Document number: | PL22.16 09/0053 = WG21 N2863 |
Date: | 2009-03-23 |
Reply to: |
William M. Miller Edison Design Group, Inc. wmm@edg.com |
This document summarizes the progress of WG21 in addressing National Body comments on Committee Draft document N2800. In total, 591 comments were received. To date, 216 have been accepted or tentatively accepted as proposed, 27 have been accepted or tentatively accepted with modifications, 99 have been rejected, and 249 remain to be addressed. The detailed breakdown:
Unresolved | Accepted | Modified | Rejected | Total | |
---|---|---|---|---|---|
CWG | 93 | 32 | 8 | 43 | 176 |
LWG | 128 | 36 | 5 | 35 | 204 |
editor | 28 | 148 | 14 | 21 | 211 |
This document consists of three tables. The first is a table of contents, listing each issue by National Body and comment number and giving its current status. The second is a list of all as-yet unresolved comments, sorted by the responsible party. The third is a comprehensive listing of all the comments received in the same order and with roughly the same organization as in document N2837; the “Disposition” field gives the response, as well as any explanatory comments provided by the responsible party. (The internal formatting of the text of the national body comments will be improved in future revisions of this document.)
In the following table, the “Date” field represents the date on which the Committee affirmed the listed disposition.
ID | Responsible | Disposition | Date | Issue | Document |
---|---|---|---|---|---|
CA 1 | CWG | accepted | |||
CA 2 | LWG | accepted | |||
CH 1 | CWG | 794 | |||
CH 2 | CWG | accepted | |||
DE 1 | CWG | accepted | |||
DE 2 | LWG | ||||
DE 3 | CWG | 735 | |||
DE 4 | CWG | 693 | |||
DE 5 | CWG | 731 | |||
DE 6 | CWG | modified | 2009-03-06 | N2845 | |
DE 7 | editor | rejected | |||
DE 8 | CWG | accepted | 2009-03-06 | N2841 | |
DE 9 | CWG | 720 | |||
DE 10 | CWG | 731 | |||
DE 11 | CWG | accepted | 2009-03-06 | N2841 | |
DE 12 | CWG | 731 | |||
DE 13 | CWG | 732 | |||
DE 14 | CWG | 730 | |||
DE 15 | CWG | ||||
DE 16 | LWG | 902 | |||
DE 17 | LWG | 1078 | |||
DE 18 | LWG | ||||
DE 20 | editor | accepted | |||
DE 21 | LWG | accepted | 817 | ||
DE 22 | LWG | accepted | 1023 | ||
DE 23 | CWG | 699 | |||
DE 24 | LWG | 922 | |||
DE 25 | CWG | 831 | |||
FI 1 | CWG | ||||
FI 2 | CWG | rejected | |||
FI 3 | CWG | rejected | |||
FI 4 | editor | accepted | |||
FI 5 | editor | accepted | |||
FI 6 | LWG | ||||
FI 7 | editor | accepted | |||
FR 1 | CWG | ||||
FR 2 | editor | ||||
FR 3 | editor | ||||
FR 4 | editor | ||||
FR 5 | CWG | 690 | |||
FR 6 | editor | ||||
FR 7 | editor | rejected | |||
FR 8 | editor | rejected | |||
FR 9 | editor | accepted | |||
FR 10 | CWG | 788 | |||
FR 11 | editor | accepted | |||
FR 12 | CWG | rejected | |||
FR 13 | editor | modified | |||
FR 14 | editor | accepted | |||
FR 15 | editor | accepted | |||
FR 16 | CWG | 481 | |||
FR 17 | CWG | 791 | |||
FR 18 | editor | accepted | |||
FR 19 | CWG | rejected | |||
FR 20 | CWG | rejected | |||
FR 21 | CWG | rejected | |||
FR 22 | CWG | rejected | |||
FR 23 | CWG | rejected | |||
FR 24 | CWG | rejected | |||
FR 25 | CWG | rejected | |||
FR 26 | CWG | ||||
FR 27 | CWG | rejected | |||
FR 28 | CWG | accepted | 2009-03-06 | N2841 | |
FR 29 | CWG | 823 | |||
FR 30 | CWG | rejected | |||
FR 31 | CWG | ||||
FR 32 | LWG | 902 | |||
FR 33 | editor | accepted | |||
FR 34 | editor | accepted | |||
FR 35 | LWG | ||||
FR 36 | editor | accepted | |||
FR 37 | editor | accepted | |||
FR 38 | LWG | accepted | |||
FR 39 | editor | ||||
JP 1 | editor | accepted | |||
JP 2 | editor | accepted | |||
JP 3 | editor | accepted | |||
JP 4 | editor | accepted | |||
JP 5 | CWG | 790 | |||
JP 6 | editor | accepted | |||
JP 7 | editor | accepted | |||
JP 8 | CWG | 743 | |||
JP 9 | CWG | rejected | |||
JP 10 | CWG | rejected | |||
JP 11 | CWG | rejected | |||
JP 12 | CWG | 699 | |||
JP 13 | editor | accepted | |||
JP 14 | CWG | 812 | |||
JP 15 | editor | rejected | |||
JP 16 | editor | accepted | |||
JP 17 | CWG | rejected | |||
JP 18 | CWG | rejected | |||
JP 19 | editor | accepted | |||
JP 20 | CWG | 825 | |||
JP 21 | LWG | rejected | |||
JP 22 | editor | rejected | |||
JP 23 | LWG | 1003 | |||
JP 24 | editor | accepted | |||
JP 25 | editor | accepted | |||
JP 26 | LWG | 1005 | |||
JP 27 | LWG | ||||
JP 28 | LWG | rejected | |||
JP 29 | LWG | 1007 | |||
JP 30 | LWG | ||||
JP 31 | LWG | 1008 | |||
JP 32 | LWG | rejected | |||
JP 33 | LWG | 1016 | |||
JP 34 | editor | accepted | |||
JP 35 | editor | accepted | |||
JP 36 | editor | accepted | |||
JP 37 | editor | accepted | |||
JP 38 | LWG | 817 | |||
JP 39 | LWG | modified | 1024 | ||
JP 40 | editor | accepted | |||
JP 41 | LWG | rejected | |||
JP 42 | editor | accepted | |||
JP 43 | editor | accepted | |||
JP 44 | LWG | accepted | 1030 | ||
JP 45 | LWG | 1032 | |||
JP 46 | LWG | 1081 | |||
JP 47 | editor | accepted | |||
JP 48 | LWG | 1081 | |||
JP 49 | LWG | 1082 | |||
JP 50 | LWG | 991 | |||
JP 51 | editor | accepted | |||
JP 52 | LWG | 1083 | |||
JP 53 | LWG | 1083 | |||
JP 54 | editor | accepted | |||
JP 55 | editor | accepted | |||
JP 56 | editor | accepted | |||
JP 57 | editor | accepted | |||
JP 58 | editor | accepted | |||
JP 59 | editor | accepted | |||
JP 60 | LWG | rejected | |||
JP 61 | editor | accepted | |||
JP 62 | LWG | rejected | |||
JP 63 | LWG | accepted | 2009-03-06 | 874 | N2836 |
JP 64 | LWG | ||||
JP 65 | LWG | ||||
JP 66 | LWG | ||||
JP 67 | LWG | ||||
JP 68 | LWG | ||||
JP 69 | LWG | ||||
JP 71 | editor | accepted | |||
JP 72 | LWG | ||||
JP 73 | LWG | ||||
JP 74 | LWG | 1014 | |||
JP 75 | LWG | rejected | |||
JP 76 | LWG | 1089 | |||
JP 77 | LWG | ||||
JP 78 | editor | accepted | |||
JP 79 | LWG | ||||
JP 80 | editor | accepted | |||
JP 81 | LWG | ||||
UK 1 | editor | rejected | |||
UK 2 | editor | rejected | |||
UK 3 | CWG | 783 | |||
UK 4 | CWG | rejected | |||
UK 5 | editor | ||||
UK 6 | CWG | 784 | |||
UK 7 | CWG | 785 | |||
UK 8 | CWG | 785 | |||
UK 9 | CWG | 787 | |||
UK 10 | CWG | accepted | 2009-03-06 | N2841 | |
UK 11 | CWG | 789 | |||
UK 12 | CWG | 787 | |||
UK 13 | CWG | 832 | |||
UK 14 | editor | accepted | |||
UK 15 | CWG | rejected | |||
UK 16 | editor | accepted | |||
UK 17 | CWG | rejected | |||
UK 18 | CWG | rejected | |||
UK 19 | CWG | rejected | |||
UK 20 | editor | ||||
UK 21 | editor | ||||
UK 22 | CWG | 633 | |||
UK 23 | CWG | 758 | |||
UK 24 | CWG | rejected | |||
UK 25 | CWG | accepted | 2009-03-06 | N2841 | |
UK 26 | CWG | 570 | |||
UK 27 | CWG | accepted | 2009-03-06 | N2841 | |
UK 28 | CWG | accepted | 2009-03-06 | N2841 | |
UK 29 | CWG | rejected | |||
UK 30 | CWG | 791 | |||
UK 31 | CWG | accepted | 2009-03-06 | N2841 | |
UK 32 | CWG | accepted | 2009-03-06 | N2841 | |
UK 33 | CWG | accepted | 2009-03-06 | N2841 | |
UK 34 | editor | accepted | |||
UK 35 | editor | accepted | |||
UK 36 | CWG | accepted | 2009-03-06 | N2841 | |
UK 37 | CWG | rejected | |||
UK 38 | CWG | rejected | |||
UK 39 | CWG | 759 | |||
UK 40 | CWG | 796 | |||
UK 41 | CWG | modified | 2009-03-06 | N2845 | |
UK 42 | CWG | 797 | |||
UK 43 | CWG | modified | 2009-03-06 | N2845 | |
UK 44 | CWG | rejected | |||
UK 45 | CWG | 795 | |||
UK 46 | CWG | modified | 2009-03-06 | N2845 | |
UK 47 | CWG | accepted | 2009-03-06 | N2841 | |
UK 48 | editor | accepted | |||
UK 49 | CWG | 699 | |||
UK 50 | CWG | 806 | |||
UK 51 | CWG | 807 | |||
UK 52 | editor | modified | |||
UK 53 | CWG | 798 | |||
UK 54 | CWG | accepted | 2009-03-06 | N2841 | |
UK 55 | CWG | 2009-03-06 | 799 | N2841 | |
UK 56 | CWG | modified | 2009-03-06 | N2841 | |
UK 57 | CWG | 800 | |||
UK 58 | CWG | 801 | |||
UK 59 | CWG | accepted | 2009-03-06 | N2841 | |
UK 60 | editor | accepted | |||
UK 61 | editor | accepted | |||
UK 62 | CWG | accepted | 2009-03-06 | N2841 | |
UK 63 | editor | accepted | |||
UK 64 | editor | accepted | |||
UK 65 | CWG | modified | 2009-03-06 | N2841 | |
UK 66 | CWG | accepted | 2009-03-06 | N2841 | |
UK 67 | CWG | accepted | 2009-03-06 | N2841 | |
UK 68 | CWG | accepted | 2009-03-06 | N2841 | |
UK 69 | CWG | 802 | |||
UK 70 | CWG | 803 | |||
UK 71 | CWG | 804 | |||
UK 72 | CWG | 805 | |||
UK 73 | CWG | rejected | |||
UK 74 | CWG | rejected | |||
UK 75 | editor | accepted | |||
UK 76 | CWG | rejected | |||
UK 77 | editor | accepted | |||
UK 78 | LWG | 1001 | |||
UK 79 | CWG | rejected | |||
UK 80 | editor | modified | |||
UK 81 | CWG | rejected | |||
UK 82 | CWG | rejected | |||
UK 83 | CWG | 808 | |||
UK 84 | CWG | accepted | 2009-03-06 | N2841 | |
UK 85 | CWG | ||||
UK 86 | CWG | 809 | |||
UK 87 | CWG | 810 | |||
UK 89 | CWG | 811 | |||
UK 90 | editor | accepted | |||
UK 91 | CWG | modified | 2009-03-06 | N2841 | |
UK 92 | editor | accepted | |||
UK 93 | editor | accepted | |||
UK 94 | editor | accepted | |||
UK 95 | CWG | 746 | |||
UK 96 | CWG | 628 | |||
UK 97 | editor | accepted | |||
UK 98 | LWG | 1055 | |||
UK 99 | CWG | rejected | |||
UK 100 | CWG | rejected | |||
UK 101 | CWG | 813 | |||
UK 102 | editor | modified | |||
UK 103 | CWG | 815 | |||
UK 104 | editor | rejected | |||
UK 105 | CWG | rejected | |||
UK 106 | editor | accepted | |||
UK 107 | editor | accepted | |||
UK 108 | CWG | 816 | |||
UK 109 | editor | accepted | |||
UK 110 | CWG | accepted | 2009-03-06 | N2841 | |
UK 111 | editor | accepted | |||
UK 112 | editor | rejected | |||
UK 113 | editor | accepted | |||
UK 114 | CWG | rejected | |||
UK 115 | CWG | 820 | |||
UK 116 | CWG | 821 | |||
UK 117 | CWG | 822 | |||
UK 118 | editor | accepted | |||
UK 119 | CWG | 791 | |||
UK 120 | editor | accepted | |||
UK 121 | editor | accepted | |||
UK 122 | CWG | ||||
UK 123 | CWG | ||||
UK 124 | editor | accepted | |||
UK 125 | CWG | ||||
UK 126 | editor | accepted | |||
UK 127 | editor | ||||
UK 128 | editor | accepted | |||
UK 129 | CWG | 826 | |||
UK 130 | CWG | 828 | |||
UK 131 | CWG | accepted | 2009-03-06 | N2844 | |
UK 132 | CWG | accepted | 2009-03-06 | N2841 | |
UK 133 | CWG | accepted | 2009-03-06 | N2841 | |
UK 134 | editor | accepted | |||
UK 135 | CWG | 829 | |||
UK 136 | CWG | 830 | |||
UK 137 | editor | accepted | |||
UK 138 | editor | rejected | |||
UK 139 | CWG | accepted | 2009-03-06 | N2841 | |
UK 140 | editor | accepted | |||
UK 141 | editor | accepted | |||
UK 142 | editor | accepted | |||
UK 143 | editor | accepted | |||
UK 144 | editor | accepted | |||
UK 145 | editor | accepted | |||
UK 146 | editor | accepted | |||
UK 147 | editor | accepted | |||
UK 148 | editor | ||||
UK 149 | editor | modified | |||
UK 150 | editor | ||||
UK 151 | editor | modified | |||
UK 152 | LWG | 1064 | |||
UK 153 | editor | accepted | |||
UK 154 | editor | rejected | |||
UK 155 | editor | accepted | |||
UK 156 | editor | accepted | |||
UK 157 | editor | accepted | |||
UK 158 | editor | accepted | |||
UK 159 | editor | accepted | |||
UK 160 | editor | ||||
UK 161 | editor | rejected | |||
UK 162 | editor | accepted | |||
UK 163 | LWG | 997 | |||
UK 164 | editor | ||||
UK 165 | LWG | ||||
UK 166 | editor | accepted | |||
UK 167 | editor | accepted | |||
UK 168 | LWG | 1065 | |||
UK 169 | LWG | 992 | |||
UK 170 | LWG | 1002 | |||
UK 171 | editor | accepted | |||
UK 172 | LWG | rejected | |||
UK 173 | LWG | ||||
UK 174 | editor | accepted | |||
UK 175 | LWG | accepted | |||
UK 176 | editor | modified | |||
UK 177 | editor | accepted | |||
UK 178 | editor | rejected | |||
UK 179 | LWG | accepted | 1004 | ||
UK 180 | LWG | rejected | |||
UK 181 | editor | accepted | |||
UK 182 | LWG | rejected | |||
UK 184 | editor | ||||
UK 185 | editor | accepted | |||
UK 186 | editor | accepted | |||
UK 187 | LWG | ||||
UK 188 | LWG | accepted | 993 | ||
UK 189 | LWG | accepted | 1066 | ||
UK 190 | LWG | accepted | 1006 | ||
UK 191 | editor | modified | |||
UK 192 | CWG | 805 | |||
UK 193 | LWG | 994 | |||
UK 194 | LWG | ||||
UK 195 | LWG | ||||
UK 196 | LWG | ||||
UK 197 | LWG | rejected | |||
UK 198 | editor | ||||
UK 199 | LWG | accepted | 1015 | ||
UK 200 | LWG | rejected | |||
UK 201 | LWG | rejected | |||
UK 202 | editor | accepted | |||
UK 203 | editor | accepted | |||
UK 204 | LWG | accepted | 1020 | ||
UK 205 | LWG | accepted | 1019 | ||
UK 206 | LWG | accepted | |||
UK 207 | editor | accepted | |||
UK 208 | LWG | rejected | |||
UK 209 | LWG | 1026 | |||
UK 210 | LWG | 1029 | |||
UK 211 | LWG | accepted | 1021 | ||
UK 212 | LWG | modified | |||
UK 213 | LWG | accepted | 1027 | ||
UK 214 | LWG | 1028 | |||
UK 215 | editor | accepted | |||
UK 216 | LWG | 1081 | |||
UK 217 | editor | accepted | |||
UK 218 | LWG | ||||
UK 219 | editor | accepted | |||
UK 220 | LWG | ||||
UK 221 | editor | accepted | |||
UK 222 | LWG | 1034 | |||
UK 223 | LWG | accepted | 2009-03-06 | N2840 | |
UK 224 | LWG | modified | 2009-03-06 | N2840 | |
UK 225 | editor | accepted | |||
UK 226 | LWG | 1035 | |||
UK 227 | editor | accepted | |||
UK 228 | editor | accepted | |||
UK 229 | editor | accepted | |||
UK 230 | LWG | rejected | |||
UK 231 | LWG | 1036 | |||
UK 232 | LWG | accepted | 1037 | ||
UK 233 | LWG | accepted | 1038 | ||
UK 234 | LWG | accepted | 1039 | ||
UK 235 | editor | accepted | |||
UK 236 | editor | modified | |||
UK 237 | editor | accepted | |||
UK 238 | LWG | modified | 1040 | ||
UK 239 | LWG | 1041 | |||
UK 240 | LWG | ||||
UK 241 | LWG | ||||
UK 242 | editor | accepted | |||
UK 243 | LWG | rejected | |||
UK 244 | LWG | 1042 | |||
UK 245 | LWG | rejected | |||
UK 246 | LWG | 1091 | |||
UK 247 | LWG | rejected | |||
UK 248 | editor | accepted | |||
UK 249 | LWG | rejected | |||
UK 250 | LWG | 1084 | |||
UK 251 | LWG | 1009 | |||
UK 252 | editor | accepted | |||
UK 253 | editor | accepted | |||
UK 254 | LWG | ||||
UK 255 | LWG | ||||
UK 256 | LWG | ||||
UK 257 | LWG | rejected | |||
UK 258 | LWG | 1085 | |||
UK 259 | LWG | rejected | |||
UK 260 | LWG | ||||
UK 261 | LWG | rejected | |||
UK 262 | LWG | ||||
UK 263 | LWG | 1010 | |||
UK 264 | LWG | ||||
UK 265 | LWG | 1079 | |||
UK 266 | LWG | ||||
UK 267 | LWG | ||||
UK 268 | editor | accepted | |||
UK 269 | editor | accepted | |||
UK 270 | LWG | 940 | |||
UK 271 | LWG | 1011 | |||
UK 272 | LWG | rejected | |||
UK 274 | editor | accepted | |||
UK 275 | LWG | rejected | |||
UK 276 | LWG | rejected | |||
UK 277 | LWG | 1012 | |||
UK 278 | LWG | rejected | |||
UK 279 | LWG | 1051 | |||
UK 280 | editor | accepted | |||
UK 281 | LWG | 1052 | |||
UK 282 | LWG | 901 | |||
UK 283 | LWG | rejected | |||
UK 284 | LWG | accepted | 1086 | ||
UK 285 | editor | ||||
UK 286 | editor | ||||
UK 287 | LWG | 788 | |||
UK 288 | editor | ||||
UK 289 | editor | ||||
UK 290 | editor | accepted | |||
UK 291 | LWG | rejected | |||
UK 292 | editor | ||||
UK 293 | editor | ||||
UK 294 | LWG | rejected | |||
UK 295 | LWG | 1053 | |||
UK 296 | LWG | rejected | |||
UK 297 | editor | accepted | |||
UK 298 | editor | accepted | |||
UK 299 | editor | ||||
UK 300 | LWG | rejected | |||
UK 301 | LWG | 1087 | |||
UK 302 | editor | rejected | |||
UK 303 | editor | rejected | |||
UK 304 | editor | rejected | |||
UK 305 | LWG | accepted | 1013 | ||
UK 306 | LWG | accepted | 2009-03-06 | N2836 | |
UK 307 | editor | accepted | |||
UK 308 | LWG | ||||
UK 309 | LWG | ||||
UK 310 | LWG | ||||
UK 311 | LWG | ||||
UK 312 | LWG | ||||
UK 313 | LWG | 926 | |||
UK 314 | LWG | ||||
UK 315 | editor | accepted | |||
UK 316 | LWG | accepted | 723 | ||
UK 317 | LWG | 1014 | |||
UK 318 | editor | accepted | |||
UK 319 | LWG | 909 | |||
UK 320 | LWG | ||||
UK 321 | editor | ||||
UK 322 | LWG | ||||
UK 323 | LWG | accepted | 929 | ||
UK 324 | LWG | rejected | 889 | ||
UK 325 | LWG | 1044 | |||
UK 326 | LWG | accepted | 1045 | ||
UK 327 | LWG | ||||
UK 328 | LWG | ||||
UK 329 | LWG | 1046 | |||
UK 330 | editor | accepted | |||
UK 331 | LWG | ||||
UK 332 | editor | ||||
UK 333 | LWG | ||||
UK 334 | LWG | accepted | 1047 | ||
UK 335 | LWG | 1048 | |||
UK 336 | LWG | ||||
UK 337 | LWG | ||||
UK 338 | LWG | ||||
UK 339 | LWG | accepted | 1049 | ||
UK 340 | LWG | accepted | 1050 | ||
UK 341 | LWG | ||||
UK 342 | LWG | 1088 | |||
UK 343 | LWG | ||||
UK 344 | CWG | rejected | |||
US 1 | CWG | accepted | |||
US 2 | LWG | accepted | |||
US 3 | editor | modified | |||
US 4 | editor | ||||
US 5 | editor | ||||
US 6 | CWG | rejected | |||
US 7 | editor | accepted | |||
US 8 | editor | accepted | |||
US 9 | editor | accepted | |||
US 10 | editor | accepted | |||
US 11 | editor | accepted | |||
US 12 | CWG | modified | 2009-03-06 | N2841 | |
US 13 | CWG | rejected | |||
US 14 | CWG | accepted | 2009-03-06 | N2841 | |
US 15 | CWG | rejected | |||
US 16 | CWG | 785 | |||
US 17 | CWG | 786 | |||
US 18 | editor | accepted | |||
US 19 | editor | accepted | |||
US 20 | CWG | rejected | |||
US 21 | editor | rejected | |||
US 22 | editor | accepted | |||
US 23 | editor | accepted | |||
US 24 | CWG | 792 | |||
US 25 | editor | rejected | |||
US 26 | CWG | 793 | |||
US 27 | editor | accepted | |||
US 28 | editor | accepted | |||
US 29 | CWG | 762 | |||
US 30 | CWG | 762 | |||
US 31 | CWG | 752 | |||
US 32 | editor | accepted | |||
US 33 | CWG | 680 | |||
US 34 | editor | accepted | |||
US 35 | editor | accepted | |||
US 36 | CWG | 717 | |||
US 37 | editor | accepted | |||
US 38 | editor | rejected | |||
US 39 | CWG | rejected | |||
US 40 | CWG | 814 | |||
US 41 | CWG | ||||
US 42 | CWG | 817 | |||
US 43 | CWG | ||||
US 44 | editor | accepted | |||
US 45 | CWG | 818 | |||
US 46 | CWG | accepted | 2009-03-06 | N2844 | |
US 49 | editor | ||||
US 50 | CWG | 819 | |||
US 51 | editor | accepted | |||
US 52 | editor | accepted | |||
US 53 | CWG | 824 | |||
US 54 | CWG | accepted | |||
US 55 | editor | accepted | |||
US 56 | editor | rejected | |||
US 57 | editor | rejected | |||
US 58 | editor | accepted | |||
US 59 | CWG | ||||
US 60 | CWG | 827 | |||
US 61 | LWG | accepted | |||
US 62 | LWG | accepted | |||
US 63 | LWG | ||||
US 64 | editor | accepted | |||
US 65 | LWG | 1075 | |||
US 66 | LWG | 1017 | |||
US 67 | editor | modified | |||
US 68 | editor | accepted | |||
US 69 | editor | ||||
US 70 | LWG | 1018 | |||
US 71 | editor | modified | |||
US 72 | LWG | 817 | |||
US 73 | CWG | accepted | 2009-03-06 | 750 | N2845 |
US 74.1 | LWG | 2009-03-06 | 1075 | N2840 | |
US 74.2 | editor | modified | |||
US 75 | LWG | modified | 2009-03-06 | N2840 | |
US 77 | LWG | ||||
US 78 | LWG | 1031 | |||
US 79 | LWG | accepted | 932 | ||
US 80 | editor | modified | |||
US 81 | LWG | accepted | 934 | ||
US 82 | editor | accepted | |||
US 83 | editor | accepted | |||
US 84 | LWG | ||||
US 85 | LWG | ||||
US 86 | LWG | ||||
US 87 | LWG | ||||
US 88 | LWG | ||||
US 89 | LWG | 937 | |||
US 90 | LWG | 908 | |||
US 91 | LWG | accepted | 1043 | ||
US 92 | LWG | rejected | |||
US 93 | LWG | ||||
US 94 | editor | accepted | |||
US 95 | LWG | rejected | |||
US 96 | LWG | rejected | |||
US 97 | LWG | 2009-03-06 | N2802 | ||
US 98 | LWG |
The following table lists the comments for which no resolution has yet been proposed.
ID | Ref | Comment | Proposed Resolution | Owner | Issue | Disposition |
---|---|---|---|---|---|---|
FR 1 | General Comment | Interactions between several new features appear obscure, and very few examples are offered to guide understanding of the formal text on interaction between these new additions. We worry about the complexity of the programming model so created. | CWG | Need more information from NB, including a suggested resolution, in order to respond effectively to this comment. | ||
US 1 | 1-16 |
The
active issues identified in WG21 N2803, C++ Standard Core
Language Active Issues, must be addressed and appropriate
action taken.
http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_active.html |
Appropriate action would include making changes to the CD, identifying an issue as not requiring a change to the CD, or deferring the issue to a later point in time. | CWG | ACCEPTED | |
CA 1 | There are quite a number of defects for the current CD recorded in SC22/WG21-N2803 and N2806 | Consider these comments and update ISO/IEC CD 14882 accordingly | CWG | ACCEPTED | ||
DE 1 | 1 through 16 | DE-1 Consider addressing a significant part of the unresolved core language issues presented in WG21 document N2791 "C++ Standard Core Language Active Issues, Revision 59", available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2791.html . | CWG | ACCEPTED | ||
CH 2 | all | The issues on the issues lists shall be addressed before the standard becomes final. | CWG | ACCEPTED | ||
US 3 | all | Latin abbreviations are presented incorrectly. | Italicize all Latin abbreviations, append commas after each occurrence of i.e. and e.g., and remove extraneous space after each such abbreviation. | editor |
ACCEPTED with MODIFICATIONS
The "Chicago Manual of Style" says that i.e. and e.g. are never italicized and always followed by a comma, so that's how they now appear. |
|
FR 3 | 1 [intro.scope] ¶ 2 | C++ is split at the end of line. | editor | |||
US 4 | 1.1 ¶ 2 | There is a bad line break in "C++". | editor | |||
UK 1 [214] |
1.1 ¶ 2 | List of additional facilities over C has been extended with this standard, so should be mentioned in the introductory material. | Add the following to the list in 1.1p2: atomic operations concurrency alignment control user-defined literals attributes | editor |
REJECTED
This list is highlights, not all differences. Okay as is. |
|
FR 4 | 1.2 [intro.refs] ¶ 1 | Is the lack of reference to ISO/CEI 9899/AC3:2007 voluntary? | editor | |||
UK 2 [215] |
1.2 ¶ 1 | We recommend taking the latest update to each listed standard, yet the C standard is quite deliberately held back to the 1990 version without comment.+ | ... not sure ... | editor |
REJECTED
1.2 (Normative references) cites 9899:1999, 9899:1999/Corl.1:2001, and 9899:1999/Cor.2:2004, which collectively constitute C99. |
|
UK 3 [217] |
1.3.1 | The definition of an argument does not seem to cover many assumed use cases, and we believe that is not intentional. | Revise the definition of argument to answer question such as: Are lambda-captures arguments? Are type names in a throw-spec arguments? 'Argument' to casts, typeid, alignof, alignas, decltype and sizeof? why in x[arg] : arg is not an agrument, but the value forwarded to operator[]() is ? Does not apply to operators as call-points not bounded by parenthises ? Similar for copy initialization and conversion? what are Deduced template 'arguments'? what are 'default arguments'? can attributes have arguments? what about concepts, requires clauses and concept_map instantiations? What about user-defined literals where parens are not used? | CWG | 783 | |
UK 4 [216] |
1.3.3 | This definition is essentially worthless, as it says nothing about what distinguished a diagnostic message from other output messages provided by the implementation | ... add something about the diagnostic message being a message issues by the implementation when translating a program that violates the rules of the standard. ... | CWG |
REJECTED
Characterizing the output of a translator is essentially a quality-of-implementation issue and beyond the scope of the Standard. |
|
FR 5 | 1.3.4 [defns. dynamic.type] | "The dynamic type of an rvalue expression is its static type." Is this true with rvalue references? | CWG | 690 | ||
US 5 | 1.3.5 | The wording is unclear as to whether it is the input or the implementation "that is not a well-formed program". | Reword to clarify that it is the input that is here considered not well-formed. | editor | ||
FR 6 | 1.3.6 [defns. impl.defined] | There is a page break between the title and the paragraph. | editor | |||
FR 7 | 1.3.13 [defns. undefined] | [intro.execution]/5 explicitly allows non causal undefined behaviour, | Adding it to the note outlying possible undefined behaviours. | editor |
REJECTED
This is covered by the first alternative in the note, "ignoring the situation completely with unpredictable results". |
|
US 6 | 1.3.14 | Unspecified behavior does not clearly state whether or not undefined behavior is permitted. (The standard says that "usually, the range of possible behaviors is delineated", but what happens if the range is not delineated? Is a crash, or worse, allowed?) | Clearly state whether or not Unspecified behavior includes undefined behavior. | CWG |
REJECTED
The cited text is in a note and thus non-normative. The acceptable range for unspecified behavior is a quality-of-implementation issue. |
|
FR 8 | 1.4 [intro. compliance] ¶ 8 | The paragraph as its stands seems to require that violations of the ODR (which make a program ill-formed) are required to be diagnosed if the program also uses an extension which defines some cases of ODR. | editor |
REJECTED
Yes, that seems to be the intention. That's the mechanism for extensions: issue a diagnostic and then do the extension. |
||
UK 5 [1] |
1.5 | Missing checklist of implementation defined behaviour (see ISO/IEC TR 10176, 4.1.1p6) | Provide a new annex with the missing checklist | editor | ||
UK 6 [2] |
1.5 | Missing annex describing potential incompatibility to previous edition of the standard (see ISO/IEC TR 10176, 4.1.1p9) | Provide a new annex with the missing documentation. See n2733(08-0243) for a starting point | CWG | 784 | |
US 7 | 1.5 ¶ 2 | There is no mention of Clause 17. | Include Clause 17 among the list of Clauses that specify the Standard Library. | editor | ACCEPTED | |
US 8 | 1.5 ¶ 2 | The paragraph omits to mention concepts and concept maps among its list of entities defined in the Standard Library. | Mention concepts and concept maps among the list of entities. | editor | ACCEPTED | |
US 9 | 1.6 ¶ 1 | The syntax description does not account for lines that wrap. | editor | ACCEPTED | ||
US 10 | 1.7 ¶ 3 | The term thread is used before defined. | Reference 1.10 [intro.multithread]. | editor | ACCEPTED | |
US 11 | 1.7 ¶ ¶ 3 last sent. | The phrase “threads of execution” should be accompanied by a reference to [intro.multithread]. | Insert the recommended reference. | editor | ACCEPTED | |
US 12 | 1.7 ¶ ¶ 3 first sent. | A memory location is not an object as the sentence claims. | Clarify that a memory location “holds” an object rather than that it “is” an object. | CWG |
ACCEPTED with MODIFICATIONS
This is the definition of the term “memory location” for reference in the following text. The current wording is appropriate for that usage; the formatting will be changed to indicate that the term is being defined. See paper N2841 |
|
US 13 | 1.7 ¶ ¶ 3 last sent. | It is unclear what is meant by memory locations that are "separate": are they distinct? non-overlapping? how much "separation" is needed? | Provide either a better definition of “separate” or reword (this and subsequent paragraphs) to avoid this term. | CWG |
REJECTED
The current text is clear enough: “separate” in the sense of “distinct” is common usage elsewhere in the Standard. |
|
US 14 | 1.7 ¶ ¶ 4 | The phrase "no matter what the sizes of the intervening bit-fields happen to be" contradicts the claim of separation "by a zero-length bit-field declaration". | Delete the “no matter…” phrase, or resolve the contradiction in a different way. | CWG |
ACCEPTED
See paper N2841 |
|
US 15 | 1.7 ¶ ¶ 5 | A struct does not “contain” memory locations. | Reword so that a struct is “held in” one or more memory locations. | CWG |
REJECTED
The phrasing is consistent with the definition of “memory location” given in paragraph 3. |
|
US 16 | 1.9 | The discussion of observable behavior in 1.9 is not consistent with the addition of threads to the language. Volatile reads and writes and other observable actions no longer occur in a single "sequence”. | Remove/replace various occurrences of "sequence" in 1.9. | CWG | 785 | |
UK 8 [222] |
1.9 ¶ 5 | With parallel execution there is no longer the idea of a single execution sequence for a program. Instead, a program may be considered a set of exectution sequences. | Update first sentance as: A conforming implementation executing a well-formed program shall produce the same observable behavior as one of the possible SETS OF execution sequences of the corresponding instance of the abstract machine CONFORMING TO THE MEMORY MODEL (1.10) with the same program and the same input. | CWG | 785 | |
UK 7 [218] |
1.9 ¶ 6 | Does the term 'sequence' imply all reads/writes through volatile memory much be serialized, and cannot occur in parallel on truly parallel hardware? Allow for multiple concurrent sequences where each sequence is constrained by this observable behaviour rule, and multiple sequences are constrained by the memory model and happens-before relationships defined in 1.10 | Replace 'sequence' with 'sequences'. | CWG | 785 | |
FR 9 | 1.9 [intro.execution] ¶ 16 | This example use int *v while the other examples seems to use notation like int* v. | editor | ACCEPTED | ||
US 17 | 1.10 ¶ 1 | This definition of “thread” is poor, and assumes the user already knows what multi-threaded means (probably true!). In particular, it does not deal adequately with the concept that all threads share the same address space. |
Replace first
sentence of para 1 as follows:
Under a hosted implementation, a C++ program can have more than one thread of execution (a.k.a. thread) running concurrently. Each thread is a single flow of control within a program. Anything whose address may be determined by a thread, including but not limited to static objects, storage obtained via new or by any dynamic allocator, directly addressable storage obtained through implementation-defined functions, and automatic variables, are accessible to all threads in the same program. |
CWG | 786 | |
UK 9 [133] |
2.1 ¶ 2, 4 | Undefined behaviour is a drastic way to silently ignore minor issues. The cases in this paragraph could be easily defined. In this case opt for conditionally supported behaviour, which mandates a diagnostic if the compiler is not prepared to handle the syntax consistently. | Replace undefined behaviour with conditionally supported behavior. Conditional behaviour may be implementation defined, although suggest there is a reasonable default in each case. For creating a universal-character name, splice text to create a universal-character. In the case of a file ending without a newline, treat as if the newline was implictly added, with an empty line to follow if the last character was a back-slash. | CWG | 787 | |
UK 10 [134] |
2.1 ¶ 3 | Implementation defined seems unnecessarily burdensome for negligible gain. I am yet to see code that depended on whether non-empty sequences of whitespace were concatenated. Better left unspecified. | How the compiler treats non-empty sequences of whitespace should be left unspecified, rather than implementation-defined. | CWG |
ACCEPTED
See paper N2841 |
|
FR 10 | 2.1 [lex.phases]/5 and 2.2 [lex.charset]/3 |
[defns.multibyte] "the
extended character set."
[lex.charset]/3 cited below implies that there is an extended character set per locale. [lex.phases]/5 "Each [...] universal-character-name [...] is converted to the corresponding member of the execution character set" [lex.charset]/3 "The values of the members of the execution character sets are implementation defined, and any additional members are locale-specific." Together they seem to imply that what is locale-specific is if a value is valid or not for the current locale, not the representation of a given universal character. This is not the behaviour of at least some compilers I've access to which are allowing different codes for the same abstract character in different locale. During phase 5, they are using an implementation defined char set. |
CWG | 788 | ||
UK 11 [135] |
2.3 | Trigraphs are a complicated solution to an old problem, that cause more problems than they solve in the modern environment. Unexpected trigraphs in string literals and occasionally in comments can be very confusing for the non-expert. | Deprecate the whole of 2.3 and move it to appendix D. | CWG | 789 | |
UK 12 [137] |
2.4, 2.8 ¶ 2 | This undefined behaviour in token concatenation is worrying and we believe hard to justify. An implementation should either support this in a defined way, or issue a diagnosic. Documenting existing practice should not break existing implementations, although unconditionally requiring a diagnostic would lead to more portable programs. | Replace undefined behaviour with conditionally supported behaviour with implementation defined semantics. | CWG | 787 | |
US 18 | 2.4 ¶ ¶ 2 | The paragraph begins with an empty line. | Delete the empty line. | editor | ACCEPTED | |
FR 11 | 2.4 [lex.pptokens] ¶ 3 | There are spurious empty lines. | editor | ACCEPTED | ||
FR 12 | 2.5 [lex.digraph] and 2.11 [lex.key]/2 | The alternative representations are reserved as such even in attribute. Is that what is wanted? | CWG |
REJECTED
The current specification is as intended. |
||
FI 2 | 2.5 ¶ Table 2 | Add eq, for spelling out == in order to distinguish it from the assignment operator. | See eq-keyword.doc, eq-keyword.ppt | CWG |
REJECTED
Adding a new keyword would break existing code that uses eq as an identifier. |
|
UK 13 [138] |
2.9 ¶ 2 | This text is confusing in isolation, as it implies pp-numbers do not have a value in translation phase 4 when evaluating #if preprocessor expressions. | Add a note with a cross-refernce to 16.1 that a pp-number may briefly acquire a value during translation phase 4 while evaluating #if expressions. | CWG | 832 | |
UK 14 [395] |
2.11 ¶ table 3 | The table is nearly sorted, but not quite. It was sorted in previous versions of the standard. | Sort the table. | editor | ACCEPTED | |
JP 1 | 2.11 ¶ Table 3 | Keywords in the table are listed disorderly. Also, a part of a frame of the table is not drawn. | Sort it in alphabetical order. Complete the table frame. | editor | ACCEPTED | |
US 19 | 2.13.1 ¶ Table 5, rows “l or L” and “ll or LL” | The final entry in the last column (“unsigned long int”) is incorrect. | Replace the incorrect entries by “unsigned long long int”. | editor | ACCEPTED | |
US 20 | 2.13.1, 2.13.3 | Long strings of digits in literals are a continuing problem in the production and maintenance of programs. | Adopt the 1983 technology of Ada and use underscores to separate digits. http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2007/n2281.html | CWG |
REJECTED
This was already considered and rejected. |
|
UK 15 [139] |
2.13.2 ¶ 2 | Inconsistency between definition of a multicharacter literal and a wide character literal containing multiple c-chars. | Define the term multicharacter wide literal for a wchar_t literal containing multiple elements, and specify its type is integer (or wider) | CWG |
REJECTED
The current specification is well-defined, and there was no consensus for changing it. |
|
UK 16 [140] |
2.13.2 ¶ 3 | Not immediately clear why the question mark needs escaping. A note would help. | Add a note explaining that the ? character may need escaping to avoid accidentally creating a trigraph. | editor | ACCEPTED | |
JP 2 | 2.13.4 ¶ 1st para, 2nd line | Typo, R"..." should be R"[...]" | Correct typo. | editor | ACCEPTED | |
JP 3 | 2.13.4 ¶ 2nd paragraph | We think that the explanation of d-char-sequence is not enough. |
Add
the following.
|
editor | ACCEPTED | |
JP 4 | 2.13.4 ¶ 3rd para, 1st line of example |
Typo. Lack of a necessary backslash in the first line of
the example as follows:
const char *p = R"[a b c]";should be const char *p = R"[a\ b c]"; |
Correct typo. | editor | ACCEPTED | |
US 21 | 2.13.4 ¶ ¶ 3 | The paragraph, marked as a Note, contains an embedded example not marked as such. | Denote the code (and perhaps also its commentary) as an Example. | editor |
REJECTED
Marking things as Note or Example distinguishes them from normative text. Examples in Notes do not need to be marked. |
|
US 22 | 2.13.4 ¶ ¶ 3 | The code does not have the effect predicted by its accompanying narrative. | Append a backslash to the first line of the code. | editor | ACCEPTED | |
JP 5 | 2.13.4 ¶ 11th para, Table 7 | It is not explicit how to combine raw-string and non-raw-string. | Add rules containing raw-string in the table 7. | CWG | 790 | |
FR 13 | 2.13.4 [lex.string] ¶ 3 |
Shouldn't the assert be
assert(std::strcmp(p, "a\nb\nc") == 0); |
editor |
ACCEPTED with MODIFICATIONS
The assert is correct but the preceding raw string is wrong. See JP 4. |
||
UK 17 [141] |
2.13.4 ¶ 10 | It would be preferred for attempts to modify string literals to be diagnosable errors. This is not possible due to the deprecated implicit conversion to pointer to null-terminated character sequence of non-const characters. If this deprecated conversion were remove (see other comments) then string literals are always accessed through const types, and the compiler can enforce the no modification rule. The only exception would be using const_cast to cast away constness, but this is already covered under the const_cast rules so needs no further detail here. | (asssuming deprecated conversion to non-const array is removed or can be turned off) Strike the sentence on undefined behaviour. | CWG |
REJECTED
The specification covers all attempts to modify the contents of a string literal, including via a pointer to the string when it is not apparent that a string literal is the target. This is consistent with the general treatment of const objects. |
|
UK 18 [142] |
2.13.4 | The addition of static_assert (7p4) to the language raises the need to concatenate string representations of integral constant expressions (typically from a sizeof or alignof expression) with narrow string literals to provide an informative error message. There is no need to support arbitrary constant expressions and especially not floating point values or formatting flags. Likewise, the need is purely to support static_assert so only narrow string literal support is required, although generalizing to other literal types would be useful. | Define a syntax to support string-ization of integral constant expressions in a form eligible for string literal concatenation, 2.13.4p6. Suggested syntax: I" integral-constant-expression ". There is no raw variant, although it could combine with type specifier in the same way that the R prefix does, supporting u8I, uI, UI and LI. | CWG |
REJECTED
There was no consensus to adopt this new functionality at this point in the standardization process. |
|
UK 19 [143] |
2.13.4 | The grammar for string literal is becoming unwieldy and could easily be refactored into the type optional specifier and the string contents. | Refactor string-literal grammar as: (note - current Drupal view loses formatting which is vital to clearly read the grammar) string-literal: string-literal-type-specifierOPT string-literal-body string-literal-type-specifier: one of u8 u U L string-literal-body: " s-char-sequenceOPT " R raw-string | CWG |
REJECTED
The Standard is correct as written. The suggestion would change the shift/reduce conflicts in the grammar and is thus not editorial. |
|
FR 14 | 3 [basic] ¶ 7 | "In general it is necessary to determine whether a name denotes one of these entities before parsing the program that contains it." |
Would prefer
"... before continuing to parse the program that contains it." or even "... to complete the parsing of the program that contains it." as some names denotes entities declared after the first occurrence. |
editor | ACCEPTED | |
FR 15 | 3 [basic] ¶ 8 | /operator-function-id/, /conversion-function-id/, /template-id/ are followed by a space and then a "s" while usually such production names aren't followed by a space when put in plural (see /identifier/). | editor | ACCEPTED | ||
UK 20 [209] |
3 | Chapter 3 ("Basic concepts") provides common definitions used in the rest of the document. Now that we have concepts as a primary feature, the title of this chapter can be confusing as it does not refer to the language feature but to definitions used in the document. | Change the title to "Basic definitions". | editor | ||
UK 21 [359] |
3 ¶ 2 | Concepts is now the name of a specific feature of the language, the term now risks confusion and ambiguity when used in the more general sense. | Rename the chapter Basic ???. THe note in p2 specifically needs similar rewording | editor | ||
UK 22 [362] |
3 ¶ 6 | References are frequently considered variables, but this definition only applies to objects. | Add "or reference" after both uses of "object" | CWG | 633 | |
UK 23 [277] |
3.1 ¶ 2 | alias-declarations are not definitions and should be added to the list | Add alias-declaration after typedef declaration. | CWG | 758 | |
UK 24 [360] |
3.1 ¶ 2 | The current words suggest the declaration of a static integral constant data member of a class cannot be a definition. Trying to fix this wording in-place will be verbose and risk raising more confusion than it solves, so suggest a footnote to call out the special case | Add a footnote attached to the static data membmer rule: *static data member delcarations of intergral type may also be definitions if a constant integral expression is provided for an initializer. | CWG |
REJECTED
The in-class declaration of such a static data member is, in fact, not a definition; an out-of-class definition must be provided if the member is "used" as described in 3.2. |
|
UK 25 [361] |
3.1 ¶ 3 | Example is misleading as implicitly defined default constructor uses default initialization, not value initialization, for non-static data members. In the case of std::String this makes no difference, but it makes a big difference for fundamental types and pointers. | Remove the : s() from the illustrated default constructor: struct C { std::string s; C() { } C(const C& x): s(x.s) { } C& operator=(const C& x) { s = x.s; return *this; } ~C() { } }; | CWG |
ACCEPTED
See paper N2841 |
|
UK 26 [363] |
3.2 ¶ 1 | THe one definition rule should cover references, and unless the term 'variable' is extended to cover references the list in this paragraph is incomplete. | Either include references in the definition of 'variable' (see earlier comment) or add reference to the list in this paragraph. | CWG | 570 | |
UK 27 [364] |
3.2 ¶ 4 | A class type must be complete when catching exceptions, even by reference or pointer. See 15.3. | Add "when used in an exception-handler (15.3)" to the list. | CWG |
ACCEPTED
See paper N2841 |
|
FR 16 | 3.3 [Declarative regions and scopes.] | The scope of function parameters is defined, but what is the scope of template parameters? | CWG | 481 | ||
UK 28 [365] |
3.3.1 ¶ 3 | Class templates are not classes, so we should include this case. | ammend "class" to "class or class template" | CWG |
ACCEPTED
See paper N2841 |
|
UK 29 [366] |
3.3.10 ¶ 3 | operators and conversion functions do not have names, yet are susceptible to 'name hiding' within a class - indeed we rely on this for the implicitly declared copy-assignment operator. | Add the additional phrase "The declaration of an operator or conversion function in a derived class (Clause 10) hides the declaration of an operator or conversion function of a base class of the same operator or type;" | CWG |
REJECTED
These entities do, in fact, have names, as described in 3¶4. |
|
FR 17 | 3.5 [Program and linkage] | This section does not specify whether concept names have linkage. Do they or not? If concept names do not have linkage, then a note is appropriate, and that would be a bit surprising and curious. What is the rationale? | CWG | 791 | ||
UK 30 [367] |
3.5 ¶ 2 | This paragraph implies concepts have no linkage (do they need it?) and that the entities behind names without linkage cannot be used in other scopes. This maybe a bigger problem for concept maps? | Add a note to clarify that concepts don't need linkage. | CWG | 791 | |
UK 31 [368] |
3.5 ¶ 4 | What is the linkage of names declared inside a namespace, in turn declared inside an anonymous namespace? It is not clear why such a namespace has no linkage, and there is no language suggesting its memmbers should lose linkage with it, which we assume is the intended consequence. | Clarify rules for namespaces inside nested namespaces, or remove the restriction. | CWG |
ACCEPTED
See paper N2841 |
|
US 23 | 3.5 ¶ 6 | Bad paragraph break. | editor | ACCEPTED | ||
FR 18 | 3.5 [basic.link] ¶ 6 | The paragraph number is not aligned with the text. | editor | ACCEPTED | ||
FR 19 | 3.6 [Start and termination] | This section completely ignores the real world and practical case of dynamically linked or loaded libraries. In current computing environments, they are ubiquitous and they cannot be ignored in practical C++ programs. The Standard should address this aspect. | CWG |
REJECTED
The Committee explicitly voted that this issue could not be addressed within the time frame for this revision of the Standard. |
||
UK 32 [369] |
3.6.1 ¶ 3 | Do we really want to allow: constexpr int main() { return 0; } as a valid program? | Add constexpr to the list of ill-formed things to annotate main | CWG |
ACCEPTED
See paper N2841 |
|
US 24 | 3.6.1 ¶ 4 | std::quick_exit is not referenced. | Reference std::quick_exit as well as std::exit in saying that automatic objects are not destroyed. It should not do so in saying that calling std::quick_exit is undefined from within destructors for static or thread duration objects. | CWG | 792 | |
US 25 | 3.6.3 ¶ ¶ 2 last sent. | The parenthesized phrase, introduced via “i.e.” is in the nature of an example. | Change “i.e.” to “e.g.” | editor |
REJECTED
The parenthesized phrase is in the nature of a clarification, so "i.e." is appropriate. |
|
JP 6 | 3.7.4.1 ¶ 4th para, 4th line |
Typo.
Lack of a comma right after “(3.7.2)” in the sentence while there are commas after any other recitations like “(3.7.1)”. It is just a unification matter. [ Note: in particular, a global allocation function is not called to allocate storage for objects with static storage duration (3.7.1), for objects or references with thread storage duration (3.7.2) for objects of type std::type_info (5.2.8), or for the copy of an object thrown by a throw expression (15.1). -end note ] should be [ Note: in particular, a global allocation function is not called to allocate storage for objects with static storage duration (3.7.1), for objects or references with thread storage duration (3.7.2), for objects of type std::type_info (5.2.8), or for the copy of an object thrown by a throw expression (15.1). -end note ] |
Correct typo. | editor | ACCEPTED | |
DE 3 | 3.7.4.3 |
DE-3 It is
unclear whether the following code has well-defined
behavior; none of the bullets in the second paragraph seem
to apply.
int& i = *new int(5); delete &i; |
Clarify that &i is considered a safely-derived pointer value. | CWG | 735 | |
US 26 | 3.8 ¶ 1 and 5 | Use of object fields during destruction is excessively and erroneously constrained. | See the attached document "Issues with the C++ Standard" under Chapter 3 "Use of objects, especially from other threads, during destruction". | CWG | 793 | |
US 27 | 3.9 ¶ ¶ 9 first sent. | There is a superfluous/extraneous “and”. | Delete “and” from the phrase “and std::nullptr_t”. | editor | ACCEPTED | |
FR 20 | 3.9 [Types] | The phrase 'effective type' is defined and used in a way that is incompatible with C99. Such a deliberate incompatible choice of terminology is both unfortunate and confusing, given past practice of the committee to maintain greater compatibility with C99. We strongly suggest that the phrase 'effective type' not be used in such an incompatible way. | CWG |
REJECTED
The choice of the term was previously discussed and no better term could be found. |
||
JP 7 | 3.9.2 ¶ 3rd para, 13th line | over-aligned type was added as new notion. So it is preferable to add the link after that. |
Add
(3.11) after over-aligned type as the link.
[ Note: pointers to over-aligned types(3.11) have no special representation, but their range of valid values is restricted by the extended alignment requirement. This International Standard specifies only two ways of obtaining such a pointer: taking the address of a valid object with an over-aligned type(3.11), and using one of the runtime pointer alignment functions. An implementation may provide other means of obtaining a valid pointer value for an over-aligned type(3.11).—end note ] |
editor | ACCEPTED | |
US 28 | 3.9.3 ¶ ¶ 5 first sent. | The closing braces of the first two sets are preceded by extraneous space. | Delete the extra spaces. | editor | ACCEPTED | |
DE 4 | 4.2 ¶ p2 | DE-4 The deprecated conversion from string literals to pointer to non-const character types should be limited to those conversions and types of string literals that were already present in ISO/IEC 14882:2003, or the deprecated conversions should be removed entirely. | Consider applying the proposed resolution presented in core issue 693 in WG21 document N2714 “C++ Standard Core Language Active Issues, Revision 58“, available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2714.html ; or remove only the conversions to "pointer to char16_t", "pointer to char32_t" in 4.2 paragraph 2 and 15.1 paragraph 3. | CWG | 693 | |
CH 1 | 4.11 and 5.2.9 | With respect to the target type, pointer to members should behave like normal pointers (least surprise principle). | The standard should allow implicit conversions from ``pointer to member of T of type cv D'' to ``pointer to member of T of type cv B'', where D is of class type and B is a public base of D, It should allow explicit conversion the other way around. | CWG | 794 | Section reference corrected from 4.9. |
DE 5 | 4.11, 5.3.1, 5.5 | DE-5 Ref-qualification has not been integrated with pointer-to-members. | Review implicit conversions (4.11), forming pointer-to-members (5.3.1), and dereferencing pointer-to-members (5.5) for type-safety concerns in the presence of ref-qualifiers on the member. | CWG | 731 | |
UK 33 [374] |
4.13 ¶ 1 | We have: "No two signed integer types shall have the same rank ..." "the rank of char shall equal the rank of signed char" Can we therefore deduce that char may not be signed? | Replace the first sentence with "No two signed integer types shall have the same rank, even if they have the same representation, except that signed char shall have the same rank as char even if char is signed (3.9.1/1)." | CWG |
ACCEPTED
See paper N2841 |
|
UK 34 [375] |
4.13 ¶ 1 | 6th bullet, "the rank of char" - first letter should be capitalised for consistency with the other bullets | The rank of char | editor | ACCEPTED | |
UK 36 [407] |
5.1 ¶ 1 | Primary expressions are literals, names, names qualified by the scope resolution operator ::, and lambda expressions. The immediately following grammar flatly contradicts this - this and (e) are also lambda expressions. | Delete this paragraph. | CWG |
ACCEPTED
The introductory sentence was deleted, but the paragraph (containing the grammar) remains. See paper N2841 |
|
UK 37 [224] |
5.1 ¶ 11 | Member function templates are not member functions, so should also be listed in the 3rd bullet | Add member function templates to the 3rd bullet | CWG |
REJECTED
The text is clear enough as it is. There are many places in the Standard where function templates are not explicitly mentioned in a reference to functions. |
|
UK 38 [230] |
5.1 ¶ 3 | this might be useful in a few more places than it is permitted, specifically in decltype expressions within a class. Two examples that would be ill-formed at class scope without changes: typedef decltype( *this ) this_type; decltype( [this]{ return this->memfun(); } ) my_lambda; | ... words to follow ... | CWG |
REJECTED
There was no consensus for this change. |
|
JP 8 | 5.1 ¶ 7th para, Syntax rules |
In
the current syntax definition, a scope operator(::) cannot
be applied to decltype, but it should be. It would be
useful in the case to obtain member type(nested-type) from
an instance as follows:
vector<int> v; decltype(v)::value_type i = 0; // int i = 0; |
Add
“decltype ( expression ) :: “ to
nested-name-specifier syntax like below.
nested-name-specifier: type-name :: namespace-name :: nested-name-specifier identifier :: nested-name-specifier templateopt simple-template-id :: nested-name-specifieropt concept-id :: decltype ( expression ) :: |
CWG | 743 | |
JP 9 | 5.1.1 |
It
would be preferable that “&&” could be
specified in a lambda expression to declare move capture.
Here is an example from N2709. template<typename F> std::unique_future<typename std::result_of<F()>::type> spawn_task(F f){ typedef typename std::result_of<F()>::type result_type; struct local_task { std::promise<result_type> promise; F func; local_task(local_task const& other)=delete; local_task(F func_): func(func_) {} local_task(local_task&& other): promise(std::move(other.promise)), f(std::move(other.f)) {} void operator() { try { promise.set_value(f()); } catch(...) { promise.set_exception(std::current_exception()); } } }; local_task task(std::move(f)); std::unique_future<result_type> res(task.promise.get_future()); std::thread(std::move(task)); return res; } This can be rewritten simply as follows if move capture can be used in a lambda expression. template<typename F> std::unique_future<typename std::result_of<F()>::type> spawn_task(F f){ typedef typename std::result_of<F()>::type result_type; std::promise<result_type> promise; std::unique_future<result_type> res(promise.get_future()); std::thread([&&promise, &&f]() { try { promise.set_value(f()); } catch(...) { promise.set_exception(std::current_exception()); } }); return res; } |
Add move capture in a lambda expression. | CWG |
REJECTED
This would be a dangerous feature, allowing “moving” from named variables. It would also contradict the change made by paper N2844, prohibiting the binding of rvalue references to lvalues, which was adopted at the March, 2009 meeting. |
|
JP 10 | 5.1.1 |
In
the current syntax definition, a returned type of a
function object cannot be obtained by using result_of from
an unnamed function object generated by a lambda expression
because it doesn’t have result type.
template <class F> void foo(F f) { typedef std::result_of<F()>::type result; // error } foo([]{}); If “Callable” or “Predicate” concept is specified, a returned type can be obtained from a function object without result_type. But it is preferable to be able to obtain it with template. |
Add result_type to the syntax of an unnamed function object generated by a lambda expression. | CWG |
REJECTED
This is not a defect. The definition of std::result_of now uses decltype and does not require a result_type member typedef, so the desired functionality is already present in the current specification. |
|
US 29 | 5.1.1 | The standard does not state whether or not direct recursion of lambdas is possible. | CWG | 762 | ||
US 30 | 5.1.1 |
The standard does not clarify the
meaning of this in
lambdas. Does it mean this lambda, or this class within
which the lambda is nested?
|
CWG | 762 | ||
US 31 | 5.1.1 | The current wording does not specify how context capturing and name resolution take place when the inner lambda refers to the outer lambda's locals variables and parameters. | CWG | 752 | ||
UK 45 [444] |
5.1.1 ¶ para 2 | Lambda is a language feature with an apparent dependency on <functional>. This increases dependency of language on library, and is inconsistent with the definition of freestanding in 17.6.2.4. | Change the text "a closure object behaves as a function object" to "a closure object is a built-in object which behaves as a function object"; and after "context.", insert " A closure object may be used without any need for <functional>." This makes clear what may already be implied, namely that lambdas can be used in freestanding implementations and don't increase dependency of language on library. (Marked as technical comment anyway because this clarity is technically important). | CWG | 795 | |
US 32 | 5.1.1 ¶ 3 | The final italic "this" in the paragraph should be a teletype "this". | editor | ACCEPTED | ||
UK 39 [408] |
5.1.1 ¶ 11 | This paragraph lists all the special member functions for the class representing a lambda. But it omits the destructor, which is awkward. | Add "F has an implicitly-declared destructor". | CWG | 759 | |
UK 40 [409] |
5.1.1 ¶ 13 | If one or more names in the effective capture set are preceded by &, the effect of invoking a closure object or a copy after the innermost block scope of the context of the lambda expression has been exited is undefined. That is too restrictive. The behaviour should be undefined iff the lifetime of any of the variables referenced has ended. This should be safe and legal; currently it has undefined behaviour: int i; reference_closure<void ()> f; if (blah) { f = [&i]() { }; } if (f) f(); | If one or more names in the effective capture set are preceded by &, the effect of invoking a closure object or a copy after the lifetime of any of the variables referenced has ended is undefined. | CWG | 796 | Paragraph reference corrected from 12. |
UK 41 [225] |
5.1.1 ¶ 12 | For argument dependant lookup (3.4.2) the associated namespaces for a class include its bases, and associated namespaces of its bases. Requiring the result of a lambda expression *to dervide from* std::reference_closure means that ADL will look in namespace std when the lambda capture is entirely by reference, which might have surprising results. Also, relying on the idea of implicitly slicing objects is uncomfortable. | Replace inheritance with implicit conversion. | CWG |
ACCEPTED with MODIFICATIONS
std::reference_closure was removed at the March, 2009 meeting, rendering the issue moot. See paper N2845 |
|
UK 42 [226] |
5.1.1 | A lambda with an empty capture list has identical semantics to a regular function type. By requiring this mapping we get an efficient lambda type with a known API that is also compatible with existing operating system and C library functions. | Add a new paragraph: "A lambda expression with an empty capture set shall be convertible to pointer to function type R(P), where R is the return type and P is the parameter-type-list of the lambda expression." Additionally it might be good to (a) allow conversion to function reference (b) allow extern "C" function pointer types | CWG | 797 | |
UK 43 [231] |
5.1.1 ¶ 12 | The note spells out the intent that objects from lambda-expressions with an effective capture list of references should be implemented as a pair of pointers. However, nothing in the rest of 5.1.1 lifts the requirement of to declare a reference member for each captured name, and a non-normative note is not enough to relax that. | ... provvide exceptions in the right places ... | CWG |
ACCEPTED with MODIFICATIONS
std::reference_closure was removed at the March, 2009 meeting, rendering the issue moot. See paper N2845 |
|
UK 44 [252] |
5.1.1 ¶ 12 | There is a strong similarity between a [&]{} lambda capturing a stack frame, and a [this]{} lambda binding a member function to a class instance. The reference_closure requirement should be extended to the second case, although we need some syntax to create such an object that is distinct from the existing pointer-to-member syntax. This would be a cleaner alternative to the new std::mem_fn library component. | Extend reference_closure requirement to cover [this] lambdas. Consider a simple syntax for creating such bound expressions. | CWG |
REJECTED
There was no consensus to make the suggested change at this point in the standardization process. Furthermore, std::reference_closure was removed at the March, 2009 meeting, rendering the issue moot. |
|
UK 46 [449] |
5.1.1 ¶ para 12 | The requirement that a lambda meeting appropriate conditions be an object derived from reference_closure makes lambda the language feature dependent on <functional>, which increases dependency of the language on the library and bloats the definition of freestanding C++. | Replace text "is publicly derived from" with "shall be implemented in a manner indistinguishable from". This places an ABI constraint on reference closures such that compiler and library implementer have to do compatible things. But it cuts the dependency of lambda syntax on <functional>. | CWG |
ACCEPTED with MODIFICATIONS
std::reference_closure was removed at the March, 2009 meeting, rendering the issue moot. See paper N2845 |
|
DE 6 | 5.1.1, 20.7.18 | DE-6 Some uses of lambda expressions refer to specializations of the unconstrained class template std::reference_closure (5.1.1). If the lambda expression appears in a constrained context and the return type or a parameter type for the lambda depend on a template parameter (see 14.10), such a use is ill-formed. | In 20.7.18, for the class template std::reference_closure, require Returnable for R and VariableType for each of the ArgTypes. | CWG |
ACCEPTED with MODIFICATIONS
std::reference_closure was removed at the March, 2009 meeting, rendering the issue moot. See paper N2845 |
|
DE 7 | 5.1.1 ¶ p10 | DE-7 The note at the end of paragraph 10 appears to be garbled. | Remove "or references" in the note. | editor |
REJECTED
Looks okay: "captured" applies to "variables or references". |
|
DE 8 | 5.1.1 ¶ p10 | DE-8 The construction of the function call operator signature is missing specifications for the ref-qualifier and the attribute-specifier. | Add bullets that say that the ref-qualifier and the attribute-specifier are absent. | CWG |
ACCEPTED
See paper N2841 |
|
US 33 | 5.1.1 ¶ 11 | There is no definition of “move constructor” or “move operation” | Since this is the first place the terms are used, a definition should either be added here, or a cross reference to one. | CWG | 680 | |
DE 9 | 5.1.1 | DE-9 There is not a single example of a lambda-expression in the standard. See also core issue 720 in WG21 document N2791 "C++ Standard Core Language Active Issues, Revision 59", available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2791.html . | Add a few well-chosen examples. | CWG | 720 | |
UK 52 [232] |
5.2 ¶ 3 | This paragraph seens out of place, assignment expressions are covered in 5.17 | Move p3 to subsection 5.17 | editor |
ACCEPTED with MODIFICATIONS
Removed the paragraph; grammar changes make it moot. |
|
UK 53 [233] |
5.2.1 | The definition in p1 makes no allowance for overloaded operator[] that treats the expression as a simple function call, and does not support the interchangability of arguments. Howver p2 relies on this definition when describing the use of brace-init-lists inside []. | Insert a new p2 describing the changed semantics for overloaded operator[]. This should be a note to avoid introducing normative text that could potentially conflict with the later definiton of these semantics. | CWG | 798 | |
UK 59 [410] |
5.2.2 ¶ 7 | When there is no parameter for a given argument, the argument is passed in such a way that the receiving function can obtain the value of the argument by invoking va_arg. That shouldn't apply to parameter packs. template <class ... Types> void f(Types ... pack); f(1, 2, 3); | Clarify that this sentence only applies where the ellipsis is used. | CWG |
ACCEPTED
See paper N2841 |
|
UK 60 [411] |
5.2.5 ¶ 3 | In the remainder of 5.2.5, cq represents either const or the absence of const vq represents either volatile or the absence of volatile. | Add "and" before vq | editor | ACCEPTED | |
UK 61 [234] |
5.2.5 ¶ p1 | Together with footnote 60 there may be confusion that the postfix expression is always evaluated - even when part of an unevaluated operand. We believe the standard does not require this, and a comment in the existing note would be a useful clarification. | Clarify in footnote 60 that this will not happen if the whole expression is an unevaluated operand. | editor | ACCEPTED | |
UK 62 [235] |
5.2.5 ¶ 4 | In the final bullet, what does 'not an lvalue' mean? Does it imply rvalue, or are there other possible meanings? Should clauses that trigger on rvalues pick up on this? | Replace 'not an lvalue' with 'is an rvalue'. | CWG |
ACCEPTED
See paper N2841 |
|
DE 10 | 5.2.5 | DE-10 If E1.E2 is referring to a non-static member function, the potential ref-qualification on E2 should be taken into account. | Adjust the presentation of the types involved as appropriate. | CWG | 731 | |
UK 63 [412] |
5.2.6 ¶ 2 | Paragraph 2 is missing its number. | Add one. | editor | ACCEPTED | |
UK 64 [413] |
5.2.7 ¶ 3 | A new name R is introduced for use in paragraphs 3 and 4. But R is the same as T. | Replace R with T and replace "the required result type (which, for convenience, will be called R in this description)" with "T". | editor | ACCEPTED | |
UK 65 [414] |
5.2.7 ¶ 8 | In the first two bullets we have "the result is a pointer (an lvalue referring) to". But para 2 makes clear that a dynamic_cast of an rvalue references produces a rvalue. (Can an lvalue refer to anything anyway?) | Replace "an lvalue referring to" with "reference", twice. | CWG |
ACCEPTED with MODIFICATIONS
See paper N2841 |
|
UK 66 [415] |
5.2.8 ¶ 1 | typeid may return "an implementation-defined class derived from std :: type_info". The derivation must be public. | an implementation-defined class publicly derived from std :: type_info | CWG |
ACCEPTED
See paper N2841 |
|
UK 67 [416] |
5.2.9 ¶ 1, 2, 3 | Paragraph 1 specifies when the result of static_cast is an lvalue; repeating it is unnecessary. | In para 2, delete "It is an lvalue if the type cast to is an lvalue reference; otherwise, it is an rvalue." and "The result is an rvalue.". In para 3, delete "The result is an lvalue if T is an lvalue reference type (8.3.2), and an rvalue otherwise." | CWG |
ACCEPTED
The sentence in question is 20.1.1/5. See paper N2841 |
|
UK 54 [417] |
5.2.10 ¶ 3, 6 | Para 3: "The mapping performed by reinterpret_cast is implementation-defined.". Para 6: "... the result of such a pointer conversion is unspecified." Which is it? | In para 6, replace unspecified with implementation-defined. Alternatively, delete paragraph 3, since individual cases are labelled appropriately. | CWG |
ACCEPTED
See paper N2841 |
|
UK 55 [236] |
5.2.10 ¶ 2 | dynamic_cast and reinterpret_cast crossreference 5.2.11 without creating an extra note. The second half of the note is unrelated to the crossrefernce, and would serve as well in normative text. | Strike the note about definition of casting away constness, preserve the cross-reference. The second sentance on reintrepret_cast to its own type should move out of the note into the normative text. | CWG | 799 |
The repetition of the reference to 5.2.11 was handled as a quasi-editorial
change. An issue will be opened for the question about casting to the
same type as the operand.
See paper N2841 |
UK 56 [237] |
5.2.10 ¶ 5 | The notion of safely derived pointers means this conversion may not be as safe in the revised standard as the original. It would be good to call attention to the changed semantics with a note. | Add: [Note: the result of such a conversion will not be a safely-derived pointer value (3.7.4.3) -- end note] | CWG |
ACCEPTED with MODIFICATIONS
See paper N2841 |
|
UK 57 [238] |
5.2.10 ¶ 8 | Conditionally supported behaviour gives a wide range or permission, so clarify relationship between safely-derived object pointers and function pointers in a note. | Add: [Note: In such cases, the implementation shall also define whether a safely-derived object pointer cast to a function pointer can be safely cast back -- end note] | CWG | 800 | |
UK 58 [418] |
5.2.11 ¶ 9 | Casting from an lvalue of type T1 to an lvalue of type T2 using a reference cast casts away constness if a cast from an rvalue of type “pointer to T1” to the type “pointer to T2” casts away constness. That doesn't cover rvalue references. | Replace lvalue with "lvalue or rvalue" twice. | CWG | 801 | |
US 34 | 5.3 ¶ 1 | The list of unary operator should be in teletype font. | editor | ACCEPTED | ||
UK 68 [419] |
5.3.1 ¶ 2-9 | All the unary operands other than * return rvalues - but this is not stated. | Add a paragraph 1a "The following unary operators all produce results that are rvalues." | CWG |
ACCEPTED
See paper N2841 |
|
UK 69 [240] |
5.3.1 ¶ 2 | If we cannot bind references/take address of functions in concept_maps, does that mean we cannot use generic bind in constrained templates? Launch threads with expressions found via concept map lookup? Hit problems creating std::function objects? Does the problem only occur if we use qualified lookup to explicitly name a concept map? Does it only kick in if we rely on the implicit function implementation provided by a concept_map, so some types will work and others won't for the same algorithm?! | ... unknown ... | CWG | 802 | |
UK 70 [420] |
5.3.3 ¶ 1 | The sizeof operator shall not be applied to ... an enumeration type before all its enumerators have been declared We should allow enum E : int; sizeof(E). | Change "an enumeration type" to "an enumeration type whose underlying type is not fixed". | CWG | 803 | |
UK 71 [254] |
5.3.4 ¶ 2 | The type of an allocated object wih the type specifier auto is determined by the rules of copy initialization, but the initialization applied will be direct initialization. This would affect classes which declare their copy constructor explicit, for instance. For consistency, use the same form of initiailization for the deduction as the new expression. | Replace T x = e; with T x(e); | CWG | 804 | |
UK 72 [257] |
5.3.4 ¶ 7 | The library headers have been carefully structured to limit the dependencies between core language and specific headers. The exception thrown should be catchable by a handler for a type lised in <exception> header in cluase 18. This might be accomplished by moving length_error into the <exception> header, but its dependency on logic_error with its std::string constructors suggest this is not a good idea. Prefer to pick an existing exception instead. | Throw std::bad_alloc instead of std::length_error. | CWG | 805 | |
UK 73 [258] |
5.3.4 ¶ 6 | A class type with conversion operator can only be used if the conversion type is constexpr and the class is a literal type. Adding the single word 'literal' before class type will clarify this. | Add 'literal' before 'class type' | CWG |
REJECTED
This is not a defect. The leftmost subscript in the noptr-new-declarator (the expression, as opposed to the constant-expression) need not be constant. |
|
UK 74 [259] |
5.3.4 ¶ 8 | operators, like constructors and destructors, do not have names. However, in certain circumstances they can be treated as if they had a name, but usually the stanadard is very clear not to actually describe their name as a distinct property. | Change "the allocation function’s name is operator new" to "the allocation function is named operator new" and similarly for operator delete. | CWG |
REJECTED
The premise is incorrect: according to 3 [basic] paragraph 4, operators do indeed have names. |
|
UK 35 [260] |
5.3.4 ¶ 9 | Missing period in middle of paragraph between "in the scope of T" and "If this lookup fails" | Add a period between "in the scope of T" and "If this lookup fails" | editor | ACCEPTED | |
UK 75 [262] |
5.3.5 ¶ 8 | A paragraph strarting with [Note... is easily skipped when reading, missing the normative text at the end. | Swap order of the note and normative text. | editor | ACCEPTED | |
FR 21 | 5.3.6 [Alignof | Should not the type of alignof-expression be of type std::max_align_t? | CWG |
REJECTED
std::max_align_t is not guaranteed to be integral, only a POD. std::size_t is the correct choice. |
||
US 35 | 5.8 ¶ 2 and 3 | There is curious spacing in the expressions "E1 <<E2" and "E1 >>E2". This is a formatting change since previous versions of the Standard. | editor | ACCEPTED | ||
UK 47 [242] |
5.14 / 5.15 ¶ 2 | Why are the descriptions of order of evaluation of expressions and side effects different between && and || operators. The interaction with the memory model should be identical, so identical words should be used to avoid accidential inconsistencies in interpretation. | Pick one form of wording as 'the best' and apply it in both places. | CWG |
ACCEPTED
See paper N2841 |
|
UK 48 [249] |
5.18 ¶ 1 | The defining feature of the comma operator is the guaranteed sequencing of two expressions. This guarantee is lost when presented with an overloaded operator, and this change is subtle enough to call attention to it. | Add: [Note: There are no guarantees on the order of value computation for an overloaded comma operator -- end note] | editor | ACCEPTED | |
UK 49 [376] |
5.19 ¶ 2 | Is an implementation permitted to reject this? constexpr int f() { return f(); } int a[f()]; AFAICT it is well-formed; f() seems to satisfy all the rules to make it a constant expression. I would hate compilation to become a potentially non-terminating experience. | Add an escape clause to allow the implementation to reject excessively deep nesting of constexpr function evaluations. (This can possibly be a note, since it is arguable that this point is handled by the general rule on resource limits in 1.4/2. A sufficiently smart compiler could use tail recursion above, meaning that it would never run out of memory given this program though.) | CWG | 699 | |
UK 50 [378] |
5.19 ¶ 2 | The following should be valid: enum E { foo = 4}; const E c = foo; int a[c]; But currently it is not - c is not an lvalue of effective integral type (4th bullet). (See also 7.1.6.1/2) | Change "effective integral type" to "effective integral or enumeration type" in the 4th bullet, 1st sub-bullet. | CWG | 806 | |
UK 51 [251] |
5.19 ¶ 2 | typeid expressions can never be constant, whether or not the operand is a polymorphic class type. The result of the expression is a reference, and the typeinfo class that the reference refers to is polymorphic, with a virtual destructor - it can never be a literal type. | Strike the words "whose operand is of a polymorphic class type" on the bullet for typeid expressions. | CWG | 807 | |
UK 76 [131] |
6.3 | Do we really need two different terms that say the same thing? | Pick either 'block' or 'compound statement' as the preferred term and use it consistently throughout the standard. | CWG |
REJECTED
Both terms are in common use, and there was no consensus for removing one of them. |
|
FR 22 | 6.4.2 [The switch statement] |
The constant-expression in
case constant-expression should be allowed to be of any constant expression of literal type for which a constexpr comparison operator (operator< and operator==) is in scope. Now that constant expressions of other integral types are evaluated at compile time, the restriction for case-labels is at best artificial. |
CWG |
REJECTED
There was no consensus for making this change at this point in the standardization process. |
||
UK 77 [132] |
6.5 ¶ 5 | The terms i/o operation, synchronize operation and atomic operation have very specific meanings within the standard. The paragraph would be much easier to understand with the terms crossreferenced. | Profide a cross-reference for the terms: i/o operation, synchronize operation and atomic operation | editor | ACCEPTED | |
JP 11 | 6.5.4 ¶ 1st para, 5th line | There is no _RangeT type in the equivalent code to “range-base for” statement. It existed in N2049. |
Add
a typedef for _RangeT in the example as follows:
{ typedef decltype( expression ) _RangeT; auto && __range = ( expression ); for ( auto __begin = std::Range<_RangeT>:: begin(__range), __end = std::Range<_RangeT>:: end(__range); __begin != __end; ++__begin ) { for-range-declaration = *__begin; statement } } |
CWG |
REJECTED
The change was intentional; the term is defined in the descriptive text. |
|
UK 78 [130] |
6.5.4 ¶ 2 | Including the header <iterator_concepts> is far too unwieldy to enable an important and (expected to be) frequently used syntax. | Merge <iterator_concepts> into <concepts> and change 6.5.4p2 to refer to <concepts>, or make the Range concept fundamental along with the other support concepts in 14.9.4 and strike any reference to including a header. | LWG | 1001 | |
UK 79 [445] |
6.5.4 | The definition of for (for-range-declaration : expression) statement is expanded in terms which require a Range concept, and the program is ill-formed if <iterator_concepts> isn't included. For users, iterating through old-fashioned arrays, this is a sledge-hammer to crack a nut and compares poorly with other languages. It's also not possible to implement this without adversely impacting the freestanding definition in 17.6.2.4. | When expression is an array a of length N whose length is known at compile time, expand range-for as 'for (... p=a, p!=a+N, p++) ...' without requiring the Range concept or <iterator_concepts>. Also, when expression is an initializer_list, expand range-for similarly without requiring <iterator_concepts>. | CWG |
REJECTED
This idea was already considered and rejected, and there was no consensus for changing it now. This suggestion would remove functionality; with the current definition, a user could provide his/her own Range concept for an array with a user-defined type. |
|
DE 11 | 6.9 ¶ p1 | DE-11 A sentence in paragraph 1 reads: "Outside of a constrained context, the late-checked block has no effect." This, at face value, specifies that the compound-statement of such a late-checked block is never executed, which appears to be unintended. | State that such a late-checked block has the same meaning as if the late_check keyword were absent. | CWG |
ACCEPTED
See paper N2841 |
|
UK 80 [370] |
7 ¶ 1 | Many of the sections and major subsections open with a sentence summarising the content. I'm not sure this is necessary; this isn't a tutorial. The problem with these summaries is that because they omit much of the detail, they tend to be inaccurate. This may not matter, but I feel the document would be improved by omitting them. There's a prime example here: "Declarations specify how names are to be interpreted." Not true for static_assert, an asm declaration nor an anonymous bit field. | Strike the first sentence. | editor |
ACCEPTED with MODIFICATIONS
Tweaked first sentence. |
|
UK 81 [371] |
7 ¶ 4 | String literal concatenation happens in phase 6, before parsing, so it is legal and useful to use it for the string literal in a static_assert. It would be useful to add a note mentioning this. | Add a note: Multiple adjacent string literals may be used instead of a single /string-literal/; see [lex.phases]. | CWG |
REJECTED
The Standard is already clear enough on this point. |
|
UK 82 [386] |
7 ¶ 2 | Paragraph 2 talks about declarations that can have nested declarations within them. It doesn't mention scoped enumerations - but according to 7.2/11, "Each scoped enumerator is declared in the scope of the enumeration." | Add "scoped enumeration" to the list in the second sentence. | CWG |
REJECTED
This suggestion was already considered and rejected. In particular, the following statement (“These nested scopes, in turn, can have declarations nested within them”) is not true of an enumeration scope. |
|
UK 83 [372] |
7.1 ¶ 2 | The longest sequence of decl-specifiers that could possibly be a type name is taken as the decl-specifier-seq of a declaration. But many sequences of decl-specifiers cannot possibly be a type name - eg the sequence "friend int", or "typedef int". | Not sure. I understand the rule, just not how to say it. | CWG | 808 | |
UK 84 [404] |
7.1 ¶ 1 | The grammar includes alignment-specifier as a production for decl-specifier, but there is no production for alignment-specifier. I suspect this is a holdover from before alignment was handled as an attribute. | Delete the production (including the duplicate in A6) | CWG |
ACCEPTED
See paper N2841 |
|
FI 3 | 7.1 ¶ [dcl. spec. auto] | While it’s considered too late for this standard revision, consider loosening the restrictions for auto specifier and making it more a mirror of a deduced template function parameter. | See restricted-auto.ppt | CWG |
REJECTED
As noted, the suggestion comes too late for consideration at this point in the standardization process. |
|
UK 85 [373] |
7.1.1 ¶ 1 | ... the init-declarator-list of the declaration shall not be empty (except for global anonymous unions, which shall be declared static). Global here means "declared at namespace scope". (cf 9.5/3 "Anonymous unions declared in a named namespace or in the global namespace shall be declared static."). | Replace "global" with "namespace scope". | CWG | ||
UK 86 [403] |
7.1.1 ¶ 2,3 | The register keyword serves very little function, offering no more than a hint that a note says is typically ignored. It should be deprecated in this version of the standard, freeing the reserved name up for use in a future standard, much like auto has been re-used this time around for being similarly useless. | Deprecate current usage of the register keyword. | CWG | 809 | |
UK 87 [405] |
7.1.1 ¶ 1, 4, 5 | Why require two keywords, where one on its own becomes ill-formed? thread_local should imply 'static' in this case, and the combination of keywords should be banned rather than required. This would also eliminate the one of two exceptions documented in 7.1.1p1. | Drop requirement to combine static keyword with thread_local at block-scope inside a function definition. | CWG | 810 | |
US 36 | 7.1.1 ¶ 4 | The permission to use thread_local static data members is missing. | Add the static members as a permitted use. | CWG | 717 | |
FR 23 | 7.1.5 [constexpr] |
'constexpr' functions should
be allowed to take const reference parameters, as long as
their uses are in a context where a constant expression may
be required. For example, the following should be allowed
template<typename T, int N> int size(const T(&)[N]) { return N; } int a[] = { 41,42,43,44 }; enum { v = size(a) }; |
CWG |
REJECTED
There was no consensus for making this change without a more complete proposal. |
||
JP 12 | 7.1.5 |
It should be
allowed to define constexpr recursively.
There is an explanation in N2235, Generalized Constant Expressions—Revision 5, as follows. We (still) prohibit recursion in all its form in constant expressions. That is not strictly necessary because an implementation limit on recursion depth in constant expression evaluation would save us from the possibility of the compiler recursing forever. However, until we see a convincing use case for recursion, we don’t propose to allow it. Then, here are the use cases where allowing recursion for constexpr is very useful. Range of problem to be handled with constexpr would become extended. For example, user defined type (e.g. Complex type) could be placed in ROM area. But with current specification, a function defined with constexpr cannot be called recursively. As a side effect is not allowed in compile-time, it cannot be implemented to repeat anything without recursion. Although it could be implemented without recursion like func0, func1, func2 in an example below, it is not elegant solution. constexpr double func0(double x) { /* ... */} constexpr double func1(double x) { /* call for func0 */ } constexpr double func2(double x) { /* call for func1 */ } /* ... */ - Compile-time and runtime As constexpr can be also evaluated both in compile-time and runtime, we need to discuss about both cases. Runtime evaluation is just to execute it. If you eliminate constexpr keyword, it is executable as of now. Any modern compiler may optimize tail recursion easily. Compile-time evaluation is the same thing as template recursion. It is necessary to support floating point operation, but it is already possible to calculate it in compile-time, so it’s ok. - Sample Here is an example to calculate a square root using constexpr recursively. /*constexpr*/ double SqrtHelper(double x, double a, int n) { return n == 0 ? a : SqrtHelper(x, (x / a + a) / 2.0, n - 1); } /*constexpr*/ double Sqrt(double x) { return SqrtHelper(x, x, 20); } /*constexpr*/ double root2 = Sqrt(2.0); // 1.41421... |
Allow constexpr recursion. | CWG | 699 | |
US 37 | 7.1.6.1 ¶ 1 | There is a "Note: 3.9.3 describes how cv-qualifiers affect object and function types." So far as I can see, 3.9.3 CV-qualifiers only describes cv-qualifiers for objects, cv-qualifiers for (member) functions being described in 8.3.5 Functions. | editor | ACCEPTED | ||
UK 89 [377] |
7.1.6.1 ¶ 2 | The two normative sentences in this paragraph appear to duplicate text elsewhere - but they aren't exact duplicates, which introduces uncertainty. 1. "An object declared in namespace scope with a const-qualified type has internal linkage unless it is explicitly declared extern or unless it was previously declared to have external linkage.". This nearly repeats 7.1.1/7: "Objects declared const and not explicitly declared extern have internal linkage." The former seems to allow more wiggle room - can an object be "previously declared to have external linkage" without having been "explicitly declared extern"? 2. "A variable of non-volatile const-qualified integral or enumeration type initialized by an integral constant expression can be used in integral constant expressions (5.19)." This nearly duplicates 5.19/2, bullet 4, 1st sub-bullet - "[... an integaral constant expression can use] an lvalue of effective integral type that refers to a non-volatile const variable or static data member initialized with constant expressions". The latter does not allow for lvalues of enumeration type (neither scoped not unscoped enumerations are integral types - 3.9.1/7, and note 44). This seems to be a flaw in 5.19/2. | Make the normative text in this section into one or more notes, with cross references, and correct the referenced text if necessary. | CWG | 811 | |
UK 90 [379] |
7.1.6.2 ¶ para 1 and table 9 | The grammar in paragraph one makes "nested-name-specifier template simple-template-id" a simple-type-specifier, but unlike all the others it is omitted from table 9. | Add a row to table 9 mentioning simple-template-id and punting to clause 14 (cf decltype(expression)). | editor | ACCEPTED | |
UK 91 [380] |
7.1.6.2 ¶ 4 | 5.1/5 says "[A] parenthesized expression can be used in exactly the same contexts as those where the enclosed expression can be used, and with the same meaning, except as otherwise indicated." When the first bullet point of this paragraph, describing the type denoted by decltype(e), says "if e is an id-expression ... decltype(e) is the type of the entity named by e", 5.1/5 is not excluded, which would imply that decltype((e)) was also the type of e. But the intention appears that it should be caught by the third bullet and treated as an lvalue expression, so decltype((e)) should be a reference to the type of e. Conversely, the second bullet point says "(parentheses around e are ignored)", which is redundant because of 5.1/5. | Insert "unparenthised" in the first bullet point - "if e is an *unparenthised* id-expression ...". In the second bullet point, move "(parentheses around e are ignored)" into a note. | CWG |
ACCEPTED with MODIFICATIONS
See paper N2841 |
|
UK 92 [382] |
7.1.6.3 ¶ 2 | The note correctly indicates that, if T is a template type parameter, then "friend class T;" is ill-formed. It might be worth pointing out at the same time that the alternative "friend T;" is now allowed - see 11.4/3. | Either strike the note or add reference to 11.4/3 and/or mention of "friend T;". | editor | ACCEPTED | |
UK 93 [454] |
7.1.6.3 ¶ Grammar before para 1 | In the third production, "enum ::opt nested-name-specifieropt identifier", enum should not be in italics; its referring to the enum keyword. | Change to keyword font | editor | ACCEPTED | |
UK 94 [383] |
7.1.6.4 ¶ 1 | The auto type-specifier signifies that the type of an object being declared shall be deduced from its initializer or specified explicitly at the end of a function declarator. A function declarator does not declare an object. | The auto type-specifier signifies that the type of an object being declared shall be deduced from its initializer or that the return type of a function is specified explicitly at the end of a function declarator. | editor | ACCEPTED | |
UK 95 [396] |
7.1.6.4 ¶ 4 | (See also c++std-core-13583) This paragraph allows auto "in the type-specifier-seq in a new-type-id (5.3.4)" (and nowhere else not listed). Specifically, it isn't allowed in a type-id in a new-expression. That allows "new auto (42)", but not "new (auto)(42)". However, 5.3.4/2 suggests the latter should be allowed "If the auto type-specifier appears in the type-specifier-seq of a new-type-id or type-id of a new-expression ...". The inconsistency should be resolved, ideally in favour of allowing both forms. | Change "in a new-type-id" to "in a new-type-id or type-id in a new-expression". | CWG | 746 | |
FR 24 | 7.1.6.4 [auto specifier] |
Now that 'auto' is finally
used in its most obvious sense to state `deduce the type of
this variable from initializer', it should also be allowed
in template parameter declarations, as in
template<auto n> struct X { /* … */ }; X<903> x; X<&Widget::callback> y; instead of the current, often verbose and cumbersome template<typename T, T n> struct X { /* … */ }; X<int,903> x; X<void (Widget::*)(),&Widget::callback> y; We understand that 'auto' is used in 14.1/18 in a different way (for constrained template), but that usable appears very strange syntax, unnatural and does not fit well with the usage in this section. |
CWG |
REJECTED
There was no consensus to make this change at this point in the standardization process. |
||
US 38 | 7.2 ¶ 1 | The discussion of attribute specifiers should be a separate paragraph. | editor |
REJECTED
It's okay where it is. It's a list of constraints on the grammar. |
||
US 39 | 7.2 ¶ 2 | The paragraph says in part "An opaque-enum-declaration declaring an unscoped enumeration shall not omit the enum-base." This statement implies that the base may be omitted for scoped enumerations, which is somewhat inconsistent with paragraph 3 and somewhat consistent with paragraph 5. | As this implication leaves no representation, it should be either affirmed here or the statement should be expanded. Perhaps a note is warranted. | CWG |
REJECTED
The current specification is consistent. A scoped enumeration with an omitted enum-base has an underlying type of int. |
|
JP 13 | 7.2 ¶ para 3 | In the description for an unscoped enumeration, enum-base in redeclaration must be the same underlying type as in the 1st declaration, but it is not described explicitly, while it is referred that all enum-bases in redeclarations must specify the same underlying type. | Replace the description, "same underlying type", with "same as underlying type of (previous) declaration." | editor | ACCEPTED | |
UK 96 [384] |
7.2 ¶ 7 | enum E { }; What are the values of E? It has neither a smallest nor largest enumerator, so paragraph 7 doesn't help. (Paragraph 6 indicates that the underlying type is as if E had a single enumerator with value 0, but that does not define the values of E.) | Add a second sentence to paragraph 7 (before "Otherwise"): "If the enumerator-list is empty, 0 is the only value of the enumeration." | CWG | 628 | |
UK 97 [385] |
7.2 ¶ 9 | Missing punctuation after "blue" in: "The possible values of an object of type color are red, yellow, green, blue these values can be converted ..." | Add a semicolon: "The possible values of an object of type color are red, yellow, green, blue; these values can be converted ..." | editor | ACCEPTED | |
UK 98 [402] |
20.5.6 | It would be useful to be able to determine the underlying type of an arbitrary enumeration type. This would allow safe casting to an integral type (especially needed for scoped enums, which do not promote), and would allow use of numeric_limits. In general it makes generic programming with enumerations easier. | Add a TransformationTrait to 20.5.6 that returns the underlying type of an enumeration type. | LWG | 1055 |
Section reference corrected from 7.2. Note that the EnumerationType concept (14.9.4 paragraph 41) has a member underlying_type that provides this information. |
UK 99 [421] |
7.2 ¶ 3 | It is unclear whether an enumeration type is complete after an opaque-enum-declaration. This paragraph only says so in a note, and the general rule in 3.9/5 ("Incompletely-defined object types ... are incomplete types") is unclear in this situation. | Move "an enumeration declared by an opaque-enum-declaration ... is a complete type" from the note to normative text. | CWG |
REJECTED
The definition of an incomplete type in 3.9 paragraph 5 does not mention opaque enumeration types; therefore, an opaque enumeration type is not an incomplete type. |
|
JP 14 | 7.3.1 |
The description of the behavior when a member that was defined with same name in other namespace was referred.
|
Add the description of the behavior when a member that was defined with same name in other namespace was referred. | CWG | 812 | |
UK 100 [387] |
7.3.3 ¶ 10 and 13 | Para 10 says "A using-declaration is a declaration and can therefore be used repeatedly where (and only where) multiple declarations are allowed." Para 13 says "Since a using-declaration is a declaration, the restrictions on declarations of the same name in the same declarative region (3.3) also apply to using-declarations." These appear to be saying exactly the same thing. | Delete para 10, moving its example into para 13. | CWG |
REJECTED
The two citations are different: paragraph 10 is referring to repeated declarations of the same entity, while paragraph 13 deals with declarations of different entities with the same name. |
|
UK 101 [388] |
7.3.3 ¶ 20 | If a using-declaration uses the keyword typename and specifies a dependent name (14.6.2), the name introduced by the using-declaration is treated as a typedef-name (7.1.3). That doesn't specify at all what the effect of using typename with a non-dependent name is. Is it allowed? What about outside any template? What if the name isn't a type? (14.6/4 doesn't cover this, I think.) | Allow typename for non-dependent names iff they refer to a type. | CWG | 813 | |
DE 12 | 7.3.3 ¶ p15 | DE-12 Overriding and hiding of member functions named in using-declarations should consider ref-qualifiers, because they are part of the function type. | CWG | 731 | ||
FR 25 | 7.3.3 [The using declaration] ¶ Para 21 | The syntax for concept map alias is unnecessarily both confused and verbose. |
We strongly suggest
simplifications, e.g.
using N1::C<int>; that fits well with existing constructs. The syntactic complexity is too high for a new feature presumably designed to support sound programming. |
CWG |
REJECTED
The complexity is considered necessary to allow concepts and concept maps to be members of different namespaces, as illustrated in the example. |
|
UK 102 [389] |
7.3.4 ¶ 6 | This paragraph says "If name lookup finds a declaration for a name in two different namespaces, and the declarations do not declare the same entity and do not declare functions, the use of the name is ill-formed." But the example uses declaration of functions, so is not covered by this paragraph. | Move the example to paragraph 7, and/or replace it with an appropriate example. | editor |
ACCEPTED with MODIFICATIONS
Changed the example slightly to make it clearer. |
|
US 40 | 7.6 |
The
list of attributes is missing an attribute to indicate that
a function with a throw() (throws nothing)
clause need not have the unexpected() catch
clause generated. This attribute was a motivating example
for the attribute syntax, and its omission is
surprising.
|
Add the attribute. | CWG | 814 | |
US 41 | 7.6 | A common problem is unintentionally declaring a new virtual member function instead of overriding a base virtual member function. | An attribute stating intent to override would enable better diagnostics. | CWG | We are awaiting a full proposal to evaluate. | |
FR 26 | 7.6[Attributes] | Are they part of object types or not? The section does not appear to indicate that clearly. | CWG | |||
FI 1 | 7.6 | Add override-attribute for functions in order to avoid mistakes when overriding functions. | See override-attribute.doc, override-attribute.ppt | CWG | We are awaiting a full proposal to evaluate. | |
FR 27 | 7.6.1 |
This section specifies that
no name lookup is performed on any identifier contained in
an attribute-token. This in particular implies that, for
example, it is impossible to define a template class
parameterized by its alignment. That restriction is
unacceptable.
The original alignment proposal made that useful construct possible. Furthermore paragraph 7.6.1/2 appears contradictory with the rest of that section -- since no name lookup is performed, how a 'type-id'is determined? |
CWG |
REJECTED
The comment is based on a misunderstanding: what is not looked up is, informally, the attribute name (e.g., align), including the attribute-namespace, if any. The attribute-argument-clause, if present, can have identifiers that are looked up, depending on the attribute. |
||
UK 103 [397] |
7.6.1 | Attributes should support pack expansion. For example, this would be extremely useful with the align attribute, directly supporting the (removed) functionality of aligned_union. NOte that aligned_union was removed as varaiant-unions were considered a complete replacement - however this is not true for variadic templates. Adding this support to attributes would remove the remaining need, and support similar attributes in the future. | Add: attribute... to the grammar for attribute-list Add to list in 14.5.3p4: "In an attribute-list(7.6.1); the pattern is an attribute." | CWG | 815 | |
UK 104 [398] |
7.6.1 ¶ 1 | It is helpful for each subclause to contain a short paragraph introducing its intent an purpose. 7.6 has such a paragraph, but it is nested under a more specific subsection. | 7.6.1p1 should move up one level to become 7.6p1. There grammar should remain under 7.6.1 | editor |
REJECTED
ISO editing guidelines insist that only leaf clauses have text. Moving this sentence to 7.6 would violate that rule (not that the standard doesn't violate it in many places, but we shouldn't make this worse now). |
|
UK 105 [448] |
7.6.1 ¶ 1 | Allowing only one level of namespaces in attributes seems unnecessarily limiting. | To: attribute-scoped-token: attribute-namespace :: identifier add attribute-namespace :: attribute-scoped-token | CWG |
REJECTED
This suggestion was discussed earlier and rejected, and there was no consensus for reconsideration. |
|
UK 106 [391] |
7.6.2 ¶ 1 | Extensive use of alignment and related terms without cross reference. | Add cross-reference to 3.11. | editor | ACCEPTED | |
JP 15 | 7.6.2 | An abbreviation of 7.6.2 should be “[decl.attr.align]” instead of “[dcl.align]”. Section name “[dcl.align]” is not consistent with others, because others in 7.6 are the form of “dcl.attr.*”. In N2761, the section name of 7.1.7 had been changed from “[dcl.align]” to “[dcl.attr.align]”, but in N2800 it was reverted to “[dcl.align]” along with a change of section number, 7.1.7 to 7.6.2. | Change "[dcl.align]" of 7.6.2 to "[decl.attr.align]". | editor |
REJECTED
Tags don't change. This section used to be somewhere else, and the tag was appropriate there. |
|
UK 107 [399] |
7.6.3 | While undefined behaviour might be the best we can guarantee, it would be helpful to encourage implementations to diagnose function definitions that might execute a return. | Adda a [Note : implementations are encouraged to issue a diagnostic where the definition of a function marked [[noreturn]] might execute a return statement -- end note] | editor | ACCEPTED | |
UK 108 [401] |
7.6.4 ¶ 2 | It is unclear why no diagnostic is required for an easily detectable violation. It is even more surprising that the associated footnote mandates behaviour for an ill-formed program. | Strike "no diagnostic required" and the associated footnote. | CWG | 816 | |
US 42 | 7.6.4 |
The
meaning of the [[final]] attribute applied
to classes is inconsistent with other languages and not
desirable in its own right.
|
Modify the semantics of [[final]] applied to
classes. See the attached paper "Issues with the C++
Standard" under Chapter 7 "Meaning of
[[final]] attribute applied to
classes".
|
CWG | 817 | |
UK 109 [392] |
7.6.5 ¶ 4 | The example code refers in comments to "Compilation unit" A and B. The term should be "Translation unit" (2/1) | Replace "Compilation" with "Translation" in two places | editor | ACCEPTED | |
UK 110 [393] |
7.6.5 ¶ 4 | The code in the example (compilation unit A) has: "foo_head[i].load(memory_order_consume)". foo_head[i] is of type foo *, so it does not have a load member. | Change the type of foo_head to atomic<foo *>[]. | CWG |
ACCEPTED
See paper N2841 |
|
US 43 | 8 |
With the introduction of late-specified return types for
functions and lambda expressions, we now have three
different syntaxes for declaring functions. The -> late
declaration is used in two. The auto keyword
is used in one, but also used differently in variable
definitions.
|
Some simplification is needed. | CWG | ||
UK 111 [457] |
8.3.5 ¶ 13 | Example missing closing bracket in template<typename... T> void f(T (* ...t)(int, int); | Add closing bracket like this: template<typename... T> void f(T (* ...t)(int, int)); | editor | ACCEPTED | |
US 44 | 8.3.5 ¶ 13 | In the Example, "template void f(T (* ...t)(int, int);" is missing a close parenthesis. | It presumably should read: "template void f(T (* ...t))(int, int);". | editor | ACCEPTED | |
US 45 | 8.3.5 ¶ 13 |
At present,
function parameter packs can only occur at the end of a
parameter-declaration-list. This restriction unnecessarily
prohibits uses of function parameter packs in cases where
template argument deduction isn’t needed, e.g.,
template<class... T> struct X { }; template<class... T1, class... T2> struct X<pair<T1, T2>...> { void f(T1..., T2...); }; More importantly, this restriction is inconsistent with the way pack expansions are handled. For example, this template is well-formed (but X<T..., int> is a non-deduced context): template<class... T> void f(X<T..., int>); Therefore, the restriction that limits function parameter packs to the end of the parameter-declaration-list should be removed. Instead, function parameter packs not at the end of the parameter-declaration-list should be considered non-deduced contexts. |
In 8.3.5p13,
remove the sentence “A function
parameter pack, if present, shall occur at the end of the
parameter-declaration-list.”
In 14.8.2.1p1, replace the phrase “For a function parameter pack” with “For a function parameter pack that occurs at the end of a parameter-declaration-list”. Replace the note text “A function parameter pack can only occur at the end of a parameter-declaration-list (8.3.5).” with “A function parameter pack that does not occur at the end of a parameter-declaration-list is a non-deduced context.” In 14.8.2.5p5, add a new bullet: “A function parameter pack that does not occur at the end of its parameter-declaration-list.” In 14.8.2.5p10, replace “If the parameter-declaration corresponding to Pi is a function parameter pack” with “If the parameter-declaration corresponding to Pi is a function parameter pack and Pi occurs at the end of the parameter-declaration-list”. Replace the note text “A function parameter pack can only occur at the end of a parameter-declaration-list (8.3.5).” with “A function parameter pack that does not occur at the end of a parameter-declaration-list is a non-deduced context.” |
CWG | 818 | |
DE 13 | 8.4 ¶ p2 | DE-13 The second paragraph, quoting the grammar for the declarator of a function declaration, is not considering late-specified return types and attributes. | Properly quote the grammar from 8.3.5. | CWG | 732 | |
JP 16 | 8.5 ¶ 15th para, 1st line |
Typo, duplicated "in"
"The initialization that occurs in in the forms" |
Remove one. | editor | ACCEPTED | |
US 46 | 8.5.3 |
The ability for
an rvalue reference to bind to an lvalue opens a
type-safety hole that becomes very dangerous with concepts.
For example, consider vector’s push_back operation:
requires MoveConstructible<T> void push_back(T&&); requires CopyConstructible<T> void push_back(const T&); For a copy-constructible T (which is also move-constructible), push_back does the right thing. However, if T is something that is move-constructible (e.g., unique_ptr<int>), the second overload is removed from considered (it is effectively SFINAE’d away), so only the first overload remains. Therefore, one could accidentally call push_back with an lvalue of type T, and push_back would silently move from the lvalue. The same problem occurs without concepts (albeit less frequently). |
Prohibit rvalue
references from binding to lvalues.
Unfortunately this change will break some current use cases of rvalue reference including the use of rvalue streams, and of the forward function itself. To resolve this we may want to consider three types of references:
|
CWG |
ACCEPTED
See paper N2844 |
|
US 49 | 8.5.4 ¶ 6 | In the Example, the comments could be improved. | See the attached paper "Issues with the C++ Standard" under "Editorial Issues" and "8.5.4/6". | editor | ||
UK 112 [440] |
9 ¶ 4-9 | We now have concepts that should (or should not?) map to the terms described in Clause 9 - these should be at least referenced. | Add appropriate forward references to 14.9.4 | editor |
REJECTED
This seems excessive. If people want to know about concepts they should read about concepts. |
|
UK 113 [250] |
9.4.2 ¶ 3 | Mis-applied edit from the paper n2756 | The term 'constant-initializer' should have been struck out when replaced by brace-or-equal-initializer. There are two occurrences in this paragraph | editor | ACCEPTED | |
US 50 | 12.1, 12.4, 12.8 |
Implicitly-declared default constructors, destructors, copy
constructors, and copy assignment operators are deleted
when their definitions would be ill-formed. However, unlike
with overloading and template argument deduction, access
control is performed as part of the check for making one of
these special function deleted. This inconsistency should
be removed.
This change would sacrifice some backward compatibility in favor of consistency. With the current wording, checking that the following class ‘A’ is CopyConstructible would proceed without error (it is not CopyConstructible): class A { A(const A&); }; With the proposed change, testing whether A is CopyConstructible would produce a diagnostic. To fix the problem, the user would have to use a deleted function: class A { A(const A&) = delete; }; |
In 12.1p5,
remove the phrase “or inaccessible from
the implicitly-declared default constructor”.
In 12.4p3, remove the phrase “or a destructor that is inaccessible from the implicitly-declared destructor,” and the phrase “or a destructor that is inaccessible from the implicitly-declared destructor”. In 12.8p5, remove the phrase “or inaccessible from the implicitly-declared copy constructor” from the two places it occurs. In 12.8p10, remove the phrase “or inaccessible from the implicitly-declared copy assignment operator” from the two places it occurs. |
CWG | 819 | |
FR 28 | 12.6.1 [Explicit initialization] | This section, in particular the example with `g' appears contradictory with the syntax for uniform initialization. | CWG |
ACCEPTED
See paper N2841 |
||
US 51 | 12.6.2 ¶ 2 | The discussion of delegating constructors should be in its own paragraph. | editor | ACCEPTED | ||
UK 114 [167] |
12.6.2 ¶ 1 | Despite all the attempts to unify initialization syntax, it is still not possible to copy-initialize base classes or non-static data members, which means the explicit keyword cannot have a bearing during evaluation of a constructor. A minimal addition to the grammar, allowing an optional = between the mem-initializer-id and braced-init-list would allow the user to choose between copy and direct initialization | Ammend the grammar for mem-initializer: mem-initializer-id =OPT braced-init-list Extend p3 to allow for Copy Initialization if the optional = is present: 3 The expression-list or braced-init-list in a mem-initializer is used to initialize the base class or non-static data member subobject denoted by the mem-initializer-id according to the initialization rules of 8.5 for direct-initialization, OR COPY-INITIALIZATION IF THE OPTIONAL = IS PRESENT BETWEEN THE MEM-INITIALIZER-ID and the BRACED-INIT-LIST. [Example:... | CWG |
REJECTED
There was no consensus for making this changes. |
|
US 52 | 13.5.8 ¶ ¶ 5 | A word is misspelled. | Change “shal” to “shall”. | editor | ACCEPTED | |
UK 115 [432] |
14 ¶ 6-11 | Exported templates were a great idea that is generally understood to have failed. In the decade since the standard was adopted, only one implementation has appeared. No current vendors appear interested in creating another. We tentatively suggest this makes the feature ripe for deprecation. Our main concern with deprecation is that it might turn out that exported constrained templates become an important compile-time optimization, as the constraints would be checked once in the exported definition and not in each translation unit consuming the exported declarations. | Consider deprecating exported templates, but no action yet. Examine interaction with constrained templates, and see if other more appropriate mechanism will support compile-time optimization. | CWG | 820 | |
UK 116 [434] |
14 ¶ 6-11 | Is it possible to export a concept map template? The current wording suggests it is possible, but it is not entirely clear what it would mean. | Either prohibit exporting concept map templates, or more directly address what it means to export a concept map. | CWG | 821 | |
UK 117 [430] |
14 ¶ 2 | It would be nice to allow template alias within a function scope, and possibly a scoped concept map. As these affect name lookup and resolution, rather than defining new callable code, they are not seen to present the same problems that prevented class and function templates in the past. | Allow template aliases to be declared inside a function scope, and consider scoped concept maps. | CWG | 822 | |
UK 118 [431] |
14 ¶ 6-11 | Exported templates are a complicated feature with surprisingly little text. To make this important text more visible, split it off into its own subclause [temp.export] | Create a new subclause [temp.export] containing 14p6-11. Move 14p12 ahead of this subclause. | editor | ACCEPTED | |
UK 119 [433] |
14 ¶ 4 | Does a concept map have linkage? Reading this paragraph and 3.5 suggests a concept map template has external linkage, but not a 'regular' concept map. Believe this is an oversight that the linkage words were not updated to provide an exception, rather than linkage of concept maps is intended. | Add an exception that concept map templates have no linkage, or add concept maps to the list of entities with linkage in 3.5 | CWG | 791 | |
UK 120 [422] |
14.1 ¶ 9 | As this is the first time the phrase “parameter pack” appears in Ch 14 I would like to see the section 8.3.5 referenced here (as well as in 14.1p17). | Insert “(8.3.5)” after “parameter pack” | editor | ACCEPTED | |
UK 121 [423] |
14.1 ¶ 18 | The example (that follows the normative text) has no begin example marker | Prefix the example code with "[ Example:" | editor | ACCEPTED | |
FR 29 | 14.3 [Template arguments] | Constant expressions of any literal type should be allowed as template arguments. | CWG | 823 | ||
US 53 | 14.5.1 ¶ 5 |
If the
requirements of a constrained member that is a copy
constructor, copy assignment operator, or destructor are
not satisfied, then that user-declared special function
will not exist. It appears that, in this case, the special
function will then be implicitly defined, which is likely to either (a) fail to compile
or (b) produce a function with the wrong semantics. For
example:
template<ObjectType T> class vector { T* first, last, end; public: requires CopyConstructible<T> vector(const vector&); }; If instantiated with a type that is not CopyConstructible, vector will get an implicitly-defined copy constructor that performs a copy of the pointers. |
Add to 14.5.1p5:
If the constrained member is a copy constructor (12.8), destructor (12.4), or copy assignment operator and its template requirements are not satisfied, then a copy constructor, destructor, or copy assignment operator, respectively, with the same signature as the constrained member (after substituting the class template’s template arguments for its template parameters) will be declared as a deleted function (8.4). |
CWG | 824 | |
UK 122 [426] |
14.5.3 ¶ 4 | Variadic templates should be supported in axioms. There are axioms in the library that rely on this feature, such as the FrontEmplacement axiom in FrontEmplacementContainer (23.1.6.1p10) | Add clarification in p2 that function parameter packs can be used to declare axioms, much like p1 clarifies they can be used to declare concepts as well as templates. | CWG | ||
FR 30 | 14.5.7 [Template aliases] |
When are two template alias
names equivalent?
E.g. given
template<template<class> class> struct X { }; template<typename,typename> struct Y { }; template<typename T> using Z1 = Y<int,T>; template<typename T> using Z2 = Y<int,T>; Are the types X<Z1> and X<Z2> equivalent? We would suggest yes (since Z1<T> and Z2<T> are the same for all T), but we do not see any wording to that effect. |
CWG |
REJECTED
This is already clear in the Standard; see 14.5.7 paragraph 2 and 14.4 paragraph 1. The last example in 14.4 paragraph 1 is very similar to the one given here. |
||
JP 17 | 14.7.2 ¶ 2nd para, 15th line |
Typo.
if that namespace is inline, any namespace from its enclosing namespace set. should be if that namespace is inline, any namespace forming its enclosing namespace set. |
Replace "from" with "forming" | CWG |
REJECTED
The current wording is correct. |
|
DE 14 | 14.7.3 ¶ p1 | DE-14 The bulleted list neither addresses "member function template of a class" nor "member class template of a class". | Add the respective bullets. | CWG | 730 | |
JP 18 | 14.7.3 ¶ 2nd para, 2nd line |
Typo,
any namespace from its enclosing namespace set should be any namespace forming its enclosing namespace set |
Replace "from" with "forming" | CWG |
REJECTED
The current wording is correct. |
|
JP 19 | 14.8.2 ¶ 6th para, 1st line |
Typo, duplicated "is"
"At certain points in the template argument deduction process it is is necessary" |
Remove one | editor | ACCEPTED | |
US 54 | 14.9 [concept], 14.10 [temp. constrained] | Concepts is of course the largest new feature in C++0x (in terms of new text inserted into the wording), and already we have found some significant defects with it. So far nothing devastating has been found, but more time is needed to shake more bugs out. | I propose no specific change here. | CWG | ACCEPTED | |
US 55 | 14.9.1 ¶ ¶ 6 | The paragraph number is in the wrong place, causing a grammar rule to be indented more than its fellows. | Move the paragraph number so as to follow the grammar rules, thus numbering the single sentence, “The body of a concept … .” | editor | ACCEPTED | |
US 56 | 14.9.1 ¶ ¶ 6 | The sentence contains two references to 14.9.1.3 [concept.req]. | Change the second such reference (at the end of the sentence) to 14.9.1.4 [concept.axiom]. | editor |
REJECTED
Yes; they apply to different terms. |
|
US 57 | 14.9.1.4 ¶ ¶ 3 | A word is misplaced, changing the intended meaning. | Change “only find … if” to “find … only if”. | editor |
REJECTED
These mean the same thing, and the latter is stilted. |
|
US 58 | 14.9.1.4 ¶ ¶ 3 | The listed phrases are not grammatically parallel. | Insert “in” before “one” so as to obtain “... in the concept, in one of its less refined concepts, or in an associated requirement.” | editor | ACCEPTED | |
US 59 | 14.9.1.4 | Axioms are under-specified and provide little benefit to programmers, so they should be removed from the working paper. The optimizations permitted by axioms (see 14.9.1.4p4-5) are not compulsory (and, therefore, programmers cannot rely on them) and the semantics expressed by axioms cannot be verified by any implementation. The resulting specification has lead to great confusion (see the reflector thread “Are floating point types Regular?” starting with c++std-lib-22717). Given the level of confusion and the inability to verify the correctness of axioms, it is likely that many axioms written by programmers (including those specified in the candidate draft) will be incorrect. |
Remove clause
14.9.1.4 [concept.axiom]
In 2.11p1, remove “axiom” from the list of keywords. In 14.5.8p7, remove “, or if the resulting concept map fails to satisfy the axioms of the corresponding concept” In 14.9.1p6, remove axiom-definition from the list of grammar productions for concept-member-specifier. Remove “, and axioms” from the final sentence, and instead “and” prior to “associated requirements” in the final sentence. Remove paragraph 14 of clause 14.9.2. In 14.10.1p6, remove the sentence, “When the concept-instance-alias-def appears in the optional requires-clause of an axiom-definition (14.9.1.4), the potential scope of the identifier begins at its point of declaration and terminates at the end of the axiom-definition.” In clauses 20.2.5, 20.2.8, 23.1.6.1, 23.1.6.2, and 24.1.4, remove the axiom-definitions and replace them with paragraphs (denoted Requires, Postconditions, or Effects, as appropriate) that express the intended semantics of the concepts from which the axiom-definitions were removed. In 24.1.4p2, replace the word “axiom” with “condition.” |
CWG | ||
FR 31 | 14.9.1.4 [Axioms] |
This section states that an
axiom-definition defines a new semantics axiom but is
unusually vague as to what those semantics might be.
The use of the '==' and '!=' with completely new semantics, unrelated to anything we have seen before in C++ is both unwise and confusing, especially if the types involved in the expressions happen to have operator== and operator!= defined. We strongly suggest use of different tokens, e.g. , and opposed to this obscure usage/overload. The description is very vague. How many times is an implementation permitted to replace one expression by another one when they have side effects? |
CWG | |||
DE 15 | 14.9.1.4 | DE-15 There is no implementation experience for axioms. Use of axioms is an area of active scientific research. It is likely that syntax changes will become necessary to make good use of axioms; having the syntax space already crowded is unhelpful. Axioms ought to be useful in concepts applicable to floating-point types (such as EqualityComparable), but IEEE floating-point types have special values such as NaN violating the axioms. | Remove section 14.9.1.4 and any reference to axioms in the rest of the proposed standard other than the keyword reservation in section 2.11. | CWG | ||
UK 123 [248] |
14.9.1.4 | auto concepts and axioms are incompatible. An axiom defines the semantics of an operaton or set of operations that describes the run time behaviour. A concept describes purely syntactic requirements at compile time. Where an auto concept will match anything that meets the syntax requirements, there is no way to know if the axioms will be met or not, and no way to opt out via some kind of negative concept map. | Add a paragraph making axioms ill-formed inside an auto concept. | CWG | ||
UK 124 [288] |
14.9.1.4 ¶ 6 | Spelling mistake, double-e in were. | weere -> were | editor | ACCEPTED | |
UK 125 [289] |
14.9.1.4 ¶ 2 | The implicit equality comparison operator available to axioms has no semantic. It is not clear what expressing the condition if( a == b ) { conditional-axiom } would mean if a and b are not truly EqualityComparable. We suspect the intent of the 'implicit defefinition' is to support declaring the equivalence of statements, a context where the operator will not actually be evaluated. | Define the semantics of the implicitly declared comparison operators, or restrict their usage to declaring equivalence between statements. | CWG | ||
UK 126 [438] |
14.9.4 ¶ 41 | This paragraph contains the only definition of the underlying_type member - but it's a note, so not normative. | Move the second sentence to the Requires clause in paragraph 42. | editor | ACCEPTED | |
UK 127 [118] |
14.9.4 | Provide a diagram clearly showing refinement relationship between the different support concepts. Several were created during development of this clause and they were very helpful. | Provide the diagram. | editor | ||
UK 128 [435] |
14.9.4 ¶ 4 | It is surprising for many people that non-copyable move-only types can be used with a return statement, and so Returnable does not always imply CopyConstructible. | A non-normative note: [Note: 'move only' types that are constructible from rvalue references may be Returnable, but not CopyConstructible(20.1.8) - end note] | editor | ACCEPTED | |
JP 20 | 14.9.4 ¶ 2nd para | Trivially copyable type was added in “3.9 Types”, so we think that it is necessary to add concept to trivially copyable type like “TriviallyCopyableType”. | Add TriviallyCopyableType that is trivially copyable type as concept. | CWG | 825 | |
UK 129 [128] |
14.10.1, 20.1.2 | It should be possible to support boolean constant expressions as requirements without resorting to defining the True concept in the library. Boolean expressions are very likely to be constraints when deadline with non-type template parameters and variadic templates, and constraints in these cases should feel just as natural as constraints on the type system. | Remove the True concept and library subclause 20.1.2. Provide support in 14.10.1 for boolean constant expressions as constraints. This may involve overloading the true keyword to disambiguate but ideally would not. | CWG | 826 | |
US 60 | 14.10.1 ¶ 1 | The use of && as the separator for a list of requirements has shown itself to be a serious teachability problem. The mental model behind ‘&&’ treats concepts as simple predicates, which ignores the role of concepts in type-checking templates. The more programmers read into the ‘&&’ (and especially try to fake || with && and !), the harder it is for them to understand the role of concept maps. Simply changing the separator to ‘,’ would eliminate a significant source of confusion. |
Replace
requirement-list: requirement-list ... [opt] && requirement requirement ... [opt] with requirement-list requirement-list ...[opt] , requirement requirement ... [opt] In 14.5.4p6, replace the first sentence with: The instantiation of an expansion produces a comma-separated list E1, E2, ..., EN, where N is the number of elements in the pack expansion parameters. |
CWG | 827 | |
UK 130 [32] |
15.1 ¶ 4 | With the new crrent_exception API it is possible to capture a reference to an exception that will outlive its last active handler. That is in conflict with the sentance "When the last remaining active handler for the exception exits by any means other than throw; the temporary object is destroyed and the implementation may deallocate the memory for the temporary object;" | Update sentence to allow for exceptions held in excpetion_ptr objects. | CWG | 828 | |
UK 131 [34] |
15.3 ¶ 3 | A handler catching its parameter by rvalue-reference is syntactically valid, but will never be activated. | Disallow handlers catching by rvalue-reference. | CWG |
ACCEPTED
See paper N2844 |
|
UK 132 [36] |
15.3 ¶ 16 | There are obscure cases whrere a copy constructor is not usually the best match to copy-initialize an object, e.g. A converting constructor template taking arguments by non-const reference. A footnote explaining such cases would be helpful, or the sentance could be rewritten using copy-initialization instead of directly calling a copy constructor. | Rewite using copy-initialization rather than directly invoking the copy constructor | CWG |
ACCEPTED
See paper N2841 |
|
UK 133 [37] |
15.4 ¶ 2 | Template aliases have the same semantics as a typedef so should also be disallowed | add "or alias-declaration" after "shall not appear in a typedef declaration". | CWG |
ACCEPTED
See paper N2841 |
|
UK 134 [38] |
15.4 ¶ 6 | The sentance "An exception-specification can also include the class std::bad_exception (18.7.2.1)." is redundant. | Either strike the quoted sentance, or add a note explaining why it is worth calling special attention to this class. | editor | ACCEPTED | |
UK 135 [39] |
15.4 ¶ 8 | Unclear if std::unexpected is called before or after the function arguments have been destroyed | Clarify the sequence of calling unexpected with respect to interesting objects, such as function arguments or partially constructed bases and members when called from a constructor or destructor | CWG | 829 | |
UK 136 [40] |
15.4 | Exception specifications have proven close to worthless in practice, while adding a measurable overhead to programs. The feature should be deprecated. The one exception to the rule is the empty throw specification which could serve a legitimate optimizing role if the requirement to call the runtime unexpected mechanism was relaxed in this case. | Move 15.4 and the parts of 15.5 that refer to it to Appendix D. Replace 15.4 with a simpler specification for empty throw specifications, where the std::unexpected call is conditionally supported allowing vendors to choose between optimizing and providing runtime checks. Ideally require vendors to provide a mode where the runtime checks are always disabled. | CWG | 830 | |
UK 137 [44] |
15.5 | There is no mention of the current_exception API which can extend the lifetime of an exception object. There should at least be a forward reference to the library clause 18.7.5 | Add another paragraph outlining 18.7.5 and the ability of an exception_ptr to extend the lifetime of an exception object | editor | ACCEPTED | |
UK 138 [41] |
15.5.1 ¶ 1 | The third bullet is redundant with the first, as it is a subset of the same conditions. | Merge the third bullet into the first bullet as a note or example. | editor |
REJECTED
They're subtly different. The first bullet is about calls made to create, copy, and destroy the exception object itself. The third bullet is about destructors of stack objects during stack unwinding. |
|
UK 139 [42] |
15.5.1 ¶ 1 | According to the first bullet it is perfectly alright for a library function to exit by throwing an exception during stack unwinding, This is clearly not true. | Strike the word 'user' from the first bullet point. | CWG |
ACCEPTED
See paper N2841 |
|
UK 140 [43] |
15.5.2 ¶ 2 | The detailed specification can fool people into thinking an exception will automatically be translated into bad_exception, where the default behaviour of std::unexcepted is to immediately call std::terminate(); | Add a note highlighting the default behaviour of std::unexpected if the user does not supply a hander-function | editor | ACCEPTED | |
UK 141 [45] |
15.6 | This whole subclause is redundant due to 15.1p5 and 15.3p17 | Strike 15.6 | editor | ACCEPTED | |
UK 142 [455] |
16.3.5 ¶ 3 | This paragraph opens with "[ Note" but has no corresponding "end note ]" | Add "end note ]" | editor | ACCEPTED | |
UK 143 [456] |
16.3.5 ¶ 7 | Example uses #define t(x,y.z) x ## y ## z | Change "x,y.z" to "x,y,z" | editor | ACCEPTED | |
US 2 | 17-30 |
The
active issues identified in WG21 N2806, C++ Standard
Library Active Issues, must be addressed and appropriate
action taken.
http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-active.html |
Appropriate action would include making changes to the CD, identifying an issue as not requiring a change to the CD, or deferring the issue to a later point in time. | LWG | ACCEPTED | |
FR 2 | General Comment ¶ Library | The adoption of the library `constexpr' proposal was not reflected in the draft, despite formal WG21 committee vote. | FR 2 | editor | ||
US 61 | 17 onward | The concepts core language feature is applied to only some of the Standard Library clauses, and even then not always consistently. | Review all clauses of the Standard Library, and consistently apply concept technology wherever possible and appropriate. The proposed wording in WG21 N2781 exemplifies the necessary level of detail. | LWG | ACCEPTED | |
CA 2 | 17 Library | “Concepts” are a significant new addition to the language, but are not exploited uniformly in the library as documented in CD 14882. | Fix the standard library so that “Concepts” are used appropriately in the library. | LWG | ACCEPTED | |
US 62 | 17-30 | Provide concepts and requirements clauses for all standard library templates | LWG | ACCEPTED | ||
US 63 | 17-30 |
The behavior of the library in the presence of threads
is incompletely specified.
For example, if thread 1 assigns to X, then writes data to file f, which is read by thread 2, and then accesses variable X, is thread 2 guaranteed to be able to see the value assigned to X by thread 1? In other words, does the write of the data "happen before" the read? Another example: does simultaneous access using operator at() to different characters in the same non-const string really introduce a data race? |
LWG | |||
DE 2 | 17 through 30 | DE-2 Marking a constructor with "explicit" has semantics even for a constructor with zero or several parameters: Such a constructor cannot be used with list-initialization in a copy-initialization context, see 13.3.1.7. The standard library apparently has not been reviewed for marking non-single-parameter constructors as "explicit". | Consider marking zero-parameter and multi-parameter constructors "explicit" in classes that have at least one constructor marked "explicit" and that do not have an initializer-list constructor. | LWG | ||
JP 21 | 21.2, 21.4, 27.2, 27.6, 27.7, 27.8.1, 28.4 |
Support of char16_t/char32_t is insufficient. The basic_xxx
classes of <iostream>, <fstream>,
<regex>, etc. does not have typedefs for
char16_t/char32_t.
Functions such as stoi, to_string in ‘21.4 Numeric Conversion’ does not support char16_t/char32_t types. |
Add commented
lines corresponding to char16_t, char32_t.
21.2 paragraph1 namespace std { ... // 21.4: numeric conversions ... int stoi(const u16string& str, size_t *idx = 0, int base = 10); long stol(const u16string& str, size_t *idx = 0, int base = 10); unsigned long stoul(const u16string& str, size_t *idx = 0, int base = 10); long long stoll(const u16string& str, size_t *idx = 0, int base = 10); unsigned long long stoull(const u16string& str, size_t *idx = 0, int base = 10); float stof(const u16string& str, size_t *idx = 0); double stod(const u16string& str, size_t *idx = 0); long double stold(const u16string& str, size_t *idx = 0); u16string to_u16string(long long val); u16string to_u16string(unsigned long long val); u16string to_u16string(long double val); int stoi(const u32string& str, size_t *idx = 0, int base = 10); long stol(const u32string& str, size_t *idx = 0, int base = 10); unsigned long stoul(const u32string& str, size_t *idx = 0, int base = 10); long long stoll(const u32string& str, size_t *idx = 0, int base = 10); unsigned long long stoull(const u32string& str, size_t *idx = 0, int base = 10); float stof(const u32string& str, size_t *idx = 0); double stod(const u32string& str, size_t *idx = 0); long double stold(const u32string& str, size_t *idx = 0); u32string to_u32string(long long val); u32string to_u32string(unsigned long long val); u32string to_u32string(long double val); } 27.2 namespace std { ... typedef basic_ios<char> ios; typedef basic_ios<wchar_t> wios; typedef basic_ios<char16_t> u16ios; typedef basic_ios<char32_t> u32ios; ... typedef basic_ifstream<wchar_t> wifstream; typedef basic_ofstream<wchar_t> wofstream; typedef basic_fstream<wchar_t> wfstream; typedef basic_streambuf<char16_t> u16streambuf; typedef basic_istream<char16_t> u16istream; typedef basic_ostream<char16_t> u16ostream; typedef basic_iostream<char16_t> u16iostream; typedef basic_stringbuf<char16_t> u16stringbuf; typedef basic_istringstream<char16_t> u16istringstream; typedef basic_ostringstream<char16_t> u16ostringstream; typedef basic_stringstream<char16_t> u16stringstream; typedef basic_filebuf<char16_t> u16filebuf; typedef basic_ifstream<char16_t> u16ifstream; typedef basic_ofstream<char16_t> u16ofstream; typedef basic_fstream<char16_t> u16fstream; typedef basic_streambuf<char32_t> u32streambuf; typedef basic_istream<char32_t> u32istream; typedef basic_ostream<char32_t> u32ostream; typedef basic_iostream<char32_t> u32iostream; typedef basic_stringbuf<char32_t> u32stringbuf; typedef basic_istringstream<char32_t> u32istringstream; typedef basic_ostringstream<char32_t> u32ostringstream; typedef basic_stringstream<char32_t> u32stringstream; typedef basic_filebuf<char32_t> u32filebuf; typedef basic_ifstream<char32_t> u32ifstream; typedef basic_ofstream<char32_t> u32ofstream; typedef basic_fstream<char32_t> u32fstream; ... template <class state> class fpos; typedef fpos<char_traits<char>::state_type> streampos; typedef fpos<char_traits<wchar_t>::state_type> wstreampos; typedef fpos<char_traits<char16_t>::state_type> u16streampos; typedef fpos<char_traits<char32_t>::state_type> u32streampos; } 27.6 namespace std { template <class charT, class traits = char_traits<charT> > class basic_istream; typedef basic_istream<char> istream; typedef basic_istream<wchar_t> wistream; typedef basic_istream<char16_t> u16istream; typedef basic_istream<char32_t> u32istream; template <class charT, class traits = char_traits<charT> > class basic_iostream; typedef basic_iostream<char> iostream; typedef basic_iostream<wchar_t> wiostream; typedef basic_iostream<char16_t> u16iostream; typedef basic_iostream<char32_t> u32iostream; } namespace std { template <class charT, class traits = char_traits<charT> > class basic_ostream; typedef basic_ostream<char> ostream; typedef basic_ostream<wchar_t> wostream; typedef basic_ostream<char16_t> u16ostream; typedef basic_ostream<char32_t> u32ostream; } 27.7 paragraph 1 namespace std { template <class charT, class traits = char_traits<charT>, class Allocator = allocator<charT> > class basic_stringbuf; typedef basic_stringbuf<char> stringbuf; typedef basic_stringbuf<wchar_t> wstringbuf; typedef basic_stringbuf<char16_t> u16stringbuf; typedef basic_stringbuf<char32_t> u32stringbuf; template <class charT, class traits = char_traits<charT>, class Allocator = allocator<charT> > class basic_istringstream; typedef basic_istringstream<char> istringstream; typedef basic_istringstream<wchar_t> wistringstream; typedef basic_istringstream<char16_t> u16istringstream; typedef basic_istringstream<char32_t> u32istringstream; template <class charT, class traits = char_traits<charT>, class Allocator = allocator<charT> > class basic_ostringstream; typedef basic_ostringstream<char> ostringstream; typedef basic_ostringstream<wchar_t> wostringstream; typedef basic_ostringstream<char16_t> u16ostringstream; typedef basic_ostringstream<char32_t> u32ostringstream; template <class charT, class traits = char_traits<charT>, class Allocator = allocator<charT> > class basic_stringstream; typedef basic_stringstream<char> stringstream; typedef basic_stringstream<wchar_t> wstringstream; typedef basic_stringstream<char16_t> u16stringstream; typedef basic_stringstream<char32_t> u32stringstream; } 27.8.1 paragraph 1 namespace std { template <class charT, class traits = char_traits<charT> > class basic_filebuf; typedef basic_filebuf<char> filebuf; typedef basic_filebuf<wchar_t> wfilebuf; typedef basic_filebuf<char16_t> u16filebuf; typedef basic_filebuf<char32_t> u32filebuf; template <class charT, class traits = char_traits<charT> > class basic_ifstream; typedef basic_ifstream<char> ifstream; typedef basic_ifstream<wchar_t> wifstream; typedef basic_ifstream<char16_t> u16ifstream; typedef basic_ifstream<char32_t> u32ifstream; template <class charT, class traits = char_traits<charT> > class basic_ofstream; typedef basic_ofstream<char> ofstream; typedef basic_ofstream<wchar_t> wofstream; typedef basic_ofstream<char16_t> u16ofstream; typedef basic_ofstream<char32_t> u32ofstream; template <class charT, class traits = char_traits<charT> > class basic_fstream; typedef basic_fstream<char> fstream; typedef basic_fstream<wchar_t> wfstream; typedef basic_fstream<char16_t> u16fstream; typedef basic_fstream<char32_t> u32fstream; } 28.4 namespace std { ... typedef basic_regex<char> regex; typedef basic_regex<wchar_t> wregex; typedef basic_regex<char16_t> u16regex; typedef basic_regex<char32_t> u32regex; ... typedef sub_match<const char*> csub_match; typedef sub_match<const wchar_t*> wcsub_match; typedef sub_match<const char16_t*> u16csub_match; typedef sub_match<const char32_t*> u16csub_match; typedef sub_match<string::const_iterator> ssub_match; typedef sub_match<wstring::const_iterator> wssub_match; typedef sub_match<u16string::const_iterator> u16ssub_match; typedef sub_match<u32string::const_iterator> u32ssub_match; ... typedef match_results<const char*> cmatch; typedef match_results<const wchar_t*> wcmatch; typedef match_results<const char16_t*> u16cmatch; typedef match_results<const char32_t*> u32cmatch; typedef match_results<string::const_iterator> smatch; typedef match_results<wstring::const_iterator> wsmatch; typedef match_results<u16string::const_iterator> u16smatch; typedef match_results<u32string::const_iterator> u32smatch; ... typedef regex_iterator<const char*> cregex_iterator; typedef regex_iterator<const wchar_t*> wcregex_iterator; typedef regex_iterator<const cha16r_t*> u16cregex_iterator; typedef regex_iterator<const char32_t*> u32cregex_iterator; typedef regex_iterator<string::const_iterator> sregex_iterator; typedef regex_iterator<wstring::const_iterator> wsregex_iterator; typedef regex_iterator<u16string::const_iterator> u16sregex_iterator; typedef regex_iterator<u32string::const_iterator> u32sregex_iterator; ... typedef regex_token_iterator<const char*> cregex_token_iterator; typedef regex_token_iterator<const wchar_t*> wcregex_token_iterator; typedef regex_token_iterator<const char16_t*> u16cregex_token_iterator; typedef regex_token_iterator<const char32_t*> u32cregex_token_iterator; typedef regex_token_iterator<string::const_iterator> sregex_token_iterator; typedef regex_token_iterator<wstring::const_iterator> wsregex_token_iterator; typedef regex_token_iterator<u16string::const_iterator> u16sregex_token_iterator; typedef regex_token_iterator<u32string::const_iterator> u32sregex_token_iterator; } |
LWG |
REJECTED
Previously considered; we decided not to do it. We believe it is not too onerous to use wbuffer_convert and wstring_convert which were added as intermediaries to avoid proliferation of types. |
|
UK 144 [72] |
17.1 ¶ 2 | List of contents of library should be extened to cover new clauses | Add "regular expressions, atomic operations and threads" | editor | ACCEPTED | |
UK 145 [73] |
17.1 ¶ 6 | Summary of numeric facilities should mention random numbers | Add random number framework to the list of library facilities | editor | ACCEPTED | |
UK 146 [74] |
17.1 | Add a summary paragraph for regular expressions | Add a summary paragraph for regular expressions | editor | ACCEPTED | |
UK 147 [75] |
17.1 | Add a summary paragraph for threads | Add a summary paragraph for threads | editor | ACCEPTED | |
UK 148 [247] |
17.2 ¶ Table 12 | Table 12 is mentioned in and relates to section 17.2, but has been pushed down to appear directly after the title of section 17.3 which is rather unfortunate/confusing for the reader. | Make sure tables are rendered in the section to which they relate. | editor | ||
UK 149 [84] |
17.3 | For consistency with narrow-oriented and wide-oriented streams, we should add terms for streams of Unicode character sequences | Define Utf16-oriented stream classes and Uft32-oriented stream classes for streams of char16_t/char32_t values. | editor |
ACCEPTED with MODIFICATIONS
The terms "narrow-oriented iostream classes" and "wide-oriented iostream classes" are never used in the standard (except in a somewhat garbled passage that I have rewritten without them), so rather than proliferate definitions of unused terms, I've removed the original culprits. |
|
UK 150 [199] |
17.3 | The addition of move semantics to the language means that many library APIs leave an object in a safely-destructible state, where no other operations can safely be performed unless it is assigned a new value. Library presentation would be simplified and made more precise if we introduce a term for this state. By analogy with singular iterators suggest the term 'singular object' or 'the object is in a singular state'. | Define 'singular state' such that an object with a singular state can only be assigned to or safely destroyed. Assiging a new value typically removes the singular state. Note that objects with a singular state may not be safely copied, so you cannot put an object into a singular state by copying another object in a singular state. Use this new term in the postcondition of all library APIs that move from an rvalue reference. It might also be used to simplify the definition of singular iterator to an iterator value with a singular state. | editor | ||
UK 151 [77] |
17.3.1 | Missing crossreference to 17.3.17 [defns.repositional.stream] | Add cross-reference in the existing empty brackets | editor |
ACCEPTED with MODIFICATIONS
Removed the reference. |
|
UK 152 [80] |
17.3.12 | Object state is using a definition of object (instance of a class) from outside the standard, rather than the 'region of storage' definiton in 1.8p1 | Clarify terms and usage | LWG | 1064 | |
UK 153 [82] |
17.3.17 | If a repositional stream can only seek to a position previously encountered, then an arbitrary-positional-stream cannot satisfy this definition, as cross-referenced in 17.3.17 | Strike the word 'only'. A note might be added to reinforce the intent | editor | ACCEPTED | |
UK 154 [83] |
17.3.20 | Missing definition of a stable partition algorithm | Add definition from 25.2.12p7 | editor |
REJECTED
Since the term "stable partition algorithm" is never used, there is no need to define it. The requirements for the algorithm stable_partition are clear as written. |
|
UK 155 [78] |
17.3.3 | Add clause 28 to list that use this definition of character | Add clause 28 to list that use this definition of character | editor | ACCEPTED | |
UK 156 [79] |
17.3.4 | Add regular expressions to set of templates using character container type | Add regular expressions to set of templates using character container type | editor | ACCEPTED | |
UK 157 [86] |
17.5.2.2 ¶ 3 | Add concepts to the ordered list of presentation | Add concepts into the sequence | editor | ACCEPTED | |
UK 158 [87] |
17.5.2.2 ¶ 3 | templates are neither classes nor functions | Replace 'classes' and 'functions' with 'classes and class templates' and 'functions and function templates' | editor | ACCEPTED | |
UK 159 [88] |
17.5.2.4 ¶ Footnote 152 | This informative footnote was relevant in 1998, not 2008. The term 'existing vendors' may imply something different now | Strike the footnote, or replace 'existing' with 'original' or similar | editor | ACCEPTED | |
UK 160 [89] |
17.5.2.4 ¶ 3 | requires is now a keyword with a specific meaning related to concepts, and its use in library specifcation may be confusing. Generally the Requires clause is used to make requirements on the caller, not the library, so typically providing runtime pre-conditions. Suggest a new name to refflect that. Note that Clause 30 already seems to be written to this convention. | Replace 'Requires' with 'Preconditions' | editor | ||
UK 161 [90] |
17.5.2.4 ¶ 4 | This paragraph is redundant as the definition of the term 'handler function' is already provided in 17.3. Are we in danger of proving two definitions of the same terms? Which is the 'controlling' definition? | Strike 17.5.2.4p4 | editor |
REJECTED
17.3 defines "handler function"; 17.5.2.4/4 imposes requirements on handler functions and replacement functions. There is no redundancy. |
|
UK 162 [170] |
17.5.2.4 ¶ 3 | Clause 30 makes use of a 'Synchronization' semantic element, that frequently appears either between Effects: and Postconditions:, or between Returns: and Throws: | Add 'Synchronization' to the list either between Effects: and Postconditions:, or between Returns: and Throws:. | editor | ACCEPTED | |
UK 163 [219] |
17.5.2.4 ¶ 3 | Many functions are defined as "Effects: Equivalent to a...", which seems to also define the preconditions, effects, etc. But this is not made clear. | Introduce an explicit "Equivalent to", which defines all of the properties of the function. | LWG | 997 | |
UK 164 [91] |
17.5.3.2.1 ¶ 1 | This phrasing predates concepts. While this kind of description is still used, the examples provided are now all concepts, and should be replaced with appropriate examples | Use better names for the examples. Ideally totally replace the need by constraining all templates in library, so that real concepts are all that is needed. Note if retained that CopyConstructible is mis-spelled. | editor | ||
UK 165 [92] |
17.5.3.2.2, 17.5.3.2.3 | constraints on bitmask and enumation types were supposed to be tightened up as part of the motivation for the constexpr feature - see paper n2235 for deails | Adopt wording in line with the motivation described in N2235 | LWG | ||
UK 166 [93] |
17.5.3.2.4.1, 17.5.3.3 | List of library clauses should go up to 30, not 27 | Replace initial refernce to ch27 with ch30 | editor | ACCEPTED | |
UK 167 [246] |
17.5.3.4 Private members | Comment marker in wrong place. | Change: // streambuf* sb; exposition only to streambuf* sb; // exposition only To reflect actual usage. | editor | ACCEPTED | |
UK 168 [406] |
17.6.2.2 ¶ 2 | We should make it clear (either by note or normatively) that namespace std may contain inline namespaces, and that entities specified to be defined in std may in fact be defined in one of these inline namespaces. (If we're going to use them for versioning, eg when TR2 comes along, we're going to need that.) | Replace "namespace std or namespaces nested within namespace std" with "namespace std or namespaces nested within namespace std or inline namespaces nested directly or indirectly within namespace std" | LWG | 1065 | |
UK 169 [95] |
17.6.2.2 | This phrasing contradicts later freedom to implement the C standard library portions in the global namespace as well as std. (17.6.2.3p4) | Resolve conflict in either place | LWG | 992 | |
UK 170 [96] |
17.6.2.3 | One of goals of C++0x is to make language easier to teach and for 'incidental' programmers. The fine-grained headers of the C++ library are valuable in large scale systems for managing dependencies and optimising build times, but overcomplicated for simple development and tutorials. Add additional headers to support the whole library through a single include statement. | Add a new header <std> that has the effect of including everything in tables 13 and 14, except <iosfwd> and <cassert>. Add an additional header <fwd> that adds all declarations from <std> but no definitions. | LWG | 1002 | |
UK 171 [98] |
17.6.2.4 ¶ 3 | Does freestanding implementation require a full implementation of all listed headers? The reference to abort, at_exit and exit is confusing. Is a conforming implementation allow to deliver partial forms of these headers? If so which ones? Are empty versions of everything but <cstdlib> conforming? | Either strike the references to abort, at_exit and exit, or clarify which headers only require partial support. | editor | ACCEPTED | |
UK 172 [99] |
17.6.2.4 ¶ 3 | No reference to new functions quick_exit and at_quick_exit | Add reference to quick_exit and at_quick_exit | LWG |
REJECTED
NAD. We do not belive at_exit and at_quick_exit should be required by freestanding implementations. |
|
UK 173 [450] |
17.6.2.4 ¶ table 15 | <initializer_list> is missing from headers required in freestanding implementations. | Add 18.8, initializer lists, <initializer_list>, to the end of the table. | LWG | ||
JP 23 | 17.6.2.4 ¶ 2nd para, Table 15 | There is a freestanding implementation including <type_traits>, <array>, <ratio>, lately added to Table 13, C++ library headers. Programmers think them useful and hope that these headers are also added to Table 15, C++ headers for freestanding implementations, that shows the set of headers which a freestanding implementation shall include at least. | Add <type_traits>, <array>, <ration> to Table 15. | LWG | 1003 | |
UK 174 [100] |
17.6.3.2 ¶ 3 | The phrasing is mildly ambiguous when using the word 'it' to refer back to the header - an unfotunate reading might confuse it with the translate unit, which is the subject of the surrounding clause. | Replace 'the first reference to any of the entities declared in that header by the translation unit' with 'the first reference to any of the entities that header declares in the translation unit' | editor | ACCEPTED | |
UK 175 [101] |
17.6.4.2.1 ¶ 2 | Local types can now be used to instantiate templates, but don't have external linkage | Remove the reference to external linkage | LWG | ACCEPTED | |
UK 176 [102] |
17.6.4.3.3 ¶ Footnote 175 | Reference to namespace ::std should be 17.6.4.2 | Change referfence from 17.6.4.3 to 17.6.4.2 | editor |
ACCEPTED with MODIFICATIONS
Removed footnote. |
|
UK 177 [103] |
17.6.4.3.4 ¶ 3 | Sentence is redundant as double underscores are reserved in all contexts by 17.6.4.3.3 | Strike the sentence | editor | ACCEPTED | |
UK 178 [104] |
17.6.4.8 ¶ 2 | The last sentence of the third bullet "Operations on such types can report a failure by throwing an exception unless otherwise specified" is redundant as behaviour is already undefined. | Strike the sentence | editor |
REJECTED
The sentence is not redundant. It points out that the behavior is sometimes circumscribed by a prohibition on throwing exceptions. |
|
UK 179 [105] |
17.6.4.8 ¶ 2 | According to the 4th bullet there is a problem if "if any replacement function or handler function or destructor operation throws an exception". There should be no problem throwing exceptions so long as they are caught within the function. | Replace the word 'throws' with 'propogates' | LWG | 1004 | ACCEPTED |
JP 22 | 17.6.5.7 ¶ 4th para, 1st line |
The
statement below describes relation among two or more
threads using words “between threads”:
[Note: This means, for example, that implementations
can’t use a static object for internal purposes
without synchronization because it could cause a data race
even in programs that do not explicitly share objects
between threads. —end note ]
In such case, “among” is preferred instead of “between”. |
Change "between threads" to "among threads".
There are same cases in 17.6.1 paragraph 2, 17.6.5.7 paragraph 6, 30.1 paragraph 1, 30.3.1 paragraph 1 also. |
editor |
REJECTED
"Between" expresses one-to-one relations of pairs in a group; "among" refers to collective relations in a group. "Between" is correct here. |
|
UK 180 [107] |
17.6.5.10 ¶ 1, 4 | It should not be possible to strengthen the exception specification for virtual functions as this could break user code. Note this is not a problem in practice as there are no virtual functions with exception specifications in the current library, other than empty throw specifications which it is not possible to strengthen. | Add restriction that exception specification of virtual functions cannot be tightened. | LWG |
REJECTED
NAD, the standard already has the desired restriction. |
|
UK 181 [108] |
17.6.5.10 ¶ Footnote 186 | This footnote is wrong. C library functions do not have any exception specification, but might be treated as if they had an empty throw specification | Clarify that this note does not mean the functions are genuinely declared with the specification, but are treated as-if. | editor | ACCEPTED | |
UK 182 [109] |
17.6.5.10 ¶ Footnote 188 | It is very helpful to assume all exceptions thrown by the standard library derive from std::exception. The 'encouragement' of this note should be made normative. | Make this footnote normative | LWG |
REJECTED
NAD. We cannot mandate all standard-library functions that might use some third-party library. |
|
UK 184 [144] |
18 -> 30 | The new alias-declaration syntax is generally easier to read than a typedef declaration. This is especially true for complex types like function pointers. | Replace all typedef declarations in the standard library with alias-declarations, except in the standard C library. | editor | ||
JP 24 | 18 ¶ 2nd para, Table 16 |
Subclauses are listed in Table 16 as:
"18.6 Type identification …" "18.8 Initializer lists …" "18.7 Exception handling …". |
Sort them in the increasing order
"18.6 Type identification …"
"18.7 Exception handling …". "18.8 Initializer lists …" |
editor | ACCEPTED | |
JP 25 | 18.1 ¶ 6th para , last line, SEE ALSO | max_align_t is described in 18.1, so add 3.11 Alignment as the reference. | Add "3.11, Alignment" to SEE ALSO. | editor | ACCEPTED | |
FR 32 | 18.2.1 [Numeric limits] | The definition of numeric_limits<> as requiring a regular type is both conceptually wrong and operationally illogical. As we pointed before, this mistake needs to be corrected. For example, the template can be left unconstrained. In fact this reflects a much more general problem with concept_maps/axioms and their interpretations. It appears that the current text heavily leans toward experimental academic type theory. | We suggest that a more pragmatic approach, in the spirit of C and C++, be taken so that calls to constrained function templates are interpreted as assertions on *values*, not necessarily semantics assertions on the carrier type. | LWG | 902 | |
DE 16 | 18.2.1 | DE-16 The class template numeric_limits should not specify the Regular concept requirement for its template parameter, because it contains functions returning NaN values for floating-point types; these values violate the semantics of EqualityComparable. See also library issue 902 in WG21 document N2794 "C++ Standard Library Active Issues List (Revision R60)", available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2794.html . | Specify a concept requirement with fewer constraints as appropriate, for example SemiRegular. | LWG | 902 | |
JP 26 | 18.2.1.1 | numeric_limits does not use concept. |
Correct as follows.
template<class T> class numeric_limits<const T>; template<class T> class numeric_limits<volatile T>; template<class T> class numeric_limits<const volatile T>; should be template<Regular T> class numeric_limits<const T>; template<Regular T> class numeric_limits<volatile T>; template<Regular T> class numeric_limits<const volatile T>; |
LWG | 1005 | |
DE 17 | 18.2.6 | DE-17 The class type_index should be removed; it provides no additional functionality beyond providing appropriate concept maps. | Specify concept maps for "const type_info *" as required by the ordered and unordered containers and remove the class type_index. | LWG | 1078 | |
UK 185 [264] |
18.3.1 ¶ 2 | There is no header <stdint>, it should either be <stdint.h> or <cstdint> | Replace <stdint> with <cstdint> | editor | ACCEPTED | |
DE 18 | 18.4 | DE-18 The proposed C++ standard makes a considerable number of existing programs that have well-defined behavior according to ISO/IEC 14882:2003 have undefined behavior without a diagnostic hint to the programmer at all. This applies to the presence of threads and to pointer safety (the latter was introduced to support garbage collection). In order to avoid requiring a full code review for user code, facilities should be present that allow the compile-time detection of the advanced features of the proposed C++ standard. It is expected that C++ implementations will provide a means (for example, a command-line switch) to turn off either or both of threads and garbage collection support, turning potentially undefined programs into well-defined ones. Note: This issue is contributing significantly to Germany's overall “no” vote. | Consider applying the changes proposed in WG21 document N2693 "Requirements on programs and backwards compatibility", available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2008/n2693.html . | LWG | ||
UK 186 [265] |
18.4 ¶ Footnote 221 | What is the purpose of this comment? The standard stream objects (cin, cerr etc.) have a peculiar lifetime that extends beyond the program. They may never be destroyed so will not be responsible for flushing buffers at the stated time. | Remove the footnote | editor | ACCEPTED | |
UK 187 [266] |
18.4 ¶ 9 | The term "thread safe" is not defined nor used in this context anywhere else in the standard. | Clarify the intended meaing of "thread safe" | LWG | ||
UK 188 [267] |
18.4 ¶ 12 | The function _Exit does not appear to be defined in this standard. Should it be added to the table of functions included-by-reference to the C standard? | Depends on where _Exit comes from | LWG | 993 | ACCEPTED |
UK 189 [273] |
18.4, 18.7 | The addition of the [[noreturn]] attribute to the language will be an important aid for static analysis tools. | The following functions should be declared in C++ with the [[noreturn]] attribute: abort exit quick_exit terminate unexpected rethrow_exception throw_with_nested | LWG | 1066 | ACCEPTED |
JP 27 | 18.4, 18.9, 18.7.2.2, 18.7.3.1 | There are Standard library functions that never return to the caller. They are explained so in the Standard but not declared explicitly. |
Consider to add the attribute [[noreturn]] to such
functions,
15.5.2 unexpected 18.4: abort(), exit(), quick_exit, 18.7.2.2: unexpected_handler, 18.7.3.1: terminate_handler, 18.7.6 rethrow_nested 18.7.6 throw_with_nested 18.9: longjmp. |
LWG | ||
UK 190 [268] |
18.5.1 ¶ various | It is not entirely clear how the current specification acts in the presence of a garbage collected implementation. | All deallocation functions taking a pointer parameter should have a Precondition : ptr is a safely-derived pointer value. | LWG | 1006 | ACCEPTED |
UK 191 [271] |
18.5.1.1 ¶ 4 | According to the second bullet, behaviour becomes undefined (for lack of a specification) if the user has not yet called set_new_handler. | Rephrase the second bullet in terms of a new handler being installed, and update any definition of handler function necessary to be sure the term 'installed' is defined. | editor |
ACCEPTED with MODIFICATIONS
Reworded, but not with "installed". |
|
UK 192 [269] |
18.5.1.2 | The declared signature is not compatible with the current requirement to throw std::length_error. It is too late to weaken the exception specification, so the only other change is to preserve new (improved) behaviour is to throw std::bad_alloc, or something derived from std::bad_alloc. | Fix 5.3.4p7 by required std::bad_alloc rather than std::length_error | CWG | 805 | |
UK 193 [272] |
18.5.2.2 ¶ 2 | quick_exit has been added as a new valid way to terminate a program in a well defined way | Change 3rd bullet: call either abort(), exit() or quick_exit(); | LWG | 994 | |
UK 194 [443] |
18.6 | The inclusion of type_index and hash<type_index> in <typeinfo> brings dependencies into <typeinfo> which are inconsistent with the definition of freestanding C++ in 17.6.2.4. | Move type_index and hash<type_index> out of <typeinfo> and into a new header, <typeindex>. | LWG | ||
JP 28 | 18.6, 18.7, 19.1 | Errors reported by Exception classes are of types char or std::string only. For example, std::exception is declared with char, std::string types, therefore types wchar_t/wstring, char16_t/u16string, or char32_t/u32string can not be used. | Consider other types. | LWG |
REJECTED
NAD. There is already guidance in the CD. It is the caller's responsibility to internationalize MB character string. |
|
JP 29 | 18.7.6 | throw_with_nested does not use concept. |
Correct as follows.
template<class T> void throw_with_nested(T&& t); // [[noreturn]] should be template<CopyConstructible T> void throw_with_nested(T&& t); // [[noreturn]] |
LWG | 1007 | |
JP 30 | 18.7.6 | To handle nested exceptions strictly, error information of tree structure will be required though, the nested_exception does not support tree structure. It is insufficient as error handling. | Consider nested_exception to support tree structure. | LWG | ||
JP 31 | 18.7.6 | It is difficult to understand in which case nested_exception is applied. | Consider to add a sample program which rethrows exception. | LWG | 1008 | |
UK 195 [451] |
18.8 | The class definition of std::initializer_list contains concept-maps to Range which should be out of the class, and in <iterator_concepts> instead. Otherwise, it's not possible to use initializer_lists in a freestanding C++ implementation. | Delete the two concept maps from std::initializer_list. | LWG | ||
UK 196 [452] |
18.8.3 | Concept maps for initializer_list to Range should not be in language support headers, but instead in iterator concepts. | Remove section 18.8.3 and put it in 24.1.8.1 instead, so that the concept_maps from initializer_list to Range are specified under Range instead of under initializer lists; also, so that they're implemented in <iterator_concepts> instead of <initializer_list>. | LWG | ||
UK 197 [275] |
19 | All the exception classes in this clause take a std::string argument by const reference. They should all be overloaded to accept std::string by rvalue rerefence for an efficient move as well. | Provide a constructor for every exception class in clause 19 accepting a std::string by rvalue reference, with the semantics that the passed string may be moved. | LWG |
REJECTED
NAD. Implementations are permitted to add the requested signature under the as-if rule. See clause 17. |
|
JP 32 | 19.1 |
Messages returned by the member function what() of standard
exception classes seem difficult to judge.
For example, following messages are returned by what() of std::bad_alloc of existing implementations: Compiler: Message returned by what() --------------------------------------------- Borland C++ 5.6.4: no named exception thrown Visual C++ 8.0: bad allocation Code Warrior 8.0: exception g++ 3.4.4: St9exception It is difficult to recognize what exception was thrown when using those compilers except Visual C++. |
Consider to add footnote that recommends what() returns message easy to recognize what exception was thrown. | LWG |
REJECTED
NAD. This is a quality of implementation issue that is beyond the scope of the standard. |
|
US 64 | 19.3 ¶ 1 | “ See also: ISO C 7.1.4, 7.2, Amendment 1 4.3.” It is unclear why this cross reference is here. Amendment 1 was to C89, not C99. | Delete this cross reference. If necessary, expand the main text to include the relevant referenced text | editor | ACCEPTED | |
US 65 | 20 | Scoped allocators and allocator propagation traits add a small amount of utility at the cost of a great deal of machinery. The machinery is user visible, and it extends to library components that don't have any obvious connection to allocators, including basic concepts and simple components like pair and tuple. |
Sketch of proposed resolution: Eliminate scoped allocators,
replace allocator propagation traits with a simple uniform
rule (e.g. always propagate on copy and move), remove all
mention of allocators from components that don't explicitly
allocate memory (e.g. pair), and adjust container
interfaces to reflect this simplification.
Components that I propose eliminating include HasAllocatorType ?, is_scoped_allocator, allocator_propagation_map, scoped_allocator_adaptor, and ConstructibleAsElement ?. |
LWG | 1075 | |
UK 198 [126] |
20 | The organization of clause 20 could be improved to better group related items, making the standard easier to navigate. | 20.6.7, 20.6.8, 20.6.9 and 20.6.10 should be grouped under a section called "operator wrappers" or similar. The goal of all 4 subsections combined is to provide a functor for every operator in the language. 20.6.17 class template hash should numerically appear immediately after the operator wrappers, as they are functors that are used in similar ways 20.6.11, 20.6.12, 20.6.13, 20.6.14, 20.6.15 are strongly related to 20.6.3, and to an extent 20.6.2. These should all come under a subheading of "function adapters" 20.7.1, 20.7.3, 20.7.4, 20.7.5, 20.7.6, 20.7.7 and 20.7.10 should all be grouped as subclauses under [20.7.2 Allocators] [20.7.12 unique_ptr] should be a sub-section of [20.7.13 smart pointers] [20.7.13 smart pointers] is important enough to have a high level bullet after [20.7 memory], suggest renumbering as [20.8 smart pointers] [20.7.13.7 Pointer safety] is nothing to do with smart pointers and should become its own subclause [20.7.14 Pointer safety] [20.9 date and time functions] should be moved under [20.8 time utilities] and retitled [20.8.6 C Library] (as per memory management/C Library) [20.6.18 reference_closure] is fundamentally a language support feature and should move to clause 18, see separate comment [20.6.5 reference_wrapper] should be simplified and moved into [2.2 utility components], see separate comment [20.6.4 result_of] should be reorganised as a type trait - see separate comment Tuples and pairs are closely related so merge tuple and pair into the same subclause - see more general comment on this | editor | ||
UK 199 [127] |
20.1.1, 20.1.2 ¶ 2 | The requirement that programs do not supply concept_maps should probably be users do not supply their own concept_map specializations. THe program will almost certainly supply concept_maps - the standard itself supplies a specialization for RvalueOf? references. Note that the term _program_ is defined in 3.5p1 and makes no account of the standard library being treated differently to user written code. | Replace the term 'program' with 'user'. | LWG | 1015 | ACCEPTED |
UK 200 [354] |
20.1.4 | All standard library use expects Predicates to be CopyConstructible, and this should be recognised easily without reatedly stating on every use-case. | Either require CopyConstructible<F> as part of Predicate, or create a refined concept, StdPredicate, used throughout the library that requires CopyConstructible as well as Callable. Consider making (Std)Predicate SemiRegular. | LWG |
REJECTED
After further discussion of UK200, we do not think adding constraints to predicates is a good idea. Straw poll on UK200: status quo, 5 pro - 2 con; remove copy-constructible, 3 pro - 4 con; pred must be copy-constructible, 4 pro - 3 con; no consensus for moving away from the status quo. Also see UK245. |
|
UK 201 [290] |
20.1.5 | The Consistency axiom for LessThanComparable will not compile. | Add a requires clause to the Consistency axiomL requires HasLessEquals<T> && HasGreaterEquals<T>, or split the Consistency axiom into two so that 'basic' consistency can be asserted regardless of the <=/>= requirement. | LWG |
REJECTED
After consultation with the submitter, we agreed that the suggestion was in error and there is nothing else to be done. |
|
JP 33 | 20.1.5 | LessThanComparable and EqualityComparable don't correspond to NaN. | Apply concept_map to these concepts at FloatingPointType | LWG | 1016 | |
US 66 | 20.1.10 | Application of the "Regular" concept to floating-point types appears to be controversial (see long discussion on std-lib reflector). | State that the “Regular” concept does not apply to floating-point types. | LWG | 1017 | |
JP 34 | 20.2 ¶ 1st para, 4th line | Though N2672 pointed at adding "#include<initializer_list>", it isn't reflected. |
add
followings
#include <initializer_list> // for concept_map |
editor | ACCEPTED | |
US 67 | 20.2.1 ¶ 5 first sent. | Some connective words are missing. | Insert “corresponding to” before “an lvalue reference type.” | editor | ACCEPTED with MODIFICATIONS | |
JP 35 | 20.2.3 ¶ 6th para, 1st line |
Typo,
"stdforward" should be "std::forward" |
Correct typo. | editor | ACCEPTED | |
UK 202 [213] |
20.2.4 | The references to pair in the tuple-like access to pair functions qualify pair with std::pair even though they are in a namespace std block. | Remove the std:: qualification from these references to pair. | editor | ACCEPTED | |
US 68 | 20.1.12 ¶ IntegralLike | The code defining the context is syntactically incorrect. | Insert a comma in two places: at the end of the third line of refinements, and at the end of the fourth line of refinements. | editor |
ACCEPTED
Section reference corrected from 20.2.12. |
|
UK 203 [229] |
20.3.2 ¶ 1-4 | The ratio_xyz types have a misplaced '}'. For example: template <class R1, class R2> struct ratio_add { typedef see below} type; ; | Move the '}' to after the typedef: template <class R1, class R2> struct ratio_add { typedef see below type; }; | editor | ACCEPTED | |
JP 36 | 20.4.2.1 ¶ 19th para, 6th line |
Typo.
"it it" should be "it is" |
Correct typo. | editor | ACCEPTED | |
UK 204 [239] |
20.5.7 [meta.trans.other] ¶ Table 41 | It is not possible to create a variant union based on a parameter pack expansion, e.g. to implement a classic discriminated union template. | Restore aligned_union template that was removed by LWG issue 856. | LWG | 1020 |
ACCEPTED
Section reference corrected from 20.5. |
US 69 | 20.5 | This section, dealing with tuple<>, should be in the same section as the similar utility pair<>. | Restructure Clause 20 so as to bring these similar components together. | editor | ||
UK 205 [253] |
20.5.3 | integral_constant objects should be usable in integral-constant-expressions. The addition to the language of literal types and the enhanced rules for constant expressions make this possible. | Add a constexpr conversion operator to class tempalate integral_constant: constexpr operator value_type() { return value; } | LWG | 1019 | ACCEPTED |
UK 206 [255] |
20.5.5 ¶ para 4 | Currently the std says: "In order to instantiate the template is_convertible<From, To>, the following code shall be well formed:" But the code shown is the requirement for the result of is_convertible to be a true_type, not a precondition on whether the template can be instantiated. | Change: "In order to instantiate the template is_convertible<From, To>, the following code shall be well formed:" To: "The template is_convertible<From, To> inherits either directly or indirectly from true_type if the following code is well formed:" | LWG | ACCEPTED | |
UK 207 [256] |
20.5.6.1 ¶ Table 36 | suffix "::type" is missing from the some of the examples. | Change: Example:remove_const<const volatile int>::type evaluates to volatile int, whereas remove_const<const int*> is const int*. —end example To: Example:remove_const<const volatile int>::type evaluates to volatile int, whereas remove_const<const int*>::type is const int*. —end example And change: Example:remove_volatile<const volatile int>::type evaluates to const int, whereas remove_volatile<volatile int*> is volatile int*. —end example To: Example:remove_volatile<const volatile int>::type evaluates to const int, whereas remove_volatile<volatile int*>::type is volatile int*. —end example And change: Example:remove_cv<const volatile int>::type evaluates to int, whereas remove_cv<const volatile int*> is const volatile int*. —end example To: Example:remove_cv<const volatile int>::type evaluates to int, whereas remove_cv<const volatile int*>::type is const volatile int*. —end example | editor | ACCEPTED | |
JP 37 | 20.5.7 ¶ Table 41 |
Typo.
There isn't a period at the end of enable_if's comments. If B is true, the member typedef type shall equal T; otherwise, there shall be no member typedef type should be If B is true, the member typedef type shall equal T; otherwise, there shall be no member typedef type. |
Add ”.” | editor | ACCEPTED | |
US 70 | 20.5 | Specifications now expressed via narrative text are more accurately and clearly expressed via executable code. | Wherever concepts are available that directly match this section’s type traits, express the traits in terms of the concepts instead of via narrative text. Where the type traits do not quite match the corresponding concepts, bring the two into alignment so as to avoid two nearly-identical notions. | LWG | 1018 | Section reference corrected from 20.6. |
US 71 | 20.6.7 ¶ Table 51, last row, column 3 | The grammar is incorrect. | Change “conversion are” to “conversion is”. | editor |
ACCEPTED with MODIFICATIONS
"conversions are" is correct. |
|
JP 38 | 20.6.12.1.3 |
add the move requirement for bind's return type.
For example, assume following th1 and th2, void f(vector<int> v) { } vector<int> v{ ... }; thread th1([v]{ f(v); }); thread th2(bind(f, v)); When function object are set to thread, v is moved to th1's lambda expression in a Move Constructor of lambda expression becuase th1's lambda expression has a Move Constructor. But bind of th2's return type doesn't have the requirement of Move, so it may not moved but copied. Add the requirement of move to get rid of this useless copy. And also, add the MoveConstructible as well as CopyConstructible. |
Add
the following requirements.
"it has a public move constructor
that performs a member-wise move."
Add the MoveConstructible. |
LWG | 817 | |
JP 39 | 20.6.16.2 | There are no requires corresponding to F of std::function. |
Correct as follows.
template<class F, Allocator A> function(allocator_arg_t, const A&, F); template<class F, Allocator A> function(allocator_arg_t, const A&, F&&); should be template<class F, Allocator A> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> function(allocator_arg_t, const A&, F); template<class F, Allocator A> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> function(allocator_arg_t, const A&, F&&); |
LWG | 1024 | ACCEPTED with MODIFICATIONS |
JP 40 | 20.6.16.2 | Thougn it's "Allocator Aloc" at other places, it's "Allocator A" only std::function constructor template parameter. |
Correct as follows.
template<class F, Allocator A> function(allocator_arg_t, const A&, F); template<class F, Allocator A> function(allocator_arg_t, const A&, F&&); should be template<class F, Allocator Alloc> function(allocator_arg_t, const Alloc &, F); template<class F, Allocator Alloc> function(allocator_arg_t, const Alloc &, F&&); |
editor | ACCEPTED | |
JP 41 | 20.6.16.2 | There are no requires corresponding to R and Args of UsesAllocator. |
Correct as follows.
template <class R, class... Args> concept_map UsesAllocator<function<R(Args...)>, Alloc> { typedef Alloc allocator_type; } should be template <Returnable R, CopyConstructible... Args> concept_map UsesAllocator<function<R(Args...)>, Alloc> { typedef Alloc allocator_type; } |
LWG |
REJECTED
This change would be redundant because function<> is already sufficiently constrained. No actions necessary. |
|
JP 42 | 20.6.16.2 |
The requires
are wrong.
R require Returnable, and ArgTypes requires CopyConstructible by a definition of function, then it's a mistake to designate followings by MoveConstructible. |
Correct as follows.
template <MoveConstructible R, MoveConstructible... ArgTypes> bool operator==(const function<R(ArgTypes...)>&, nullptr_t); template <MoveConstructible R, MoveConstructible... ArgTypes> bool operator==(nullptr_t, const function<R(ArgTypes...)>&); template <MoveConstructible R, MoveConstructible... ArgTypes> bool operator!=(const function<R(ArgTypes...)>&, nullptr_t); template <MoveConstructible R, MoveConstructible... ArgTypes> bool operator!=(nullptr_t, const function<R(ArgTypes...)>&); template <MoveConstructible R, MoveConstructible... ArgTypes> void swap(function<R(ArgTypes...)>&, function<R(ArgTypes...)>&); should be template <Returnable R, CopyConstructible... ArgTypes> bool operator==(const function<R(ArgTypes...)>&, nullptr_t); template <Returnable R, CopyConstructible... ArgTypes> bool operator==(nullptr_t, const function<R(ArgTypes...)>&); template <Returnable R, CopyConstructible... ArgTypes> bool operator!=(const function<R(ArgTypes...)>&, nullptr_t); template <Returnable R, CopyConstructible... ArgTypes> bool operator!=(nullptr_t, const function<R(ArgTypes...)>&); template <Returnable R, CopyConstructible... ArgTypes> void swap(function<R(ArgTypes...)>&, function<R(ArgTypes...)>&); |
editor | ACCEPTED | |
UK 208 [338] |
20.6.17 ¶ 1 | std::hash should be implemented for much more of the standard library. In particular for pair, tuple and all the standard containers. | . | LWG |
REJECTED
NAD. Consensus is that this is an expansion beyond the scope of C++0X. |
|
UK 209 [111] |
20.7 | Smart pointers cannot be used in constrained templates | Provide constraints for smart pointers | LWG | 1026 | |
UK 213 [357] |
20.7.6 | std::allocator should be constrained to simplify its use on constrained contexts. This library component models allocation from free store via the new operator so choose constraints to match. The Allocator concept allows for a wider variety of allocators that users may choose to supply if their allocation model does not require operator new, without impacting the requirements of this template. | The primary allocator template should be constrained to require ObjectType<T> and FreeStoreAllocatable<T>. Further operations to be constrained as required. | LWG | 1027 | ACCEPTED |
UK 214 [125] |
20.7.8 | raw_storage_iterator needs constraining as an iterator adaptor to be safely used in constrained templates | Constrain the raw_storage_iterator template | LWG | 1028 | |
UK 210 [124] |
20.7.11 | Specialized algorithms for memory managenment need requirements to be easily usable in constrained templates | Provide constraints for all algorithms in 20.7.11 | LWG | 1029 | |
DE 20 | 20.6.12 | DE-20 The section heading and the first sentence use the term "template function", which is undefined. | Replace "template function" by "function template". | editor |
ACCEPTED
Section reference corrected from 20.7.12. |
|
US 72 | 20.6.12.1.3 [func.bind.bind] | bind should support move-only functors and bound arguments. | LWG | 817 | Section reference corrected from 20.7.12. | |
DE 21 | 20.6.12.1.3 [func.bind.bind] | DE-21 The specification for bind claims twice that "the values and types for the bound arguments v1, v2, ..., vN are determined as specified below". No such specification appears to exist. | Add the missing specification in the same section, or add a cross-reference indicating the section where the specification appears. | LWG | 817 |
ACCEPTED
Section reference corrected from 20.7.12.1.3. |
UK 211 [428] |
20.6.12.2.3 [unique.ptr.single.asgn] ¶ 11 | The nullptr_t type was introduced to resolve the null pointer literal problem. It should be used for the assignemnt operator, as with the constructor and elsewhere through the library. | Change signature here and in the synopsis to: unique_ptr& operator=(nullptr_t); Strike the sentance an note before the Effects clause. | LWG | 1021 |
ACCEPTED
Section reference corrected from 20.7.12.2.3. |
UK 212 [270] |
20.6.13.7 [util.dynamic.safety] | The pointer-safety API is nothing to do with smart pointers, so does not belong in 20.7.13. In fact it is a set of language support features are really belongs in clause 18, with the contents declared in a header that deals with language-support of memory management. | Move this specification to 18.5. Move the declarations into the header <new>. | LWG |
ACCEPTED with MODIFICATIONS
Section reference corrected from 20.7.13.7. |
|
DE 22 | 20.6.16.2 [func.wrap.func] | DE-22 The conditions for deriving from std::unary_function and std::binary_function are unclear: The condition would also be satisfied if ArgTypes were std::vector<T1>, because it (arguably) "contains" T1. | Consider stating the conditions in normative prose instead of in comments in the class definition. Use "consists of" instead of "contains". Consider using "if and only if" instead of "iff". | LWG | 1023 |
ACCEPTED
Section reference corrected from 20.7.16.2. |
US 73 | 20.6.18 |
The
std::reference_closure template is redundant with
std::function and should be removed.
std::reference_closure is a premature optimization that provides a limited subset of the functionality of std::function intended to improve performance in a narrow use case. However, the “parallel application performance” benchmark used to motivate the inclusion of std::reference_closure was flawed in several ways:
|
Remove 20.7.18
[func.referenceclosure].
Remove 5.1.1 paragraph 12. |
CWG | 750 |
ACCEPTED
Section reference corrected from 20.7.18. See paper N2845 |
US 74.1 | 20.8 |
Scoped
allocators represent a poor trade-off for standardization,
since (1) scoped-allocator--aware containers can be
implemented outside the C++ standard library but used with
its algorithms, (2) scoped allocators only benefit a tiny
proportion of the C++ community (since few C++ programmers
even use today’s allocators), and (3) all C++ users,
especially the vast majority of the C++ community that
won’t ever use scoped allocators are forced to cope
with the interface complexity introduced by scoped
allocators. In essence, the larger community will suffer to
support a very small subset of the community who can
already implement their own data structures outside of the
standard library. Therefore, scoped allocators should be
removed from the working paper.
Some evidence of the complexity introduced by scoped allocators: 20.3.3, 20.5: large increase in the number of pair and tuple constructors 23: confusing “AllocatableElement” requirements throughout. |
Remove support
for scoped allocators from the working paper. This includes
at least the following changes:
Remove 20.8.3 [allocator.element.concepts] Remove 20.8.7 [allocator.adaptor] Remove 20.8.10 [construct.element] In Clause 23: replace requirements naming the AllocatableElement concept with requirements naming CopyConstructible, MoveConstructible, DefaultConstructible, or Constructible, as appropriate. |
LWG | 1075 | See paper N2840 |
US 74.2 | 20.8.2.2 ¶ (a) synopsis (b) after ¶ 14 | A concept name is twice misspelled. | Change “Hasconstructor” to “HasConstructor” (twice). | editor |
ACCEPTED with MODIFICATIONS
The text in question is in 20.7.2.2, not 20.8.2.2. |
|
US 75 | 20.8.2.2 | Allocator concepts are incomplete | See paper: http://www.halpernwightsoftware.com/WG21/n2810_allocator_defects.pdf | LWG |
ACCEPTED with MODIFICATIONS
See paper N2840 |
|
JP 43 | 20.8.3 |
Typo.
"alloc" should be "Alloc" |
Correct as follows.
auto concept UsesAllocator<typename T, typename Alloc> { requires Allocator<alloc>; typename allocator_type = T::allocator_type; should be auto concept UsesAllocator<typename T, typename Alloc> { requires Allocator<Alloc>; typename allocator_type = T::allocator_type; |
editor | ACCEPTED | |
UK 215 [356] |
20.8.3.3 ¶ 6,8 | Extra pair of {}, presumably some formatting code gone awry : duration& operator-{-}(); | Remove the {} or fix formatting | editor | ACCEPTED | |
US 77 | 20.8.4 |
Allocator-specific move and copy behavior for containers
(N2525) complicates a little-used and already-complicated
portion of the standard library (allocators), and breaks
the conceptual model of move-constructor and
move-assignment operations on standard containers being
efficient operations. The extensions for allocator-specific
move and copy behavior should be removed from the working
paper.
With the introduction of rvalue references, we are teaching programmers that moving from a standard container (e.g., a vector<string>) is an efficient, constant-time operation. The introduction of N2525 removed that guarantee; depending on the behavior of four different traits (20.8.4), the complexity of copy and move operations can be constant or linear time. This level of customization greatly increases the complexity of standard containers, and benefits only a tiny fraction of the C++ community. |
Remove 20.8.4.
Remove 20.8.5. Remove all references to the facilities in 20.8.4 and 20.8.5 from clause 23. |
LWG | ||
US 78 | 20.8.12, 20.8.13.2 |
There is presently no way to
convert directly from a shared_ptr to a
unique_ptr .
|
Add
an interface that performs the conversion. See the
attached, Issues with the C++ Standard" paper under Chapter
20, "Conversion from shared_ptr to
unique_ptr".
|
LWG | 1031 | |
US 79 | 20.8.12.2.1 |
[unique.ptr.single.ctor]/5 no longer requires for D not to
be a pointer type. This restriction needs to be put
back in. Otherwise we have a run time failure that
could have been caught at compile time:
unique_ptr<int, void(*)(void*)> p(malloc(sizeof(int))); // should not compile unique_ptr<int, void(*)(void*)> p(malloc(sizeof(int)), free); // ok |
LWG | 932 | ACCEPTED | |
JP 44 | 20.7.13.6 [util.smartptr.shared.atomic] |
The
1st parameter p and 2nd parameter v is now
shared_ptr<T> *.
It should be shared_ptr<T>&, or if these are shared_ptr<T>* then add the "p shall not be a null pointer" at the requires. |
Change shared_ptr<T>& or add the "p shall not be a null pionter" at the requires. | LWG | 1030 |
ACCEPTED
Section reference corrected from 20.8.13.6. |
JP 45 | 20.9 |
Rep, Period, Clock and Duration don't correspond to
concept.
template <class Rep, class Period = ratio<1>> class duration; template <class Clock, class Duration = typename Clock::duration> class time_point; |
Make concept for Rep, Period, Clock and Duration, and fix 20.9 and wait_until and wait_for's template parameter at 30. | LWG | 1032 | |
US 80 | 20.8.2.1 [time.traits.is_fp] ¶ Heading | The section heading does not describe the contents. | Change the heading “is_floating_point” to “treat_as_floating_point”. Optionally adjust the section’s label similarly. | editor |
ACCEPTED with MODIFICATIONS
Section reference corrected from 20.9.2.1. |
|
US 81 | 20.9.3 |
chrono::duration is missing the modulous operator for both
member and non-member arithmetic. This operator is useful
for finding the position of a duration within a bounded
time frame. Having it be built in to duration is safer than
requiring the client to extract and operate directly on the
duration’s representation as the latter will not
enforce the correct units of the operation.
Ex: milliseconds ms(...); microseconds us(...); ms % us; // microseconds us % ms; // microseconds ms % 5; // milliseconds 5 % ms; // does not compile |
Add:
template <class Rep, class Period = ratio<1>> class duration { public: ... duration& operator%(const rep&); duration& operator%(const duration&); .. }; template <class Rep1, class Period, class Rep2> duration<typename common_type< Rep1, Rep2>::type, Period> operator%(const duration<Rep1, Period>& d, const Rep2& s); template <class Rep1, class Period1, class Rep2, class Period2> typename common_type<duration<Rep1, Period1>, duration<Rep2, Period2>>::type operator%(const duration<Rep1, Period1>& lhs, const duration<Rep2, Period2>& rhs); |
LWG | 934 | ACCEPTED |
US 82 | 20.9.5.3 ¶ after ¶ 1 | The code synopsis has a minor alignment error. | Align “rep” with the other symbols defined in this synopsis. | editor | ACCEPTED | |
UK 216 [282] |
21 | All the containers use concepts for their iterator usage, exect for basic_string. This needs fixing. | Use concepts for iterator template parameters throughout the chapter. | LWG | 1081 | |
JP 46 | 21.2, 21.3 | The basic_string does not use concept. |
The
"class Alloc" is changed to "Allocator Alloc".
The "class InputIterator" is changed to "InputIterator Iter". // 21.3, basic_string: template<class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_string; template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc> operator+(const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(basic_string<charT,traits,Alloc>&& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(const basic_string<charT,traits,Alloc>& lhs, basic_string<charT,traits,Alloc>&& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(basic_string<charT,traits,Alloc>&& lhs, basic_string<charT,traits,Alloc>&& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc> operator+(const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(const charT* lhs, basic_string<charT,traits,Alloc>&& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc> operator+(charT lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(charT lhs, basic_string<charT,traits,Alloc>&& rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc> operator+(const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(basic_string<charT,traits,Alloc>&& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc> operator+(const basic_string<charT,traits,Alloc>& lhs, charT rhs); template<class charT, class traits, Allocator Alloc> basic_string<charT,traits,Alloc>&& operator+(basic_string<charT,traits,Alloc>&& lhs, charT rhs); template<class charT, class traits, Allocator Alloc> bool operator==(const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator==(const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator==(const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator!=(const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator!=(const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator!=(const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator< (const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator< (const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator< (const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator> (const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator> (const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator> (const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator<=(const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator<=(const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator<=(const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator>=(const basic_string<charT,traits,Alloc>& lhs, const basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> bool operator>=(const basic_string<charT,traits,Alloc>& lhs, const charT* rhs); template<class charT, class traits, Allocator Alloc> bool operator>=(const charT* lhs, const basic_string<charT,traits,Alloc>& rhs); // 21.3.8.8: swap template<class charT, class traits, Allocator Alloc> void swap(basic_string<charT,traits,Alloc>& lhs, basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> void swap(basic_string<charT,traits,Alloc>&& lhs, basic_string<charT,traits,Alloc>& rhs); template<class charT, class traits, Allocator Alloc> void swap(basic_string<charT,traits,Alloc>& lhs, basic_string<charT,traits,Alloc>&& rhs); // 21.3.8.9: inserters and extractors template<class charT, class traits, Allocator Alloc> basic_istream<charT,traits>& operator>>(basic_istream<charT,traits>&& is, basic_string<charT,traits,Alloc>& str); template<class charT, class traits, Allocator Alloc> basic_ostream<charT, traits>& operator<<(basic_ostream<charT, traits>&& os, const basic_string<charT,traits,Alloc>& str); template<class charT, class traits, Allocator Alloc> basic_istream<charT,traits>& getline(basic_istream<charT,traits>&& is, basic_string<charT,traits,Alloc>& str, charT delim); template<class charT, class traits, Allocator Alloc> basic_istream<charT,traits>& getline(basic_istream<charT,traits>&& is, basic_string<charT,traits,Alloc>& str); // 21.3 Class template basic_string namespace std { template<class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_string { public: // types: typedef traits traits_type; typedef typename traits::char_type value_type; typedef Alloc allocator_type; typedef typename Alloc::size_type size_type; typedef typename Alloc::difference_type difference_type; typedef typename Alloc::reference reference; typedef typename Alloc::const_reference const_reference; typedef typename Alloc::pointer pointer; typedef typename Alloc::const_pointer const_pointer; typedef implementation-defined iterator; // See 23.1 typedef implementation-defined const_iterator; // See 23.1 typedef std::reverse_iterator<iterator> reverse_iterator; typedef std::reverse_iterator<const_iterator> const_reverse_iterator; static const size_type npos = -1; // 21.3.2 construct/copy/destroy: explicit basic_string(const Alloc& a = Alloc()); basic_string(const basic_string& str); basic_string(basic_string&& str); basic_string(const basic_string& str, size_type pos, size_type n = npos, const Alloc& a = Alloc()); basic_string(const charT* s, size_type n, const Alloc& a = Alloc()); basic_string(const charT* s, const Alloc& a = Alloc()); basic_string(size_type n, charT c, const Alloc& a = Alloc()); template<InputIterator Iter> basic_string(Iter begin, Iter end, const Alloc& a = Alloc()); basic_string(initializer_list<charT>, const Alloc& = Alloc()); basic_string(const basic_string&, const Alloc&); basic_string(basic_string&&, const Alloc&); ~basic_string(); basic_string& operator=(const basic_string& str); basic_string& operator=(basic_string&& str); basic_string& operator=(const charT* s); basic_string& operator=(charT c); basic_string& operator=(initializer_list<charT>); // 21.3.3 iterators: ... // 21.3.4 capacity: ... // 21.3.5 element access: ... // 21.3.6 modifiers: ... basic_string& append(const basic_string& str); basic_string& append(const basic_string& str, size_type pos, size_type n); basic_string& append(const charT* s, size_type n); basic_string& append(const charT* s); basic_string& append(size_type n, charT c); template<InputIterator Iter> basic_string& append(Iter first, Iter last); basic_string& append(initializer_list<charT>); void push_back(charT c); basic_string& assign(const basic_string& str); basic_string& assign(basic_string&& str); basic_string& assign(const basic_string& str, size_type pos, size_type n); basic_string& assign(const charT* s, size_type n); basic_string& assign(const charT* s); basic_string& assign(size_type n, charT c); template<InputIterator Iter> basic_string& assign(Iter first, Iter last); basic_string& assign(initializer_list<charT>); basic_string& insert(size_type pos1, const basic_string& str); basic_string& insert(size_type pos1, const basic_string& str, size_type pos2, size_type n); basic_string& insert(size_type pos, const charT* s, size_type n); basic_string& insert(size_type pos, const charT* s); basic_string& insert(size_type pos, size_type n, charT c); iterator insert(const_iterator p, charT c); void insert(const_iterator p, size_type n, charT c); template<InputIterator Iter> void insert(const_iterator p, Iter first, Iter last); void insert(const_iterator p, initializer_list<charT>); ... basic_string& replace(size_type pos1, size_type n1, const basic_string& str); basic_string& replace(size_type pos1, size_type n1, const basic_string& str, size_type pos2, size_type n2); basic_string& replace(size_type pos, size_type n1, const charT* s, size_type n2); basic_string& replace(size_type pos, size_type n1, const charT* s); basic_string& replace(size_type pos, size_type n1, size_type n2, charT c); basic_string& replace(iterator i1, iterator i2, const basic_string& str); basic_string& replace(iterator i1, iterator i2, const charT* s, size_type n); basic_string& replace(iterator i1, iterator i2, const charT* s); basic_string& replace(iterator i1, iterator i2, size_type n, charT c); template<InputIterator Iter> basic_string& replace(iterator i1, iterator i2, Iter j1, Iter j2); basic_string& replace(iterator, iterator, initializer_list<charT>); ... // 21.3.7 string operations: ... template <class charT, class traits, Allocator Alloc> struct constructible_with_allocator_suffix< basic_string<charT, traits, Alloc> > : true_type { }; |
LWG | 1081 | |
JP 47 | 21.3 |
Typo. Missing ”>”
template <class charT, class traits, Allocator Alloc should be template <class charT, class traits, Allocator Alloc> |
Correct typo. | editor | ACCEPTED | |
JP 48 | 21.3 |
char_traits does not use concept.
For example, create CharTraits concept and change as follows: template<class charT, CharTraits traits = char_traits<charT>, class Allocator = allocator<charT> > class basic_string { |
Add a concept for char_traits. | LWG | 1081 | |
UK 217 [340] |
21.3 | basic_string refers to constructible_with_allocator_suffix, which I thought was removed? | Remove the lines: template <class charT, class traits, class Alloc struct constructible_with_allocator_suffix< basic_string<charT, traits, Alloc> > : true_type { }; | editor | ACCEPTED | |
UK 218 [228] |
21.3.1 ¶ 3 | The identity "&*(s.begin() + n) == &*s.begin() + n" relies on operator& doing the "right thing", which (AFAICS) there is no requirement for. See my comment under clauses "23.2.1, 23.2.6" (p1 in both cases) - this is the same issue, but I've created a separate comment for basic_string because it is in a different chapter. | See my recommendations for "23.2.1, 23.2.6". | LWG | ||
UK 219 [342] |
21.3.6.6 [string.replace] ¶ 11 | Effects refers to "whose first begin() - i1 elements" However i1 is greater than begin()... | Effects refers to "whose first i1 - begin() elements" | editor | ACCEPTED | |
UK 220 [339] |
21.3.8.9 | The operator<< for basic_string takes the output stream by r-value reference. This is different from the same operator for error_code (19.4.2.6), bitset (20.2.6.3), shared_ptr (20.7.13.2.8), complex (26.3.6) and sub_match (28.4) | Use the same reference type for the all the library types. This should be the r-value reference. There are other places in the standard where TR1, and new classes, did not receive an 'R-value' update. | LWG | ||
FR 33 | 22.1.1 [locale] ¶ 3 |
ios_base::iostate err = 0;
iostate is a bitmask type and so could be an enumeration. Probably using goodbit is the solution. |
editor | ACCEPTED | ||
JP 49 | 22.1.3.2.2 |
codecvt does not use concept. For example, create
CodeConvert concept and change as follows.
template<CodeConvert Codecvt, class Elem = wchar_t> class wstring_convert { |
Add a concept for codecvt. | LWG | 1082 | |
JP 50 | 22.1.3.2.2 | Add custom allocator parameter to wstring_convert, since we cannot allocate memory for strings from a custom allocator. |
Correct as follows.
template<class Codecvt, class Elem = wchar_t> class wstring_convert { public: typedef std::basic_string<char> byte_string; typedef std::basic_string<Elem> wide_string; should be template<class Codecvt, class Elem = wchar_t, Allocator WideAllocator = allocator<Elem>, Allocator ByteAllocator = allocator<char>> class wstring_convert { public: typedef std::basic_string<char, char_traits<char>, ByteAllocator> byte_string; typedef std::basic_string<Elem, char_traits<Elem>, WideAllocator> wide_string; |
LWG | 991 | |
FI 4 | 22.2.1.4.1 22.2.1.4.2 | to_end and to_limit are both used. Only one is needed. | Change to_limit to to_end. | editor | ACCEPTED | |
FI 5 | 22.2.1.4.2 ¶ #3 |
[ Note: As
a result of operations on state, it can return ok or
partial and set next == from and to_next != to. —end
note ]
"next" should be "from_next." Also, the sentence applies to all the examples, including do_in and do_out. Reasoning: When reading one element of multibyte source data such as UTF-8, it is possible that from_next is incremented, to_next is unaltered, state is altered and return value is partial. When reading one element of wide character data, the output can be several multibyte characters, so it is possible that from_next is unaltered, to_next is advanced, state is altered and return value is partial. |
[ Note: As a result of operations on state, do_in and do_out can return ok or partial and set from_next == from and/or to_next != to. —end note ] | editor | ACCEPTED | |
FI 6 |
22.2.1.5
See also 22.2.1.4 (1,2,3) |
codecvt_byname is only specified to work with locale names. There is no built-in means to find a codecvt with a character set's name. A locale and a character set are not the same. If the user has data which is encoded in a certain character set and she wants to create a codecvt which can convert from that character set to another one, she must iterate through locales to find one, or use external means (iconv, ICU's uconv). Specifying a locale with the character set is not a suitable solution, since there is no built-in mapping from character sets to locales. It is only possible to inquire the character set once the locale is known. | There should be a built-in means to find a codecvt with a pair of character set names. | LWG | ||
FI 7 | 22.2.1.4 ¶ 1,2,3 | The word "codeset" is used, whereas the word "character set" is used elsewhere in the text. The words are intended to convey the same concept, so only one should be used (or always both together). | Change "codeset" to "character set." | editor | ACCEPTED | |
JP 51 | 22.2.5.1.1 ¶ 7th para, 1st line |
A
parameter `end’ should be `fmtend’.
get() function had two `end’ parameters at N2321.
iter_type get (iter_type s, iter_type end, ios_base& f, ios_base::iostate& err, tm* t, const char_type* fmt, const char_type *end) const; The function prototype of get() has been corrected at N2800, but a Requires statement still refers `end’ parameter. |
Correct as follows.
Requires: [fmt,end) shall be a valid range. should be Requires: [fmt,fmtend) shall be a valid range. |
editor | ACCEPTED | |
JP 52 | 22.2.5.1, 22.2.5.2, 22.2.6.1 | InputIterator does not use concept. |
Correct as follows.
22.2.5.1 template <class charT, class InputIterator = istreambuf_iterator<charT> > class time_get : public locale::facet, public time_base { public: typedef charT char_type; typedef InputIterator iter_type; should be template <class charT, InputIterator InputIter = istreambuf_iterator<charT> > class time_get : public locale::facet, public time_base { public: typedef charT char_type; typedef InputIter iter_type; 22.2.5.2 template <class charT, class InputIterator = istreambuf_iterator<charT> > class time_get_byname : public time_get<charT, InputIterator> { public: typedef time_base::dateorder dateorder; typedef InputIterator iter_type; should be template <class charT, InputIterator InputIter = istreambuf_iterator<charT> > class time_get_byname : public time_get<charT, InputIter> { public: typedef time_base::dateorder dateorder; typedef InputIter iter_type; 22.2.6.1 template <class charT, class InputIterator = istreambuf_iterator<charT> > class money_get : public locale::facet { public: typedef charT char_type; typedef InputIterator iter_type; should be template <class charT, InputIterator InputIter = istreambuf_iterator<charT> > class money_get : public locale::facet { public: typedef charT char_type; typedef InputIter iter_type; |
LWG | 1083 | |
JP 53 | 22.2.5.3 , 22.2.5.4 | OutputIterator does not use concept. |
Correct as follows.
22.2.5.3 template <class charT, class OutputIterator = ostreambuf_iterator<charT> > class time_put : public locale::facet { public: typedef charT char_type; typedef OutputIterator iter_type; should be template <class charT, OutputIterator OutputIter = ostreambuf_iterator<charT> > class time_put : public locale::facet { public: typedef charT char_type; typedef OutputIter iter_type; 22.2.5.4 template <class charT, class OutputIterator = ostreambuf_iterator<charT> > class time_put_byname : public time_put<charT, OutputIterator> { public: typedef charT char_type; typedef OutputIterator iter_type; should be template <class charT, OutputIterator OutputIter = ostreambuf_iterator<charT> > class time_put_byname : public time_put<charT, OutputIter> { public: typedef charT char_type; typedef OutputIter iter_type; |
LWG | 1083 | |
JP 54 | 23 ¶ 2nd para, Table 79 | There is not <forward_list> in Table 79. | Add <forward_list> between <deque> and <list>. | editor | ACCEPTED | |
UK 221 [287] |
23 ¶ Table 79 | The table is missing the new <forward_list> header. | Add <forward_list> to the table for sequence containers. Alternative (technical) solution might be to merge <forward_list> into <list>. | editor | ACCEPTED | |
UK 222 [295] |
23 | It is not clear what purpose the Requirement tables serve in the Containers clause. Are they the definition of a library Container? Or simply a conventient shorthand to factor common semantics into a single place, simplifying the description of each subsequent container? This becomes an issue for 'containers' like array, which does not meet the default-construct-to-empty requirement, or forward_list which does not support the size operation. Are these components no longer containers? Does that mean the remaining requirements don't apply? Or are these contradictions that need fixing, despite being a clear design decision? | Clarify all the tables in 23.1 are there as a convenience for documentation, rather than a strict set of requirements. Containers should be allowed to relax specific requirements if they call attention to them in their documentation. The introductory text for array should be expanded to mention a default constructed array is not empty, and forward_list introduction should mention it does not provide the required size operation as it cannot be implemented efficiently. | LWG | 1034 | |
JP 55 | 23.1.1 ¶ 3rd para, 4th line | It seems that “the MinimalAllocator concep” is the typo of “the MinimalAllocator concept”. | Change to … models the MinimalAllocator concept. | editor | ACCEPTED | |
UK 223 [285] |
23.1.1 ¶ 3 | The library does not define the MinimalAllocator or ScopedAllocator concepts, these were part of an earlier Allocators proposal that was rejected. | Remove the references to MinimalAllocator and ScopedAllocator, or add definitions for these concepts to clause 20.7. | LWG |
ACCEPTED
See paper N2840 |
|
UK 224 [298] |
23.1.1 ¶ 8 | This paragraph implicitly requires all containers in clause 23 to support allocators, which std::array does not. | Add an 'unless otherwise specified' rider somewhere in p8, or move the whole array container from clause 23 [containers] to clause 20 [utilies] to accompany bitset, pair and tuple. | LWG |
ACCEPTED with MODIFICATIONS
See paper N2840 |
|
UK 225 [299] |
23.1.1 ¶ Table 81 | Inconsistent words used to say the same thing. Table 80 describes iterator/const_iterator typedef as returning an "iterator type whose value type is T". Table 81 expresses the same idea as an "iterator type pointing to T". Express identical ideas with the same words to avoid accidentally introducing subtlety and confusion | Change return types for X::(const)_reverse_iterator to say "iterator type whose value type is (const) T". | editor | ACCEPTED | |
UK 226 [336] |
23.1.1 ¶ 10 | <array> must be added to this list. In particular it doesn't satisfy: - no swap() function invalidates any references, pointers, or iterators referring to the elements of the containers being swapped. and probably doesn't satisfy: — no swap() function throws an exception. | If <array> remains a container, this will have to also reference array, which will then have to say which of these points it satisfies. | LWG | 1035 | |
UK 227 [146] |
23.1.1 ¶ Table 80 | The post-condition for a = rv uses the word “construction” when it means “assignment” | Replace the word “construction” with the word “assignment” | editor | ACCEPTED | |
UK 228 [283] |
23.1.1 ¶ 3 | Line 4 contains a spelling mistake in the fragment "MinimalAllocator concep." | Replace "concep" with "concept" | editor | ACCEPTED | |
UK 229 [284] |
23.1.1 ¶ 3 | The fragment "A container may directly call constructors" is not technically correct as constructors are not callable. | Replace "A container may directly call constructors and destructors for its stored objects" with something similar to "A container may directly construct its stored objects and call destructors for its stored objects" | editor | ACCEPTED | |
UK 230 [147] |
23.1.2 ¶ 1 | “implementations shall consider the following functions to be const” - what does this mean? I don't understand what it means by implementations considering the functions to be const – surely they are either declared const or not? | Clarify what is meant and what requirements an implementation must satisfy. | LWG |
REJECTED
After considerable discussion and consideration, we do not feel this is a defect given the reference to [res.on.data.races]. |
|
JP 56 | 23.1.3 ¶ 12th para, Table 84 | `array’ is unstated in Table 84 - Optional sequence container operations. |
Add
`array’ to Container field for the following
Expression.
a.front() a.back() a[n] a.at(n) |
editor | ACCEPTED | |
UK 231 [300] |
23.1.3 ¶ 9-11 | These paragraphs are redundant now that Concepts define what it means to be an Iterator and guide overload resolution accordingly. | Strike 23.1.3p9-11. Make sure std::basic_string has constraints similar to std::vector to meet this old guarantee. | LWG | 1036 | |
UK 232 [301] |
23.1.3 ¶ Table 84 | match_results may follow the requirements but is not listed a general purpose library container. | Remove reference to match_results against a[n] operation | LWG | 1037 | ACCEPTED |
UK 233 [302] |
23.1.3 ¶ Table 84 | Add references to the new containers. | Add reference to array to the rows for: a.front(), a. back(), a[n] a.at(n). Add reference to forward_list to the rows for: a.front(), a.emplace_front(args), a.push_front(t), a.push_front(rv), a.pop_front(). Add reference to basic_string to the row for: a.at(n). | LWG | 1038 | ACCEPTED |
UK 234 [346] |
23.1.3 ¶ Table 84 | Ther reference to iterator in semantics for back should also allow for const_iterator when called on a const-qualified container. This would be ugly to specify in the 03 standard, but is quite easy with the addition of auto in this new standard. | Replace iterator with auto in semantics for back: { auto tmp = a.end(); --tmp; return *tmp; } | LWG | 1039 | ACCEPTED |
UK 235 [148] |
23.1.3 ¶ 1 | “The library provides three basic kinds of sequence containers: vector, list, and deque” - text appears to be out of date re addition of array and forward_list | Change the text to read: “The library provides five basic kinds of sequence containers: array, deque, forward_list, list and vector”. | editor | ACCEPTED | |
UK 236 [150] |
23.1.3 ¶ 2 | [ I've moved (1) into a separate comment because I believe it is editorial in the simple sense, whereas (2) and (3) are not so straight forward ] (2) “vector is the type of sequence container that should be used by default” -- As I understand it vector was considered first port of call because the things it has in common with the native array make programmers (especially those new to the container library) feel like they are on familiar territory. However, we now have the array container, so I think this should be recommended as the first port of call. (3) This paragraph is actually giving guidance on the use of the containers and should not be normative text | Remove this paragraph | editor |
ACCEPTED with MODIFICATIONS
Reworded the paragraph and its predecessor to reflect five basic sequence containers. |
|
UK 237 [334] |
23.1.3 ¶ 2 | vector, list, and deque offer the programmer different complexity trade-offs and should be used accordingly - this ignores array and forward_list | Modify the text to read: "array, deque, forward_list, list and vector offer the programmer different complexity trade-offs and should be used accordingly" | editor | ACCEPTED | |
UK 238 [347] |
23.1.4 ¶ 6 | Leaving it unspecified whether or not iterator and const_iterator are the same type is dangerous, as user code may or may not violate the One Definition Rule by providing overloads for both types. It is probably too late to specify a single behaviour, but implementors should document what to expect. Observing that problems can be avoided by users restricting themselves to using const_iterator, add a note to that effect. | Change 'unspecified' to 'implementation defined'. Add "[Note: As itererator and const_iterator have identical semantics in this case, and iterator is convertible to const_iterator, users can avoid violating the One Definition Rule by always using const_iterator in their function parameter lists -- end note] | LWG | 1040 | ACCEPTED with MODIFICATIONS |
UK 239 [447] |
23.1.4 ¶ 85 | It is not possible to take a move-only key out of an unordered container, such as (multi)set or (multi)map, or the new hashed containers. | Add below a.erase(q), a.extract(q), with the following notation: a.extract(q), Return type pair<key, iterator> Extracts the element pointed to by q and erases it from the set. Returns a pair containing the value pointed to by q and an iterator pointing to the element immediately following q prior to the element being erased. If no such element exists,returns a.end(). | LWG | 1041 | |
UK 240 [427] |
23.1.6.1 ¶ 12 | The axiom EmplacePushEquivalence should be asserting the stronger contract that emplace and insert return the same iterator value, not just iterators that dereference to the same value. This is a similar statement that is easier to express and should be equivalent - the idea that insert and emplace might return iterator values that do not compare equal but point to the same element should fail somewhere in the iterator concepts. Also, this axiom should be renamed to reflect its connection with insert, rather than push_front/push_back, | Remove the * to deference the returned iterator either side of the == in the EmplacePushEquivalence axiom, rename the axiom EmplacementInsertionEquivalence : requires InsertionContainer<C> && Constructible<value_type, Args...> axiom EmplacementInsertionEquivalence(C c, const_iterator position, Args... args) { emplace(c, position, args...) == insert(c, position, value_type(args...)); } | LWG | ||
JP 57 | 23.1.6.3 ¶ 1st para, 4th line |
Typo, duplicated "to"
"to to model insertion container concepts." |
Remove one. | editor | ACCEPTED | |
UK 241 [286] |
23.2.1 | std::array does not have an allocator, so need to document an exception to the requirements of 23.1.1p3 | add exception to 23.1.1p3 | LWG | ||
UK 242 [355] |
23.2.1 ¶ 3 | std:: qualification no longer needed for reverse_iterator. | remove std:: qualification from std::reverse_iterator<iterator> and std::reverse_iterator<const_iterator> | editor | ACCEPTED | |
UK 243 [337] |
23.2.1 ¶ 3 | Most containers, and types in general have 3 swaps: swap(T&, T&) swap(T&&, T&) swap(T&, T&&) But array only has swap(T&, T&). | add the other two swaps. | LWG |
REJECTED
After discussing D2844, it was decided to remove all the r-value-ref swap overloads from containers.
Therefore adding them to array has no benefit. So the final
disposition is NAD.
|
|
UK 244 [153] |
23.2.1, 23.2.6 ¶ 1 | The validity of the expression &a[n] == &a[0] + n is contingent on operator& doing the “right thing” (as captured by the CopyConstructible requirements in table 30 in C++2003). However this constraint has been lost in the Concepts of C++0x. This applies to vector and array (it actually applies to string also, but that's a different chapter, so I'll file a separate comment there and cross-reference). | Define a ContiguousStrorage and apply it to vector,array and string. The Concept (supplied by Alisdair M) looks like this: Concept< typename C > ContiguousStrorage { requires Container<C>; typename value_type = C::value_type; value_type * data( C ); axiom Contiguous { C c; true = equal_ranges( data( c), data(c) + size(c), begin(c)); } }; | LWG | 1042 | |
UK 245 [358] |
23.2.3 ¶ 2 | The predicate types used in special member function of forward_list should be CopyConstructible, as per the algorithms of the same name. Note: an alternate solution would be to require these callable concepts to be CopyConstructible in clause 20, which would simplify the library specification in general. See earlier comment for details, that would render this one redundant. | Add CopyConstructible requirement to the following signatures: template <Predicate<auto, T> Pred> requires CopyConstructible<Pred> void remove_if(Pred pred); template <EquivalenceRelation<auto, T> BinaryPredicate> requires CopyConstructible<BinaryPredicate> void unique(BinaryPredicate binary_pred); template <StrictWeakOrder<auto, T> Compare> requires CopyConstructible<Compare> void merge(forward_list<T,Alloc>&& x, Compare comp); template <StrictWeakOrder<auto, T> Compare> requires CopyConstructible<Compare> void sort(Compare comp); | LWG |
REJECTED
UK245 with additional comments on UK200 in clause 20: After further discussion of UK200, we do not think adding constraints to predicates is a good idea. Straw poll on UK245: status quo, 5 pro - 0 con; add copy-constructible, 1 pro - 3 con; no consensus for moving away from the status quo. |
|
JP 58 | 23.2.3.2 ¶ 1st line before 1st para | Unnecessary "{" exists before a word iterator like "{iterator before_begin()". | Remove "{" | editor | ACCEPTED | |
JP 59 | 23.2.4.4 | Types of the third and forth parameter of splice() are iterator at 23.2.4.4, though types of them are const_iterator at 23.2.4. (They are both const_iterator on N2350) |
Correct as follows.
void splice(const_iterator position, list<T,Allocator>&& x, iterator i); void splice(const_iterator position, list<T,Allocator>&& x, iterator first, iterator last); should be void splice(const_iterator position, list<T,Allocator>&& x, const_iterator i); void splice(const_iterator position, list<T,Allocator>&& x, const_iterator first, const_iterator last); |
editor | ACCEPTED | |
US 83 | 23.2.6.2 ¶ 7 | "shrink_to_fint" should be "shrink_to_fit". | editor | ACCEPTED | ||
UK 246 [350] |
23.3.2.2 | The content of this sub-clause is purely trying to describe in words the effect of the requires clauses on these operations, now that we have Concepts. As such, the desctiption is more confusing than the signature itself. The semantic for these functions is adequately covered in the requirements tables in 23.1.4. | Strike 23.3.2.2 entirely. (but do NOT strike these signatures from the class template definition!) | LWG | 1091 | This comment was submitted as editorial, but the project editor has classified it as technical. |
UK 247 [46] |
24.1 | Iterator concepts are not extensive enough to merit a whole new header, and should be merged into <concpts>. This is particularly important for supporting the new for loop syntax which requires access to the Range concept. The required header to enable this syntax shoud have a simple name, like <concepts>, rather than something awkward to type like <iterator_concepts>. | Move the concepts of <iterator_concepts> into the <concepts> header. We take no position on moving the text from Clause 24 to Clause 20 though. | LWG |
REJECTED
NAD. We believe the separate header to have value. |
|
UK 248 [47] |
24.1 ¶ 6 | The text "so for any iterator type there is an iterator value that points past the last element of a corresponding container" is slightly misleading. Iterators can refer into generalised ranges and sequences, not just containers. A broader term than 'container' should be used. | Replace the reference to container with a more appropriate concept | editor | ACCEPTED | |
UK 250 [50] |
24.1.1 | A default implementation should be supplied for the post-increment operator to simplify implementation of iterators by users. | Copy the Effects clause into the concept description as the default implementation. Assumes a default value for postincrement_result | LWG | 1084 | |
UK 251 [51] |
24.1.1 ¶ 3 | The post-increment operator is dangerous for a general InputIterator. The multi-pass guarantees that make it meaningful are defined as part of the ForwardIterator refinement. Any change will affect only constrained templates that have not yet been written, so should not break existing user iterators which remain free to add these operations. This change will also affect the generalised OutputIterator, although there is no percieved need for the post-increment operator in this case either. | Move declaration of postincrement operator and postincrement_result type from Interator concept to the ForwardIterator concept | LWG | 1009 | |
UK 252 [53] |
24.1.2 ¶ 3 | istream_iterator is not a class, but a class template | Change 'class' to 'class template' in the note. | editor | ACCEPTED | |
UK 253 [54] |
24.1.3 ¶ 1 | First sentance does not make gramatical sense, Seems to be missing the words 'if it' by comparison with similar sentance in other subsections | Add the words 'if it' : "X satisfies the requirements of an output iterator IF IT meets the syntactic and semantic requirements" | editor | ACCEPTED | |
UK 254 [55] |
24.1.3 ¶ 5 | This postcondition for pre-increment operator should be written as an axiom | Move the postcondition into the concept definition as an axiom | LWG | ||
UK 255 [56] |
24.1.4 ¶ 4 | This postcondition for pre-increment operator should be written as an axiom | Move the postcondition into the concept definition as an axiom | LWG | ||
UK 256 [57] |
24.1.5 ¶ 3, 4, 5 | The relationship between pre- and post- decrement should be expressed as an axiom. | Move the text specification of pre/post-conditions and behaviour into an axiom within the BidirectionalIterator concept | LWG | ||
UK 257 [58] |
24.1.5 | There is a reasonable default for postdecrement_result type, which is X. X is required to be regular, therefore CopyConstructible and a valid ResultType. Together with the next comment this simplifies user defined iterator implentations | Add the default : typename postincrement_result = X; | LWG |
REJECTED
NAD, postdecrement_result is deduced. |
|
UK 258 [50] |
24.1.5 | A default implementation should be supplied for the post-decrement operator to simplify implementation of iterators by users. | Copy the Effects clause into the concept description as the default implementation. Assumes a default value for postincrement_result | LWG | 1085 | |
UK 259 [60] |
24.1.5 | postdecrement_result is effectively returning a copy of the original iterator value, so should have similar constraints, rather than just HasDereference. If Concepts do not support this circularity of definition suggest that concepts feature may want a little more work | Add the requirement: requires Iterator< postdecrement_result >; | LWG |
REJECTED
NAD unless Alisdair comes back with more motivation. |
|
UK 260 [61] |
24.1.5 ¶ 6 | The effects clause for post-decrement iterator should be available as an axiom and a default implementation, where the compiler can make better use of it. | Move the Effects clause into the BidirectionalIterator concept definition as an axiom, and as the default implementation for the operation. | LWG | ||
UK 249 [64] |
24.1.6 ¶ 2 | The semantic for operator+= should also be provided as a default implementation to simplify implementation of user-defined iterators | Copy the text from the effects clause into the RandomAccessIterator concept as the default implementaiton. | LWG |
REJECTED
NAD, violates complexity requirements. |
|
UK 261 [62] |
24.1.6 | To simplify user defined random access iterator types, the subscript_reference type should default to reference | typename subscript_reference = reference; | LWG |
REJECTED
NAD, subscript reference isn't used. In c++std-lib-23104, Daniel Krügler commented, “this [proposed change] would disable "auto-deduction" of the return type of any operator[] as described by [concept.map.assoc]/4+5.” |
|
UK 262 [65] |
24.1.6 ¶ 3, 4 | Effects and post-conditions for operator+ are more useful if expressed as axioms, and supplied as default implementations. | Move the effects and Postcondition definitions into the RandomAccessIterator concept and copy the code in the specification as the default implementation of these operations. | LWG | ||
UK 263 [66] |
24.1.6 ¶ 5 | This requirement on operator-= would be better expressed as a default implementation in the concept, with a matching axiom | Move the specification for operator-= from the returns clause into an axiom and default implementation within the RandomAccessIterator concept | LWG | 1010 | |
UK 264 [67] |
24.1.6 ¶ 6 | Effects clauses are better expressed as axioms where possible. | Move code in operator- effects clause into RandomAccessIterator concept as both a default implementation and an axiom | LWG | ||
UK 265 [68] |
24.1.6 ¶ 8 | This effects clause is nonesense. It looks more like an axiom stating equivalence, and certainly an effects clause cannot change the state of two arguments passed by const reference | Strike the Effects clause | LWG | 1079 | |
UK 266 [69] |
24.1.6 ¶ 9 | This sentance should be provided as a default definition, along with a matching axiom | Move the Returns clause into the spefication for RandomAccessIterator operator- as a default implementation. Move the Effects clause as the matching axiom. | LWG | ||
UK 267 [70] |
24.1.6 ¶ 24.1.6 | The code in the Requires clause for RandomAccessIterator operator[] would be better expressed as an axiom. | Rewrite the Requires clause as an axiom in the RandomAccessIterator concept | LWG | ||
UK 268 [71] |
24.1.6 ¶ 12 | This note is potentialy confusing as __far enters the syntax as a clear language extension, but the note treats it as a regular part of the grammar. It might be better expressed using attributes in the current wording. | Clean up the note to either avoid using language extension, or spell out this is a constraint to support extensions. | editor | ACCEPTED | |
JP 60 | 24.1.8 ¶ 1st para |
Capability of an iterator is too much restricted by
concept.
Concept of std::Range is defined as: concept Range<typename T> { InputIterator iterator; iterator begin(T&); iterator end(T&); } So the following code generates an error. template <std::Range Rng> void sort(Rng& r) { // error! Rng::iterator does not satisfy requirements of a random // access iterator. std::sort(begin(r), end(r)); } std::vector<int> v; // vector::iterator is a random access iterator. sort(v); This is because the concept of an iterator of std::Range is InputIterator. For this reason, a random access iterator loses its capability when passed to a std::Range parameter. To be able to work the above code, we need to write as follows: template <std::Range T> requires std::RandomAccessIterator<T::iterator> && std::ShuffleIterator<T::iterator> && std::LessThanComparable<T::iterator::value_type> void sort(T& r) { sort(begin(r), end(r)); } std::vector<int> v; sort(v); It needs quiet a few amount of codes like this only to recover random access capability from InputIterator concept. We can write the following valid code with Boost.Range, which is implemented with using C++03 SFINAE. template <class Range> void sort(Range& r) { std::sort(boost::begin(r), boost::end(r)); } std::vector<int> v; sort(v); // OK One of the motivation to introduce concepts are supporting template programming techniques by language directly to eliminate hacky techniques such as tag-dispatch, SFINAE and Type Traints. But SFINAE will be kept using because it needs quite a few amount of codes without using SFAINAE. |
Add InputRange, OutputRange, ForwardRange, BidirectionalRange and RandomAccessRange. | LWG |
REJECTED
NAD, beyond the scope of the Standard for C++0x because we are not supplying range-based algorithms. |
|
UK 269 [16] |
24.3 ¶ 3 | 'decrements for negative n' seems to imply a negative number of decrement operations, which is odd. | Need simple, clearer wording | editor | ACCEPTED | |
UK 270 [17] |
24.3 ¶ 4 | The reachability constraint in p5 means that a negavite result, implying decrements operations in p4, is not possible | Split the two overloads into separate descriptions, where reachability is permitted to be in either direction for RandomAccessIterator | LWG | 940 | |
UK 271 [145] |
24.3 ¶ 6,7 | next/prev return an incremented iterator without changing the value of the original iterator. However, even this may invalidate an InputIterator. A ForwardIterator is required to guarantee the 'multipass' property. | Replace InputIterator constraint with FOrwardIterator in next and prev function templates. | LWG | 1011 | |
UK 272 [159] |
24.4 | reverse_iterator and move_iterator use different formulations for their comparison operations. move_iterator merely requires the minimal set of operations, < and ==, from its underlying iterator and synthesizes all oprations from these two. reverse_iterator relies on the undelying iterator to support all comparision operations directly. In practice, move_iterator can only be implemented this way as it must support iterator types that are merely InputIterators, and so SemiRegular and not Regular. However, reserse_iterator has an existing specification and any change of semantic could change behaviour of conforming programs today - although an iterator that yields different results for (a > b) than (b < a) may fall foul of some semantic consistency requirements, even if the syntax is met. | Rephrase the reverse_iterator comparison operations using only operators < and ==, as per the move_iterator specification. | LWG |
REJECTED
NAD, no consensus. |
|
UK 274 [119] |
24.4, 24.5 | The subclauses for standard iterator adaptors could be better organised. There are essentially 3 kinds of iterator wrappers provided, stream iterators adapt streams and get their own subsection. insert iterators adapt containers, and get their own subsection but it is inserted into the middle of 24.4 Predifined iterators. reverse_iterator and move_iterator adpat other iterators, but their presentation is split by insert iterators | Promote 24.4.2 [insert.iterators] up one level to 24.6. Emphasize that insert iterators adapt containers Retarget 24.4 [predef.iterators] as iterator adapters for iterator templates that wrap other iterators. | editor | ACCEPTED | |
UK 275 [18] |
24.4.1.1 | The constructor template taking a single Iterator argument will be selected for Copy Initialization instead of the non-template constructor taking a single Iterator argument selected by Direct Initialization. | The reverse_iterator template constructor taking a single Iterator argument should be explicit. | LWG |
REJECTED
NAD, withdrawn by submitter. |
|
UK 276 [19] |
24.4.1.1 | It is odd to have a mix of declaration stlyes for operator+ overloads. Prefer if either all are member functions, or all are 'free' functions. | Make the member operators taking a difference_type argument non-member operators | LWG |
REJECTED
NAD, not editorial, withdrawn by submitter. |
|
UK 277 [20] |
24.4.1.2.1 ¶ 1 | The default constructor default-initializes current, rather than value-initializes. This means that when Iterator corresponds to a trivial type, the current member is left un-initialized, even when the user explictly requests value intialization! At this point, it is not safe to perform any operations on the reverse_iterator other than assign it a new value or destroy it. Note that this does correspond to the basic definition of a singular iterator. | i/ Specify value initialization rather than default intialization or ii/ specify = default; rather than spell out the semantic. This will at least allow users to select value initialization and copy the iterator value. or iii/ Add a note to the description emphasizing the singular nature of a value-initialized reserve iterator. | LWG | 1012 | |
UK 278 [21] |
24.4.1.2.1 ¶ 3 | There is an inconsistency between the constructor taking an iterator and the constructor template taking a reverse_iterator where one passes by value, and the other by const reference. The by-value form is preferred as it allows for efficient moving when copy elision fails to kick in. | Change the const reverse_iterator<U> & parameter to pass-by-value | LWG |
REJECTED
NAD, we don't believe that copy elision is a sufficiently high priority for iterator types. |
|
UK 279 [24] |
24.4.1.2.12, 24.4.3.2.12 | The reason the return type became unspecified is LWG issue 386. This reasoning no longer applies as there are at least two ways to get the right return type with the new language facilities added since the previous standard. | Specify the return type using either decltype or the Iter concept map | LWG | 1051 | |
UK 280 [22] |
24.4.1.2.4 | The presence of the second iterator value is surprising for many readers who underestimate the size of a reverse_iterator object. Adding the exposition only member that is required by the semantic will reduce confusion. | Add reverse_iterator expsoition only member tmp as a comment to class declaration, as a private member | editor | ACCEPTED | |
UK 281 [23] |
24.4.1.2.5 | The current specification for return value will always be a true pointer type, but reverse_iterator supports proxy iterators where the pointer type may be some kind of 'smart pointer' | Replace the existing returns specification with a copy of the operator* specification that returns this->tmp.operator-> | LWG | 1052 | |
UK 282 [5] |
24.4.2.1, 24.4.2.2.2, 24.4.2.3, 24.4.2.4.2, 24.4.2.5, 24.4.2.6.2 ¶ n/a | Insert iterators of move-only types will move from lvalues | Add an additional constrained overload for operator= that requires !CopyConstructible<Cont::value_type> and mark it =delete. | LWG | 901 | |
UK 283 [156] |
24.4.2.5, 24.4.2.6.4 | postincrement operator overloads traditionally return by value - insert_iterator is declared as return by reference. The change is harmless in this case, but either front/back_insert_iterator should take the matching change for consistency, or this function should return-by-value | change operator++(int) overload to return by value, not reference. Affects both class summary and operator definition in p | LWG |
REJECTED
NAD. This is compatible with C++03; and we lack a consensus for change. |
|
JP 61 | 24.4.3.2.1 ¶ 2nd para, 1st line |
Typo.
"intializing" should be "initializing" |
Add "i" | editor | ACCEPTED | |
UK 284 [26] |
24.5 | The stream iterators need constraining with concepts/requrires clauses. | Provide constraints | LWG | 1086 | ACCEPTED |
UK 285 [27] |
24.5.1 ¶ 1,2 | Much of the content of p1 and the whole of p2 is a redundant redefinition of InputIterator. It should be simplified | Strike p2. Simplify p1 and add a cross-reference to the definition of InputIterator concept. | editor | ||
UK 286 [28] |
24.5.1 ¶ 3 | To the casual reader it is not clear if it is intended to be able to assign to istream_iterator objects. Specifying the copy constructor but relying on the implicit copy-assign operator is confusing. | Either provide a similar definition to the copy-assign operator as for the copy constructor, or strike the copy constructor | editor | ||
UK 287 [25] |
24.5.1.1 ¶ 2 | It is not clear what the intial state of an istream_iterator should be. Is _value_ initialized by reading the stream, or default/value initialized? If it is initialized by reading the stream, what happens if the initialization is deferred until first dereference, when ideally the iterator value should have been that of an end-of-stream iterator which is not safely dereferencable? | Specify _value_ is initialized by reading the stream, or the iterator takes on the end-of-stream value if the stream is empty | LWG | 788 | |
UK 288 [29] |
24.5.1.1 ¶ 3 | The provided specification is vacuous, offering no useful information. | Either strike the specification of the copy constructor, or simply replace it with an = default definition. | editor | ||
UK 289 [30] |
24.5.1.2 ¶ 6 | It is very hard to pick up the correct specification for istream_iterator::operator== as the complete specification requires reading two quite disconnected paragraphs, 24.5.1p3, and 24.5.1.2p6. Reading just the specifaction of the operation itself suggests that end-of-stream iterators are indistinguishable from 'valid' stream iterators, which is a very dangerous misreading. | Merge 24.5.1p3, equality comparison of end-of-stream-iterators, into 24.5.1.2p6, the specification of the equality operator for istream_iterator. | editor | ||
UK 290 [160] |
24.5.2 ¶ 1 | The character type of a string delimiter for an ostream_iterator should be const charT *, the type of the elements, not char *, a narrow character string. | Replace char * with const charT * | editor | ACCEPTED | |
UK 291 [161] |
24.5.2.2 ¶ 2 | ostream_iterator postincrement operator returns by reference rather than by value. This may be a small efficiency gain, but it is otherwise unconventional. Prefer return-by-value. | ostream_iterator operator++(int); | LWG | REJECTED | |
FR 34 | 24.5.3 [istreambuf. iterator] | There are two public sections, and the content of the second one is indented with respect to the first. I don't it should be. | editor | ACCEPTED | ||
UK 292 [163] |
24.5.3 ¶ 1 | Prefer the use of the new nullptr constant to the zero literal when using a null pointer in text. | Change istreambuf_iterator(0) to istreambuf_iterator(nullptr) | editor | ||
UK 293 [164] |
24.5.3 ¶ 2,3,4 | The listed paragraphs redundantly redefine an input iterator, and redundancy can be a dangerous thing in a specification. Suggest a simpler phrasing below. | 2. The result of operator*() on an end of stream is undefined. For any other iterator value a char_type value is returned. 3. Two end of stream iterators are always equal. An end of stream iterator is not equal to a non-end of stream iterator. 4. As istreambuf_iterator() is an InputIterator but not a ForwardIterator, istreambuf_iterators object can be used only for one-pass algorithms. It is not possible to assign a character via an input iterator. | editor | ||
UK 294 [166] |
24.5.3.2 ¶ 2 | Implicit converting constructors can be invoked at surprising times, so there should always be a good reason for declaring one. | Mark the two single-argument constructors take a stream or stream buffer as explicit. The proxy constructor should remain implicit. explicit istreambuf_iterator(basic_istream<charT,traits>& s) throw(); explicit istreambuf_iterator(basic_streambuf<charT,traits>* s) throw(); | LWG | REJECTED | |
UK 295 [173] |
25 | THere is a level of redundancy in the library specification for many algorithms that can be eliminated with the combination of concepts and default parameters for function templates. Eliminating redundancy simplified specification and reduces the risk of inttroducing accidental inconsistencies. | Adopt n2743, or an update of that paper. | LWG | 1053 | |
JP 62 | 25, 25.3.1.5, 26.3.6.5 |
The return types of is_sorted_until function and
is_heap_until function are iterator. But basically, the
return type of is_xxx function is bool. And the return type
of lower_bound function and upper_bound is iterator.
So we think that it is reasonable to change those two functions. |
Change "is_sorted_until" to "sorted_bound"
Change "is_heap" to "heap_bound" |
LWG |
REJECTED
No consensus to make the change. |
|
UK 296 [183] |
25.1.8 ¶ 1 | The 'Returns' of adjacent_find requires only HasEqualTo, or a Predicate. Requiring EqualityComparable or EquivalenceRelation seems too strong and not useful. | Change EqualityComparable to HasEqualTo and EquivalnceRelation to Predicate | LWG |
REJECTED
No consensus to make the change. |
|
UK 297 [185] |
25.2.11 ¶ 6 | The definition of rotate_copy is very complicated. It is equivalent to: return copy(first, middle, copy(middle, last, result)); | Change 'effects' to, returns, requires, complexity to: effects: equivalent to: return copy(first, middle, copy(middle, last, result)); | editor | ACCEPTED | |
UK 298 [186] |
25.2.13 ¶ 13 | partition_point requires a partitioned array | requires: is_partitioned(first, last, pred) != false; | editor | ACCEPTED | |
UK 299 [352] |
25.2.2 | Should be consistent in style use of concepts in template parameter lists. The auto-OutputIterator sytle used in std::copy is probably preferred. | Change way signature is declared: template<InputIterator InIter, OutputIterator<auto, RvalueOf<InIter::reference>::type> OutIter> OutIter move(InIter first, InIter last, OutIter result); | editor | ||
UK 300 [351] |
25.2.3 | Since publishing the original standard, we have learned that swap is a fundamental operation, and several common idioms rely on it - especially those related to exception safety. As such it belongs in the common <utility> header rather than the broader <algorithm> header, and likewise should move to clause 20. For backwards compatiblility the algorithm header should be required to #include <utility>, which would be covered in the resolution of LWG issue 343. There are already dependencies in <algorithm> on types declared in this header, so this comment does not create a new dependency. | Move primary swap template from <algorithm> into <utility>. Move 25.2.3 to somewhere under 20.2. Require <algorithm> to #include <utility> to access pair and provide legacy support for finding swap. | LWG |
REJECTED
No consensus to make the change. |
|
UK 301 [184] |
25.2.5 | replace and replace_if have the requirement: OutputIterator<Iter, Iter::reference> Which implies they need to copy some values in the range the algorithm is iterating over. This is not however the case, the only thing that happens is const T&s might be copied over existing elements (hence the OutputIterator<Iter, const T&> | Remove OutputIterator<Iter, Iter::reference> from replace and replace_if | LWG | 1087 | |
UK 302 [187] |
25.3 ¶ 4 | the concept StrictWeakOrder covers the definition of a strict weak ordering, described in paragraph 4. | Remove 4, and mention StrictWeakOrder in paragraph 1. | editor |
REJECTED
There are at least a dozen places in the standard that refer to this defintion of stict weak ordering, including the concepts StrictWeakOrder and LessthanComparable. |
|
UK 303 [188] |
25.3 ¶ 6 | This paragraph just describes is_partitioned | 6 A sequence [start,finish) is partitioned with respect to an expression f(e) if is_partitioned(start, finish, e) != false | editor |
REJECTED
The paragraph describes something different. is_partitioned takes a predicate; "partitioned with respect to an expression" deals with an expression. While it's possible to wrap an expresion in a predicate, that would result in a circumlocution in the places where "partitioned with respect to an expression" is used. |
|
UK 304 [189] |
25.3.6 | The requires clauses of push_heap, pop_heap and make_heap are inconsistently formatted, dispite being identical | Format them identically. | editor |
REJECTED
Seems to be a comment about an earlier version. In N2800 the requires clauses of push_heap and pop_heap are similar, and are formatted the same. make_heap has no requires clause. |
|
UK 305 [335] |
25.3.7 ¶ 1, 9, 17 | The negative requirement on IsSameType is a hold-over from an earlier draught with a variadic template form of min/max algorith. It is no longer necessary. | Strike the !IsSameType<T, Compare> constraint on min/max/minmax algorithms | LWG | 1013 | ACCEPTED |
US 84 | 26 | Parts of the numerics chapter are not concept enabled. | LWG | |||
FR 35 | 26.3 [Complex numbers] | Instantiations of the class template complex<> have to be allowed for integral types, to reflect existing practice and ISO standards (LIA-III). | LWG | |||
UK 306 [113] |
26.4 | Random number library component cannot be used in constrained templates | Provide constraints for the random number library | LWG |
ACCEPTED
See paper N2836 |
|
JP 63 | 26.4.8.5.1 |
No
constructor of discrete_distribution that accepts
initializer_list.
discrete_distribution initialize distribution by a given range (iterators), but temporary variable of a container or an array is needed in the following case. int ar[] = {1, 2, 3}; discrete_distribution<> dist(ar, ar+3); Other libraries also accept initializer_list, so change discrete_distribution library to accept initializer_list too. |
Add
the following constructer.
discrete_distribution(initializer_list<result_type>); |
LWG | 874 |
ACCEPTED
See paper N2836 |
JP 64 | 26.5.2 | “valarray<T>& operator+= (initializer_list<T>);” is not defined. | Add valarray<T>& operator+= (initializer_list<T>); | LWG | ||
UK 307 [394] |
26.7 ¶ Footnote 288 | The footnote refers to TR1, which is not a defined term in this standard. Drop the reference to TR1, those templates are a regular part of the standard now and how they were introduced is no longer relevant. | Drop the reference to TR1. | editor | ACCEPTED | |
US 85 | 27 | The input/output chapter is not concept enabled. | LWG | |||
UK 308 [114] |
27 | iostreams library cannot be used from constrained templates | Provide constraints for the iostreams library, clause 27 | LWG | ||
JP 65 | 27.4.4 | Switch from “unspecified-bool-type” to “explicit operator bool() const”. | Replace "operator unspecified-bool-type() const;" with "explicit operator bool() const;" | LWG | ||
JP 66 | 27.4.4.3 ¶ 1st para | Switch from “unspecified-bool-type” to “explicit operator bool() const” | Replace "operator unspecified-bool-type() const;" with "explicit operator bool() const;" | LWG | ||
FR 36 | 27.6.1.2.2 [istream. formatted. arithmetic] ¶ 1, 2, and 3 |
iostate err = 0;
iostate is a bitmask type and so could be an enumeration. Probably using goodbit is the solution. |
editor | ACCEPTED | ||
FR 37 | 27.6.1.2.2 [istream. formatted. arithmetic] ¶ 3 |
else if (lval <
numeric_limits<int>::min()
|| numeric_limits<int>::max() < lval)) The parentheses aren't balanced. |
editor | ACCEPTED | ||
JP 67 | 27.7.1 | basic_stringbuf dose not use concept. |
Replace “class Allocator” to “Allocator
Alloc”.
Correct as follows. namespace std { template <class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_stringbuf : public basic_streambuf<charT,traits> { public: ... // 27.7.1.1 Constructors: explicit basic_stringbuf(ios_base::openmode which = ios_base::in | ios_base::out); explicit basic_stringbuf (const basic_string<charT,traits,Alloc>& str, ios_base::openmode which = ios_base::in | ios_base::out); basic_stringbuf(basic_stringbuf&& rhs); ... // 27.7.1.3 Get and set: basic_string<charT,traits,Alloc> str() const; void str(const basic_string<charT,traits,Alloc>& s); ... }; template <class charT, class traits, Allocator Alloc> void swap(basic_stringbuf<charT, traits, Alloc>& x, basic_stringbuf<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_stringbuf<charT, traits, Alloc>&& x, basic_stringbuf<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_stringbuf<charT, traits, Alloc>& x, basic_stringbuf<charT, traits, Alloc>&& y); } |
LWG | ||
JP 68 | 27.7.2 | basic_istringstream dose not use concept. |
Replace “class Allocator” to “Allocator
Alloc”.
Correct as follows. namespace std { template <class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_istringstream : public basic_istream<charT,traits> { public: typedef charT char_type; typedef typename traits::int_type int_type; typedef typename traits::pos_type pos_type; typedef typename traits::off_type off_type; typedef traits traits_type; typedef Alloc allocator_type; // 27.7.2.1 Constructors: explicit basic_istringstream(ios_base::openmode which = ios_base::in); explicit basic_istringstream( const basic_string<charT,traits,Alloc>& str, ios_base::openmode which = ios_base::in); basic_istringstream(basic_istringstream&& rhs); // 27.7.2.2 Assign and swap: basic_istringstream& operator=(basic_istringstream&& rhs); void swap(basic_istringstream&& rhs); // 27.7.2.3 Members: basic_stringbuf<charT,traits,Alloc>* rdbuf() const; basic_string<charT,traits,Alloc> str() const; void str(const basic_string<charT,traits,Alloc>& s); private: // basic_stringbuf<charT,traits,Alloc> sb; exposition only }; template <class charT, class traits, Allocator Alloc> void swap(basic_istringstream<charT, traits, Alloc>& x, basic_istringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_istringstream<charT, traits, Alloc>&& x, basic_istringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_istringstream<charT, traits, Alloc>& x, basic_istringstream<charT, traits, Alloc>&& y); } |
LWG | ||
JP 69 | 27.7.3 | basic_ostringstream dose not use concept. |
Replace “class Allocator” to “Allocator
Alloc”.
Correct as follows. namespace std { template <class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_ostringstream : public basic_ostream<charT,traits> { public: // types: typedef charT char_type; typedef typename traits::int_type int_type; typedef typename traits::pos_type pos_type; typedef typename traits::off_type off_type; typedef traits traits_type; typedef Alloc allocator_type; // 27.7.3.1 Constructors/destructor: explicit basic_ostringstream(ios_base::openmode which = ios_base::out); explicit basic_ostringstream( const basic_string<charT,traits,Alloc>& str, ios_base::openmode which = ios_base::out); basic_ostringstream(basic_ostringstream&& rhs); // 27.7.3.2 Assign/swap: basic_ostringstream& operator=(basic_ostringstream&& rhs); void swap(basic_ostringstream&& rhs); // 27.7.3.3 Members: basic_stringbuf<charT,traits,Alloc>* rdbuf() const; basic_string<charT,traits,Alloc> str() const; void str(const basic_string<charT,traits,Alloc>& s); private: // basic_stringbuf<charT,traits,Alloc> sb; exposition only }; template <class charT, class traits, Allocator Alloc> void swap(basic_ostringstream<charT, traits, Alloc>& x, basic_ostringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_ostringstream<charT, traits, Alloc>&& x, basic_ostringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_ostringstream<charT, traits, Alloc>& x, basic_ostringstream<charT, traits, Alloc>&& y); } |
LWG | ||
JP 71 | 27.7.3 |
Typo.
"template" is missing in "class basic_ostringstream" of the title of the chapter. |
Correct as follows.
27.7.3 Class basic_ostringstream should be 27.7.3 Class template basic_ostringstream |
editor | ACCEPTED | |
JP 72 | 27.7.4 | basic_stringstream dose not use concept. |
Replace "class Allocator" to "Allocator Alloc".
Correct as follows. namespace std { template <class charT, class traits = char_traits<charT>, Allocator Alloc = allocator<charT> > class basic_stringstream : public basic_iostream<charT,traits> { public: // types: typedef charT char_type; typedef typename traits::int_type int_type; typedef typename traits::pos_type pos_type; typedef typename traits::off_type off_type; typedef traits traits_type; typedef Alloc allocator_type; // constructors/destructor explicit basic_stringstream( ios_base::openmode which = ios_base::out|ios_base::in); explicit basic_stringstream( const basic_string<charT,traits,Alloc>& str, ios_base::openmode which = ios_base::out|ios_base::in); basic_stringstream(basic_stringstream&& rhs); // 27.7.5.1 Assign/swap: void swap(basic_stringstream&& rhs); // Members: basic_stringbuf<charT,traits,Alloc>* rdbuf() const; basic_string<charT,traits,Alloc> str() const; void str(const basic_string<charT,traits,Alloc>& str); private: // basic_stringbuf<charT, traits> sb; exposition only }; template <class charT, class traits, Allocator Alloc> void swap(basic_stringstream<charT, traits, Alloc>& x, basic_stringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_stringstream<charT, traits, Alloc>&& x, basic_stringstream<charT, traits, Alloc>& y); template <class charT, class traits, Allocator Alloc> void swap(basic_stringstream<charT, traits, Alloc>& x, basic_stringstream<charT, traits, Alloc>&& y); } |
LWG | ||
JP 73 | 27.8.1.14 | It is a problem from C++98, fstream cannot appoint a filename of wide character string(const wchar_t and const wstring&). | Add interface corresponding to wchar_t, char16_t and char32_t. | LWG | ||
US 86 | 28 | The regular expressions chapter is not concept enabled. | LWG | |||
UK 309 [115] |
28 | Regular expressions cannot be used in constrained templates | Provide constraints for the regular expression library, clause 28 | LWG | ||
UK 310 [281] |
28 | The regex chapter uses iterators in the old pre-concept style, it should be changed to use concepts instead. | Use concepts for iterator template arguments throughout. | LWG | ||
UK 314 [343] |
28.4 | The swap overloads for regex classes are only supplied for l-value references. Other sections of the library (eg 21 strings or 23 containers) provide two extra overloads taking an r-value reference as the first and second argument respectively. | Add the missing overloads to 28.4 and the corresponding later sections in 28 for each swap function. We want to accept AMs paper which proposes a single overload with two r-value references | LWG | ||
UK 315 [344] |
28.4 ¶ p6 | 6 Effects: string_type str(first, last); return use_facet<collate<charT> >( getloc()).transform(&*str.begin(), &*str.end()); Is it legal to dereference str.end() ? | Reword to effect clause to use legal iterator dereferences | editor | ACCEPTED | |
UK 316 [345] |
28.4 ff | The constructors for regex classes do not have an r-value overload. | Add the missing r-value constructors to regex classes. | LWG | 723 | ACCEPTED |
UK 317 [278] |
28.8 | basic_string has both a constructor and an assignment operator that accepts an initializer list, basic_regex should have the same. | In the basic_regex synopsis, after: basic_regex& operator=(const charT* ptr); add: basic_regex& operator=(initializer_list<charT> il); And after paragraph 20 add: basic_regex& operator=(initializer_list<charT> il); Effects: returns assign(il.begin(), il.end()); | LWG | 1014 | |
JP 74 | 28.8 | “basic_regx & operator= (initializer_list<T>);” is not defined. | Add basic_regx & operator= (initializer_list<T>); | LWG | 1014 | |
UK 318 [279] |
28.8.2 ¶ para 22 | Constructor definition should appear with the other constructors not after assignment ops. | Move para 22 to just after para 17. | editor | ACCEPTED | |
UK 319 [280] |
28.12.2 | It was always expected that regex_token_iterator would be constructible from an array literal: indeed ideally this is the prefered method of initializing such an object. However, the best we could do in C++0x was: template <std::size_t N> regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, const int (&submatches)[N], regex_constants::match_flag_type m = regex_constants::match_default); Now that we have initializer_lists we should use them to remove this particular wart. | To the synopsis for regex_token_iterator, after template <std::size_t N> regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, const int (&submatches)[N], regex_constants::match_flag_type m = regex_constants::match_default); add regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, initializer_list<int> submatches, regex_constants::match_flag_type m = regex_constants::match_default); In 28.12.2.1 add the declaration: regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, initializer_list<int> submatches, regex_constants::match_flag_type m = regex_constants::match_default); And to the end of para 3 add: The forth constructor initializes the member subs to hold a copy of the sequence of integer values in the range [submatches.begin(), submatches.end()). | LWG | 909 | |
US 87 | 29 | The atomics chapter is not concept enabled. The adopted paper, N2427, did have those concepts. | LWG | |||
UK 311 [116] |
29 | Atomic types cannot be used generically in a constrained template | Provide constraints for the atomics library, clause 29 | LWG | ||
UK 312 [157] |
29 | The contents of the <stdatomic.h> header are not listed anywhere, and <cstdatomic> is listed as a C99 header in chapter 17. If we intend to use these for compatibility with a future C standard, we should not use them now. | Remove <cstdatomic> from the C99 headers in table 14. Add a new header <atomic> to the headers in table 13. Update chapter 29 to remove reference to <stdatomic.h> and replace the use of <cstdatomic> with <atomic>. If and when WG14 adds atomic operations to C we can add corresponding headers to table 14 with a TR. | LWG | ||
JP 75 | 29 | A definition of enum or struct is the style of C using typedef. |
Change to a style of C++.
Correct as follows. 29.1 namespace std { typedef enum memory_order { memory_order_relaxed, memory_order_consume, memory_order_acquire, memory_order_release, memory_order_acq_rel, memory_order_seq_cst } memory_order; } should be namespace std { enum memory_order { memory_order_relaxed, memory_order_consume, memory_order_acquire, memory_order_release, memory_order_acq_rel, memory_order_seq_cst }; } 29.3.1 namespace std { typedef struct atomic_bool { ... } atomic_bool; } should be namespace std { struct atomic_bool { ... }; } namespace std { typedef struct atomic_itype { ... } atomic_itype; } should be namespace std { struct atomic_itype { ... }; } 29.3.2 namespace std { typedef struct atomic_address { ... } atomic_address; } should be namespace std { struct atomic_address { ... }; } 29.5 namespace std { typedef struct atomic_flag { ... } atomic_flag; } should be namespace std { struct atomic_flag { ... }; } |
LWG |
REJECTED
NAD: We believe the current syntax is most appropriate for an interface that is intended to be highly compatible with C. |
|
UK 313 [220] |
29.1 | seq_cst fences don't necessarily guarantee ordering http://home.twcny.rr.com/hinnant/cpp_extensions/issues_preview/lwg-active.html#926 | Add a new paragraph after 29.1 [atomics.order]p5 that says For atomic operations A and B on an atomic object M, where A and B modify M, if there are memory_order_seq_cst fences X and Y such that A is sequenced before X, Y is sequenced before B, and X precedes Y in S, then B occurs later than A in the modifiction order of M. | LWG | 926 | |
US 88 | 29.2 | The "lockfree" facilities do not tell the programmer enough. | Expand the "lockfree" facilities. See the attached paper "Issues with the C++ Standard" under Chapter 29, "atomics.lockfree doesn't tell the programmer enough" | LWG | ||
US 89 | 29.3.1 ¶ Table 122 | The types in the table "Atomics for standard typedef types" should be typedefs, not classes. These semantics are necessary for compatibility with C. | Change the classes to typedefs. | LWG | 937 | |
US 90 | 29.4 | Are atomic functions allowed to have non-volatile overloads? | Allow non-volatile overloads. See the attached paper "Issues with the C++ Standard, under Chapter 29, "Are atomic functions allowed to have non-volatile overloads?" | LWG | 908 | |
US 91 | 29.4 | Whether or not a failed compare_exchange is a RMW operation (as used in 1.10 [intro.multithread]) is unclear. | Make failing compare_exchange operations not be RMW. See the attached paper under "atomic RMW status of failed compare_exchange" | LWG | 1043 | ACCEPTED |
US 92 | 29.4 | The effect of memory_order_consume with atomic RMW operations is unclear. | Follow the lead of fences [atomics.fences], and promote memory_order_consume to memory_order_acquire with RMW operations. | LWG |
REJECTED
NAD. We can not see the issue being suggested by the comment. |
|
JP 76 | 30 |
A
description for "Throws: Nothing." are not unified.
At the part without throw, "Throws: Nothing." should be described. |
Add
"Throws: Nothing." to the following.
30.2.1.6 , 1st paragraph 30.3.3.1 , 4th paragraph 30.3.3.2.1 , 6th paragraph 30.4.1 , 7th and 8th paragraph 30.4.2 , 6th, 7th,19th,21th and 25th paragraph |
LWG | 1089 | |
US 93 | 30 | The thread chapter is not concept enabled. | LWG | |||
UK 320 [117] |
30 | Threads library cannot be used in constrained templates | Provide constraints for the threads library, clause 30 | LWG | ||
UK 321 [172] |
30 | Throughout this clause, the term Requires: is used to introduce compile time requirements, which we expect to be replaced with concepts and requires in code. Run-time preconditions are introduced with the term "Preconditions:" which is not a defined part of the library documentation structure (17.5.2.4). However, this is exactly the direction that BSI would like to see the organisation move, replacing Requires: clauses with Preconditions: clasues throught the library. See comment against clause 17 for more details. | Decument Preconditions: paragraphs in 17.5.2.4, and use consistently through rest of the library. | editor | ||
US 94 | 30.1.2 ¶ 1 | The first sentence of para 1 suggests that no other library function is permitted to call operating system or low level APIs. | Rewrite para 1 as: “ Some functions described in this Clause are specified to throw exceptions of type system_error (19.4.5 ). Such exceptions shall be thrown if a call to an operating system or underlying API results in an error that prevents the library function from satisfying its postconditions or from returning a meaningful value.” | editor | ACCEPTED | |
US 95 | 30.1.3 ¶ 1 | “native_handle_type” is a typedef, not a class member. |
Several classes
described in this Clause have a member native_handle (of
type native_handle_type) . The
presence of this member and its semantics is implementation defined. [ Note: This member allows implementations to provide access to implementation details. The name of the member and the type are specified to facilitate portable compile-time detection. Actual use of this member or the corresponding type is inherently non-portable. —end note ] |
LWG |
REJECTED
Not a defect. native_handle_type is a typedef, but it is also a member of the classes in question. |
|
US 96 | 30.1.4 ¶ 2 | There is no definition here for monotonic clock. | Implementations should use a monotonic clock to measure time for these functions. A monotonic clock measures real time, but cannot be set, and is guaranteed to have no negative clock jumps. | LWG | REJECTED | |
UK 322 [168] |
30.1.4 ¶ 2 | Not all systms can provide a monotonic clock. How are they expected to treat a _for function? | Add at least a note explaining the intent for systems that do not support a monotonic clock. | LWG | ||
UK 323 [171] |
30.2.1 ¶ 1 | The presence of a non-explicit variadic template constructor alongside an explicit single-argument constructor can lead to behaviour that is not intended: the variadic constructor will be selected for implicit conversions, defeating the purpose of the explicit single-argument constructor. Additionally the single-argument constructor is redundant; the variadic constructor can provide identical functionality with one *fewer* copies if the supplied func is an lvalue. | Mark constructor template <class F, class ...Args> thread(F&& f, Args&&... args); as explicit and remove the single-argument constructor. | LWG | 929 | ACCEPTED |
UK 324 [169] |
30.2.1.1 | thread::id objects should be as useable in hashing containers as they are in ordered associative containers. | Add thread::id support for std::hash | LWG | 889 |
REJECTED
Not a defect. See [unord.hash], where std::thread:id is already listed as a required specialization for std::hash(). |
JP 77 | 30.2.1.2 | "CopyConstructible" and "MoveConstructible" in "Requires: F and each Ti in Args shall be CopyConstructible if an lvalue and otherwise MoveConstructible." are reflected by interface. | Add a concept for constructor of thread. | LWG | ||
JP 78 | 30.2.1.2 ¶ 4th para, 1st line | In "F and each Ti in Args", 'Ti' is not clear. | Replace "Ti" with "args" | editor | ACCEPTED | |
US 97 | 30.2.1.3 ¶ 1 | detach-on-destruction may result in “escaped” threads accessing objects with bounded lifetime after the end of their lifetime. | See document WG21 N2802=08-0312 written by Hans Boehm. | LWG | See paper N2802 | |
US 98 | 30.2.1.3, 30.2.1.4 | The current defined behavior for the std::thread destructor is to detach the thread. Unfortunately, this behavior exposes programmers to tricky, hard-to-diagnose, undefined behavior. | Change destruction behavior to undefined behavior, with a note strongly encouraging termination. See the attached paper "Issues with the C++ Standard" under Chapter 30, "Implicit thread detach is harmful". | LWG | ||
UK 325 [174] |
30.3.3 ¶ 2 | We believe constexpr literal values should be a more natural expression of empty tag types than extern objects as it should improve the compilers ability to optimize the empty object away completely. | Replace the extern declarations: extern const defer_lock_t defer_lock; extern const try_to_lock_t try_to_lock; extern const adopt_lock_t adopt_lock; with constexpr values constexpr defer_lock_t defer_lock{}; constexpr try_to_lock_t try_to_lock{}; constexpr adopt_lock_t adopt_lock{}; | LWG | 1044 | |
UK 326 [176] |
30.3.3.2.1 ¶ 7 | The precondition that the mutex is not owned by this thread offers introduces the risk of un-necessary undefined behaviour into the program. The only time it matters whether the current thread owns the mutex is in the lock operation, and that will happen subsequent to construction in this case. The lock operation has the identical pre-condition, so there is nothing gained by asserting that precondition earlier and denying the program the right to get into a valid state before calling lock. | Strike 30.3.3.2.1p7 | LWG | 1045 | ACCEPTED |
UK 327 [179] |
30.3.3.2.2 ¶ 4, 9, 14, 19 | Not clear what the specification for error condition resource_deadlock_would_occur means. It is perfectly possible for this thread to own the mutex without setting owns to true on this specific lock object. It is also possible for lock operations to succeed even if the thread does own the mutex, if the mutex is recursive. Likewise, if the mutex is not recursive and the mutex has been locked externally, it is not always possible to know that this error condition should be raised, depending on the host operating system facilities. It is possible that 'i.e.' was supposed to be 'e.g.' and that suggests that recursive locks are not allowed. That makes sense, as the exposition-only member owns is boolean and not a integer to count recursive locks. | Add a precondition !owns. Change the 'i.e.' in the error condition to be 'e.g.' to allow for this condition to propogate deadlock detection by the host OS. | LWG | ||
UK 328 [182] |
30.3.3.2.2 ¶ 20 | There is a missing precondition that owns is true, or an if(owns) test is missing from the effect clause | Add a precondition that owns == true. Add an error condition to detect a violation, rather than yield undefined behaviour. | LWG | ||
UK 329 [203] |
30.5 | future, promise and packaged_task provide a framework for creating future values, but a simple function to tie all three components together is missing. Note that we only need a *simple* facility for C++0x. Advanced thread pools are to be left for TR2. | Provide a simple function along the lines of: template< typename F, typename ... Args > requires Callable< F, Args... > future< Callable::result_type > async( F&& f, Args && ... ); Semantics are similar to creating a thread object with a packaged_task invoking f with forward<Args>(args...) but details are left unspecified to allow different scheduling and thread spawning implementations. It is unspecified whether a task submitted to async is run on its own thread or a thread previously used for another async task. If a call to async succeeds, it shall be safe to wait for it from any thread. The state of thread_local variables shall be preserved during async calls. No two incomplete async tasks shall see the same value of this_thread::get_id(). [Note: this effectively forces new tasks to be run on a new thread, or a fixed-size pool with no queue. If the library is unable to spawn a new thread or there are no free worker threads then the async call should fail.] | LWG | 1046 | |
UK 330 [424] |
30.5.1 | 30.5.1 (and then 30.5.7) refer to a specialisation of constructible_with_allocator_prefix<> However this trait is not in the CD, so references to it should be removed. | Remove the reference to constructible_with_allocator_prefix in 30.5.1 Remove paragraph 30.5.7 | editor | ACCEPTED | |
JP 79 | 30.5.1 | The concept of UsesAllocator and Allocator should be used. |
Correct as follows.
template <class R, class Alloc> struct uses_allocator<promise<R>, Alloc>; template <class R> struct constructible_with_allocator_prefix<promise<R> >; should be template<class R, Allocator Alloc> concept_map UsesAllocator<promise<R>, Alloc>; |
LWG | ||
UK 331 [192] |
30.5.3 | Not clear what it means for a public constructor to be 'exposition only'. If the intent is purely to support the library calling this constructor then it can be made private and accessed through friendship. Otherwise it should be documented for public consumption. | Declare the constructor as private with a note about intended friendship, or remove the exposition-only comment and document the semantics. | LWG | ||
UK 332 [194] |
30.5.4 | It is not clear without reference to the original proposal how to use a future. In particular, the only way for the user to construct a future is via the promise API which is documented after the presentation of future. Most library clauses feature a small description of their components and intended use, it would be most useful in this case. | Provide a small introductory paragraph, docuenting intended purpose of the future class template and the way futures can only be created via the promise API. | editor | ||
UK 333 [195] |
30.5.4 ¶ 5 | We expect the complicated 3-signature specifcation for future::get() to be simplified to a single signature with a requires clause by the application of concepts. | Requires fully baked concepts for clause 30 | LWG | ||
UK 334 [196] |
30.5.4 ¶ 5 | Behaviour of get() is undefined if calling get() while not is_ready(). The intent is that get() is a blocking call, and will wait for the future to become ready. | Effects: If is_ready() would return false, block on the asynchronous result associated with *this. | LWG | 1047 | ACCEPTED |
UK 335 [436] |
30.5.4 | std::unique_future is MoveConstructible, so you can transfer the association with an asynchronous result from one instance to another. However, there is no way to determine whether or not an instance has been moved from, and therefore whether or not it is safe to wait for it. std::promise<int> p; std::unique_future<int> uf(p.get_future()); std::unique_future<int> uf2(std::move(uf)); uf.wait(); // oops, uf has no result to wait for. | Add a "waitable()" function to unique_future (and shared_future) akin to std::thread::joinable(), which returns true if there is an associated result to wait for (whether or not it is ready). Then we can say: if(uf.waitable()) uf.wait(); | LWG | 1048 | |
UK 336 [437] |
30.5.4 | It is possible to transfer ownership of the asynchronous result from one unique_future instance to another via the move-constructor. However, it is not possible to transfer it back, and nor is it possible to create a default-constructed unique_future instance to use as a later move target. This unduly limits the use of unique_future in code. Also, the lack of a move-assignment operator restricts the use of unique_future in containers such as std::vector - vector::insert requires move-assignable for example. | Add a default constructor with the semantics that it creates a unique_future with no associated asynchronous result. Add a move-assignment operator which transfers ownership. | LWG | ||
JP 80 | 30.5.4, 30.5.5 |
Typo, duplicated ">"
"class Period>>" |
Remove one | editor | ACCEPTED | |
UK 337 [198] |
30.5.5 | shared_future should support an efficient move constructor that can avoid unnecessary manipulation of a reference count, much like shared_ptr | Add a move constructor | LWG | ||
UK 338 [223] |
30.5.5 | shared_future is currently CopyConstructible, but not CopyAssignable. This is inconsistent with shared_ptr, and will surprise users. Users will then write work-arounds to provide this behaviour. We should provide it simply and efficiently as part of shared_future. Note that since the shared_future member functions for accessing the state are all declared const, the original usage of an immutable shared_future value that can be freely copied by multiple threads can be retained by declaring such an instance as "const shared_future". | Remove "=delete" from the copy-assignment operator of shared_future. Add a move-constructor shared_future(shared_future&& rhs), and a move-assignment operator shared_future& operator=(shared_future&& rhs). The postcondition for the copy-assignment operator is that *this has the same associated state as rhs. The postcondition for the move-constructor and move assignment is that *this has the same associated as rhs had before the constructor/assignment call and that rhs has no associated state. | LWG | ||
UK 339 [200] |
30.5.6 ¶ 6, 7 | Move assignment is goiing in the wrong direction, assigning from *this to the passed rvalue, and then returning a reference to an unusable *this | Strike 6. 7 Postcondition: associated state of *this is the same as the associated state of rhs before the call. rhs has no associated state. | LWG | 1049 | ACCEPTED |
UK 340 [201] |
30.5.6 ¶ 11, 12, 13 | There is an implied postcondition that the state of the promise is transferred into the future leaving the promise with no associated state. It should be spelled out. | Postcondition: *this has no associated state. | LWG | 1050 | ACCEPTED |
UK 341 [122] |
30.5.6 | promise::swap accepts its parameter by lvalue reference. This is inconsistent with other types that provide a swap member function, where those swap functions accept an rvalue reference | Change promise::swap to take an rvalue reference. | LWG | ||
UK 342 [123] |
30.5.6 | std::promise is missing a non-member overload of swap. This is inconsistent with other types that provide a swap member function | Add a non-member overload void swap(promise&& x,promise&& y){ x.swap(y); } | LWG | 1088 | |
UK 343 [439] |
30.5.6 ¶ 3 | The move constructor of a std::promise object does not need to allocate any memory, so the move-construct-with-allocator overload of the constructor is superfluous. | Remove the constructor with the signature template <class Allocator> promise(allocator_arg_t, const Allocator& a, promise& rhs); | LWG | ||
JP 81 | 30.5.8 | There are not requirements for F and a concept of Allocator dose not use. |
Correct as follows.
template <class F> explicit packaged_task(F f); template <class F, class Allocator> explicit packaged_task(allocator_arg_t, const Allocator& a, F f); template <class F> explicit packaged_task(F&& f); template <class F, class Allocator> explicit packaged_task(allocator_arg_t, const Allocator& a, F&& f); should be template <class F> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> explicit packaged_task(F f); template <class F, Allocator Alloc> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> explicit packaged_task(allocator_arg_t, const Alloc& a, F f); template <class F> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> explicit packaged_task(F&& f); template <class F, Allocator Alloc> requires CopyConstructible<F> && Callable<F, ArgTypes...> && Convertible<Callable<F, ArgTypes...>::result_type, R> explicit packaged_task(allocator_arg_t, const Alloc& a, F&& f); |
LWG | ||
DE 23 | Annex B ¶ p2 | DE-23 Recursive use of constexpr functions appears to be permitted. Since such functions may be required to be evaluated at compile-time, Annex B "implementation quantities" should specify a maximum depth of recursion. | In Annex B, specify a recursion depth of 256 or a larger value. | CWG | 699 | |
DE 24 | Annex B ¶ p2 | DE-24 The number of placeholders for "bind" is implementation-defined in 20.7.12.1.4, but no minimum is suggested in Annex B. | Add a miminum of 10 placeholders to Annex B. | LWG | 922 | |
DE 25 | Annex B ¶ p2 | DE-25 Specifying a minimum of 17 recursively nested template instantiations is too small for practical purposes. The limit is too high to effectively limit compiler resource usage, see http://ubiety.uwaterloo.ca/~tveldhui/papers/2003/turing.pdf . The conclusion is that the metric "number of recursively nested template instantiations" is inapposite. | Remove the bullet "Recursively nested template instantiations [17]". | CWG | 831 | |
FR 38 | C.2 [diffs.library] ¶ 1 |
What is ISO/IEC 1990:9899/DAM
1? My guess is that's a typo for ISO/IEC
9899/Amd.1:1995 which I'd have expected to be referenced here (the tables make reference to things which were introduced by Amd.1). |
One need probably a reference
to the document which introduce char16_t and
char32_t in C (ISO/IEC TR 19769:2004?). |
LWG | ACCEPTED | |
UK 344 [136] |
Appendix D | It is desirable to allow some mechanism to support phasing out of deprecated features in the future. Allowing compilers to implement a mode where deprecated features are not available is a good first step. | Add to the definition of deprecated features permission for compilers to maintain a conditionally supported mode where deprecated features can be disabled, so long as they also default to a mode where all deprecated features are supported. | CWG |
REJECTED
Compiler switches and modes are quality-of-implementation issues and outside the scope of the Standard. |
|
FR 39 | Index | Some definitions seem not indexed (such as trivially copyable or sequenced before), indexing them would be useful (and marking specially the page -- say bold or italic -- which reference to the definition would increase the usefulness; having a separate index of all definitions is something which could also be considered). | editor |