From peren!uunet!aussie.com!rex Sun Feb 23 13:38 PST 1997 Return-Path: Received: by cobra. (5.x/SMI-SVR4) id AA14102; Sun, 23 Feb 1997 13:38:37 -0800 Received: from peren by cobra; Sun, 23 Feb 1997 13:38 PST Received: by (SMI-8.6/SMI-SVR4) id NAA20601; Sun, 23 Feb 1997 13:15:08 -0800 Received: from uunet by peren; Sun, 23 Feb 1997 13:15 PST Received: from alterdial.UU.NET by relay1.UU.NET with ESMTP (peer crosschecked as: alterdial.UU.NET [192.48.96.22]) id QQceej17748; Sun, 23 Feb 1997 15:28:18 -0500 (EST) Received: from 205.230.244.7 by alterdial.UU.NET with SMTP (peer crosschecked as: pool007.Max1.FFX1.VA.DYNIP.ALTER.NET [205.230.244.7]) id QQceej11789; Sun, 23 Feb 1997 15:28:13 -0500 (EST) From: peren!uunet!aussie.com!rex >Received: by (SMI-8.6/SMI-SVR4) id NAA20601; Sun, 23 Feb 1997 13:15:08 -0800 >Received: from alterdial.UU.NET by relay1.UU.NET with ESMTP (peer crosschecked as: alterdial.UU.NET [192.48.96.22]) id QQceej17748; Sun, 23 Feb 1997 15:28:18 -0500 (EST) Date: Sun, 23 Feb 1997 15:28:13 -0500 (EST) Message-Id: Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Subject: C9x draft 8 comments from Kona meeting To: uunet!peren.com!jb, uunet!farance.com!frank, uunet!Eng.Sun.COM!dew X-Mailer: SPRY Mail Version: 04.00.06.17 Content-Type: text/plain Content-Length: 15332 Status: RO Here's the final version of the comments I gathered during the draft walk-through. Except for spell checking and formatting, this is the same as the version I gave several of you at the meeting. John, please include this in the next mailing. Rex C9x Draft #8 Issues WG14 N662 J11/97-025 1 C9x Draft #8 Issues =================== Rex Jaeschke These notes were recorded at the Kona, Hawaii meeting during the walk-through of C9x Draft 8. General issues 1) Make sure all comments in code are set in italics. 2) Footnote superscripts should follow any punctuation, not precede it. 3) Force footnote text to be on same page as references to it. 4) Check use of Example vs Examples subhead. ----------------------------------------------------------- Specific page references pp. Introduction [Seymour] Introduction: [6] clause 7 becomes 6, [7] clause 8 becomes 7. (done) pp. 2 [Nestor] 3.6 [1] Definition of Constraint Change start of sentence to be "Restrictions, both syntactic and semantic, by which the exposition ..." pp. 2 [Martin] Add ISO 2382 and IEC559 to normative references list. pp. 3 [Seebach] 3.15 [1] Exclude padding from the bytes being discussed here. Seymour will write a paper for London. pp. 11 [MacDonald] pp -3, bad break in a + b expression. In [...,...] put space after comma. pp. 18 [Gwyn] C9x Draft #8 Issues WG14 N662 J11/97-025 2 Values in limits.h should be aligned consistently. pp. 19 [MacDonald/Seymour] [3] spacing issue at front, ",and" should be ", and" and "and" should be in roman. [4] right-justify digit column footnote 12: For example , if [FLT_EVAL_METHOD] is 1 ... (done) footnote 13: fesetround needs font change pp. 20 [MacDonald] [5] misc typesetting issues. pp. 22 [MacDonald] 2. change ANSI/IEEE ref to IEC559. pp. 26 [MacDonald] [2] "complex" needs font change [2] Change last sentence to "... but not otherwise; if the header is included, then any use of the token *complex* prior to the first inclusion results in undefined behavior." pp. 30 [Seymour] [3] Change "." to "," on line 2. pp. 32 [MacDonald] [16] and [17] need overhauling. Walls/Thomas papers address this. pp. 33 [Seymour] footnote 23: remove parenthetical remark and move footnote to a definition in clause 3. [18] "type domain" to be in italics. pp. 36 [Tydeman] hex-fractional-constants formatting issue. Addressed by Walls paper. C9x Draft #8 Issues WG14 N662 J11/97-025 3 pp. 37 [Martin] [5] Replace with the following text: The implementation emits a diagnostic if a hexadecimal constant cannot be represented exactly in its evaluation format. Emitting such a diagnostic should not prevent a valid program from being correctly translated. pp. 38 [Seymour] [1] syntax needs long long fixed (done). pp. 41 [Seymour] [12] 3. Change this para to the following: Even if eight bits are used for objects that have type char, the construction '\x123' specifies an integer character constant containing only one character, since a hexadecimal escape sequence is terminated only by a nonhexadecimal character. To specify an integer character constant containing the two characters whose values are '\x12' and '3', the construction '\0223' may be used, since an octal escape sequence is terminated after three octal digits. (The value of this two-character integer character constant is implementation-defined also.) pp. 43 [Seymour] [7] need single quotes around \x12. pp. 46 [Nestor] [2] add p, P, and +/-. pp. 48 [Martin/Thomas] 6.2.1.3 Add a new sentence to para 2: "If the value being converted is outside the range of values that can be represented, the behavior is undefined." pp. 49 [MacDonald] [1] typesetting issues. longdouble. (done) pp. 57 [Seymour] [6] line 2, add a "," after "argument". pp. 61 [MacDonald] footnote: remove paragraph number (done) C9x Draft #8 Issues WG14 N662 J11/97-025 4 pp. 66 [MacDonald] footnote: remove paragraph number. (done) pp. 68 [MacDonald] [5] line -2, change "higher" to "greater". pp. 69 [MacDonald/Thomas] [3] last sentence. Make the following insertion: "... Any two values of arithmetic types [from different type-domains] are equal ..." pp. 73 [Tydeman] [8] Make table more obviously a table by adding vertical lines pp. 87 [Plum] Paragraph numbers in examples not consistently numbered/formatted. pp. 88 [Seymour] [2] Replace "same specifiers list of qualifier list" with "same specifier-qualifier-list". Change "shall be" to "is". Move [2] after the first sentence in semantics. pp. 89 [Nestor] [9] Move last sentence outside of parenthetical remark. pp. 91 [Martin] 3. Make arrays d and e static so they are initialized. pp. 100 [Seymour] [1] Change "designation-opt" to "designation" with opt as a subscript. pp. 102 [Seymour] [12] line 2, make "subobject" plural. pp. 104 [Martin] Change the type of all float arrays in examples to int to match the type of the values in the init list. Simply a C9x Draft #8 Issues WG14 N662 J11/97-025 5 style issue. pp. 117 [Seymour] [3] Add "type" to the end. pp. 121 [MacDonald] pragma operator is missing from the syntax. MacDonald will write a paper for London. pp. 126 [Tydeman] [2, 3] Italicize "lparen". [3] should be changed according to the direction given in TC2, to be like [2]. pp. 129 [Tydeman] [4] Replace ## with hash_hash. pp. 132 [Seebach] 6. Change FTN prefix to something more indicative of function, such as FUN; editor to chose. pp. 134 [Martin] __STDC_VERSION__, make discussion of 199409L a footnote for historical reference purposes. Make new value of this macro be 199901L. pp. 135 [Seymour] [1] make \ be \\. Also "user" should be "unary". "replacing" is misspelled. pp. 137 [Meyers] 7.1.1 [3] "setlocale typeset incorrectly as setlocal with superscript of a degree or some such symbol. pp. 138 [Tydeman] [2] Add complex.h and fenv.h to this list. pp. 143 [Seymour] Strike last example now that we don't allow implicit function declarations. pp. 152 [Martin] C9x Draft #8 Issues WG14 N662 J11/97-025 6 7.4.8 [1] Make examples consistent with 5.2.4.2.1 limits.h w.r.t. to specifying values for the macros. The numerical values should be in roman. pp. 153 [Martin] line 2, change 65536U to 65535U. pp. 154 [Walls] 7.4.10 [1] Replace "of of" with "of". pp. 155 [Seebach] Gwyn will write a paper for London to clean up this whole section. pp. 163 [Seymour] int_frac_digits: change "a" to "an". pp. 165 [Seymour] In the table at top, the entry for Switzerland has formatting problems. The second column runs into the third. Split between C and CHF. pp. 166 [Tydeman] [3] footnote 123 reference has too much space. pp. 168 [Seymour] line 2, spacing issue. (Walls paper) pp. 176 [Seymour] 7.6.5.6 [2] Fix/remove HU typesetting command. pp. 178 [Seymour] 7.6.6.7 [2] Change "2x" to "2^x". Remove space in front of footnote reference. 7.6.6.8 [2] Change "ex" to "e^x". Make "e" italic in all base-e. pp. 179 [Seymour] line 2, [3] Change "ex" to "e^x". C9x Draft #8 Issues WG14 N662 J11/97-025 7 7.6.6.9 [2] Replace em dash with a minus sign. (Two occurrences). pp. 180 [Seymour] 7.6.6.12 Fix ...RADIXn in two places. pp. 182 [Seymour] Last line, change "loge" to "log subscript e". Change g to gamma symbol. Also in the previous paragraph. pp. 184 [Seymour] 7.6.9.6 [3] longint becomes long int. (Walls paper) pp. 187 [Seymour] 7.6.10.3 [2] line 2, formatting issue. pp. 190 [Seymour] 7.6.12.2 [3] footnote 136: move to 7.6.12.3 [3]. 7.6.13 [1] Add em dashes around phrase that currently has what looks like a hyphen. pp. 191 [Martin] 7.6.13.2 [2] Bad spacing. pp. 194 [Seymour] 7.7 [1] line -1 longdouble becomes long double, same in footnote 138. pp. 195 [Seymour] 7.7.2.1 [1, 2] Annex numbers are wrong. line 4, delete sentence that begins "When inside a compound statement ..." pp. 196 [Seymour] 7.7.2.2 [3] End of sentence, formatting issue. (done) pp. 201 [Seymour] 7.7.2.18 [2, 3] "xy" becomes "x^y". pp. 214 [Seymour] C9x Draft #8 Issues WG14 N662 J11/97-025 8 2nd bullet, remove the name of the standard and say defined that this is defined in this subclause just like the previous bullets. pp. 234 [Martin] footnote 165: The list of functions is incomplete so after the first two or three, add ", etc." pp. 254 [Seymour] [3] after 2nd bullet, text missing (in Walls paper). 0x needs a font change. pp. 266 [Martin] 7.12.6.4 [1] argument should have type long long int. (done) pp. 292 [Tydeman] 7.16.2.1 [2] line 2, L' ' should be L' '. pp. 302 [Martin/Benito] In subclause 7.17.2.1 , page 302, # conversion specification, second sentence change: For o conversion, it increases the precision to force the first digit of the results to be a zero, if necessary. To: For o conversion, it increases the precision, if and only if necessary, to force the first digit of the result to be a zero. In subclause 7.17.2.2, page 306, paragraph [9], second sentence change: An input item is defined as the longest sequence of input wide characters, not exceeding any specified field width, which is, or is a prefix of, a matching sequence. To: An input item is defined as the longest sequence of input wide characters which does not exceed any specified field width and which is, or if a prefix of, a matching input sequence. C9x Draft #8 Issues WG14 N662 J11/97-025 9 In subclause 7.17.2.2, page 309, New paragraph after paragraph [18]: If conversion terminates on a conflicting input character, the offending input character is left unread in the input stream.[footnote] [footnote] Same footnote as footnote 165, page 234 - as changed in another edit. pp. 323 [Martin] Change "to to" to "to". pp. 369 [Seymour] [1] The long long macros are missing. pp. 371 [Martin] F.2 [1] bullet 3, formatting problems (in Walls paper) pp. 410 [Seymour] I.2 2nd bullet. Qualify that this should only apply to a hosted environment. pp. 412 [Tydeman] 5th bullet from bottom, change "type-qualifier" to "type qualifier". pp. 415 [Thomas] 4th bullet. add conversion specifiers a, A, and F to the list 5th bullet. add conversion specifiers a, A, and F to the list 6th bullet. add conversion specifiers a, A, and F to the list 9th bullet. add conversion specifier a to the list pp. 416 [Tydeman] 3rd bullet, This entry is out of order w.r.t. section numbers. Also, the last reference should be 7.12.2.1.