...
28080   <del>// comparisons</del>
28081   <del>template&lt;class Y&gt; bool operator&lt;(weak_ptr&lt;Y&gt; const&amp;) const = delete;</del>
28082   <del>template&lt;class Y&gt; bool operator&lt;=(weak_ptr&lt;Y&gt; const&amp;) const = delete;</del>
28083   <del>template&lt;class Y&gt; bool operator&gt;(weak_ptr&lt;Y&gt; const&amp;) const = delete;</del>
28084   <del>template&lt;class Y&gt; bool operator&gt;=(weak_ptr&lt;Y&gt; const&amp;) const = delete;</del>
28085 };
28086 ...
28087 </pre></blockquote>
28088
28089
28090
28091
28092
28093 <hr>
28094 <h3><a name="1257"></a>1257. Header &lt;ios&gt; still contains a <code>concept_map</code></h3>
28095 <p><b>Section:</b> 27.5 [iostreams.base] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
28096  <b>Submitter:</b> Beman Dawes <b>Opened:</b> 2009-11-04  <b>Last modified:</b> 2009-11-04</p>
28097 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#iostreams.base">issues</a> in [iostreams.base].</p>
28098 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
28099 <p><b>Discussion:</b></p>
28100 <p>
28101 The current WP still contains a <tt>concept_map</tt>.
28102 </p>
28103
28104
28105 <p><b>Proposed resolution:</b></p>
28106 <p>
28107 Change Iostreams base classes 27.5 [iostreams.base], Header &lt;ios&gt; synopsis,
28108 as indicated:
28109 </p>
28110
28111 <blockquote><pre><del>concept_map ErrorCodeEnum&lt;io_errc&gt; { };</del>
28112 <ins>template &lt;&gt; struct is_error_code_enum&lt;io_errc&gt; : true_type { }</ins>
28113 error_code make_error_code(io_errc e);
28114 error_condition make_error_condition(io_errc e);
28115 const error_category&amp; iostream_category();
28116 </pre></blockquote>
28117
28118
28119
28120
28121
28122
28123 <hr>
28124 <h3><a name="1258"></a>1258. std::function Effects clause impossible to satisfy</h3>
28125 <p><b>Section:</b> 20.7.15.2.2 [func.wrap.func.mod] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
28126  <b>Submitter:</b> Daniel Krügler <b>Opened:</b> 2009-11-05  <b>Last modified:</b> 2009-11-05</p>
28127 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
28128 <p><b>Discussion:</b></p>
28129 <p>
28130 As of 20.7.15.2.2 [func.wrap.func.mod]/2+ we have the following
28131 prototype description:
28132 </p>
28133
28134 <blockquote><pre>template&lt;class F, Allocator Alloc&gt;
28135   requires Callable&lt;F, ArgTypes...&gt;
28136     &amp;&amp; Convertible&lt;Callable&lt;F, ArgTypes...&gt;::result_type, R&gt;
28137   void assign(F, const Alloc&amp;);
28138 </pre>
28139 <blockquote>
28140 <i>Effects:</i> <tt>function(f, a).swap(*this)</tt>
28141 </blockquote>
28142 </blockquote>
28143
28144 <p>
28145 Two things: First the concept debris needs to be removed, second and
28146 much more importantly, the effects clause is now impossible to satisfy,
28147 because there is no constructor that would match the parameter sequence
28148 (<tt>FunctionObject</tt>, <tt>Allocator</tt>) [plus the fact that no
28149 <tt>f</tt> and no <tt>a</tt> is part of the signature]. The most
28150 probable candidate is
28151 </p>
28152
28153 <blockquote><pre>template&lt;class F, class A&gt; function(allocator_arg_t, const A&amp;, F);
28154 </pre></blockquote>
28155
28156 <p>
28157 and the effects clause needs to be adapted to use this signature.
28158 </p>
28159
28160
28161 <p><b>Proposed resolution:</b></p>
28162 <p>
28163 Change in 20.7.15.2.2 [func.wrap.func.mod] the complete prototype description as
28164 indicated
28165 </p>
28166 <p><i>[
28167 Question to
28168 the editor: Shouldn't there a paragraph number in front of the Effects clause?
28169 ]</i></p>
28170
28171
28172 <blockquote><pre><del>template&lt;class F, Allocator Alloc&gt;
28173   requires Callable&lt;F, ArgTypes...&gt;
28174     &amp;&amp; Convertible&lt;Callable&lt;F, ArgTypes...&gt;::result_type, R&gt;
28175   void assign(F, const Alloc&amp;);</del>
28176 <ins>template&lt;class F, class A&gt; void assign(F f, const A&amp; a);</ins>
28177 </pre>
28178 <blockquote>
28179 <ins>3</ins> <i>Effects:</i> <tt>function(<del>f, a</del><ins>allocator_arg, a,
28180 f</ins>).swap(*this)</tt>
28181 </blockquote>
28182 </blockquote>
28183
28184
28185
28186
28187
28188 <hr>
28189 <h3><a name="1259"></a>1259. Should initializer-list constructors move elements?</h3>
28190 <p><b>Section:</b> 23.2.3 [sequence.reqmts] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
28191  <b>Submitter:</b> Sean Hunt <b>Opened:</b> 2009-11-05  <b>Last modified:</b> 2009-11-06</p>
28192 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#sequence.reqmts">active issues</a> in [sequence.reqmts].</p>
28193 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#sequence.reqmts">issues</a> in [sequence.reqmts].</p>
28194 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
28195 <p><b>Discussion:</b></p>
28196 <p>
28197 According to 23.2.3 [sequence.reqmts], <tt>X(il)</tt> is
28198 equivalent to <tt>X(il.begin(), il.end())</tt>. Should it instead be
28199 equivalent to <tt>X(move_iterator(il.begin()),
28200 move_iterator(il.end()))</tt> so that needless copies are not made? This
28201 doesn't seem ideal either - it may make more sense to provide two
28202 overloads for the constructor, one for move and one for copy.
28203 </p>
28204
28205
28206 <p><b>Proposed resolution:</b></p>
28207 <p>
28208 </p>
28209
28210
28211
28212
28213
28214 <hr>
28215 <h3><a name="1260"></a>1260. <tt>is_constructible&lt;int*,void*&gt;</tt> reports true</h3>
28216 <p><b>Section:</b> 20.6.4.3 [meta.unary.prop] <b>Status:</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html#New">New</a>
28217  <b>Submitter:</b> Peter Dimov <b>Opened:</b> 2009-11-07  <b>Last modified:</b> 2009-11-08</p>
28218 <p><b>View other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index-open.html#meta.unary.prop">active issues</a> in [meta.unary.prop].</p>
28219 <p><b>View all other</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-index.html#meta.unary.prop">issues</a> in [meta.unary.prop].</p>
28220 <p><b>View all issues with</b> <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-status.html#New">New</a> status.</p>
28221 <p><b>Discussion:</b></p>
28222 <p>
28223 The specification of <tt>is_constructible&lt;T,Args...&gt;</tt> in
28224 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n3000.pdf">N3000</a>
28225 uses
28226 </p>
28227
28228 <blockquote><pre>static_cast&lt;T&gt;(create&lt;Args&gt;()...)
28229 </pre></blockquote>
28230
28231 <p>
28232 for the one-argument case, but <tt>static_cast</tt> also permits
28233 unwanted conversions such as <tt>void*</tt> to <tt>T*</tt> and
28234 <tt>Base*</tt> to <tt>Derived*</tt>.
28235 </p>
28236
28237
28238 <p><b>Proposed resolution:</b></p>
28239 <p>
28240 Change 20.6.4.3 [meta.unary.prop], p6:
28241 </p>
28242
28243 <blockquote>
28244 <p>
28245 the predicate condition for a template specialization
28246 <tt>is_constructible&lt;T, Args&gt;</tt> shall be satisfied, if and only
28247 if the following <del>expression <i>CE</i></del> <ins>variable
28248 definition</ins> would be well-formed:
28249 </p>
28250
28251 <ul>
28252 <li>
28253 <p>
28254 if <tt>sizeof...(Args) == <ins>0</ins> <del>1</del></tt><del>, the expression</del>:
28255 </p>
28256 <blockquote><pre><del>static_cast&lt;T&gt;(create&lt;Args&gt;()...)</del>
28257 <ins>T t;</ins>
28258 </pre></blockquote>
28259 </li>
28260 <li>
28261 <p>
28262 otherwise <del>the expression</del>:
28263 </p>
28264 <blockquote><pre>T<ins> t</ins>(create&lt;Args&gt;()...);
28265 </pre></blockquote>
28266 </li>
28267 </ul>
28268 </blockquote>
28269
28270
28271
28272
28273
28274 </body></html>