Document Number: AUSTIN/88 Title: XSHd6 Aardvark Change Request Report Revision Date: 2001-06-01 Source: Andrew Josey, Chair Action: for review This report contains the dispositions of the aardvark comments submitted against the XSH Draft 6. Aardvark Summary Table ______________________ ERN 1 Accept ERN 2 Accept ERN 3 Accept as marked ERN 4 Accept ERN 5 Duplicate of 6 ERN 6 Accept ERN 7 Accept ERN 8 Accept ERN 9 Duplicate of 8 ERN 10 Accept ERN 11 Reject ERN 12 Accept ERN 13 Accept ERN 14 Reject ERN 15 Accept ERN 16 Accept ERN 17 Accept ERN 18 Reject ERN 19 Reject ERN 20 Accept ERN 21 Accept ERN 22 Reject ERN 23 Accept ERN 24 Accept as marked ERN 25 Accept as marked ERN 26 OPEN ERN 27 Reject ERN 28 Reject ERN 29 Accept as marked ERN 30 Accept ERN 31 Accept ERN 32 Accept as marked ERN 33 Reject ERN 34 Accept as marked ERN 35 Accept ERN 36 Accept ERN 37 Reject ERN 38 Accept ERN 39 Accept ERN 40 Accept as marked ERN 41 Reject ERN 42 Accept as marked ERN 43 Accept ERN 44 Accept ERN 45 Accept ERN 46 Accept ERN 47 Accept as marked ERN 48 Reject ERN 49 Accept as marked ERN 50 Accept ERN 51 Accept as marked ERN 52 Accept ERN 53 Accept ERN 54 Accept ERN 55 Accept ERN 56 Reject ERN 57 Accept ERN 58 Accept ERN 59 Reject ERN 60 Accept as marked ERN 61 Accept as marked ERN 62 Accept ERN 63 Reject ERN 64 Accept ERN 65 Accept ERN 66 Accept ERN 67 Accept ERN 68 Accept ERN 69 Accept ERN 70 Accept ERN 71 Accept ERN 72 Accept ERN 73 Accept ERN 74 Accept ERN 75 Accept as marked ERN 76 OPEN ERN 77 Accept ERN 78 Accept ERN 79 Accept as marked ERN 80 Accept ERN 81 Accept ERN 82 Accept ERN 83 Accept _____________________________________________________________________________ COMMENT Enhancement Request Number 1 gwc@unisoft.com BUG in XSHd6 2.1 (rdvk# 16) [gwc implicit decl] Thu, 26 Apr 2001 19:00:59 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 463 Line: 498 Section: 2.1 Problem: C99 does not allow implicit function declarations. Action: Change ", either explicitly or implicitly," to "explicitly" _____________________________________________________________________________ COMMENT Enhancement Request Number 2 drepper@redhat.com Bug in XSHd6 The Compilation Environment (rdvk# 76) {ud-10} Mon, 21 May 2001 08:31:14 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 469 Line: 702- Section: The Problem: (Draft 5 ERN reference: NA ) The table on page 469 lists the functions which are reserved by ISO C for future extensions (or are defined today). This list is not complete. See the Library Future Directions section (7.26) in C99. Action: Add to the table in the correct places: (table updated as per austin-review-l:archive/latest/1034 and in meeting) cerf cerff cerfl cerfc cerfcf cerfcl cexp2 cexp2f cexp2l cexmp1 cexpm1f cexpm1l clog10 clog10f clog10l clog1p clog1pf clog1pl clog2 clog2f clog2l clgamma clgammaf clgammal ctgamma ctgammaf ctgammal [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 3 gwc@unisoft.com BUG in XSHd6 2.3 (rdvk# 20) [gwc rlimit EMFILE] Fri, 27 Apr 2001 18:07:45 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change "maximum number of {OPEN_MAX} file descriptors" to "maximum number of file descriptors". (the soft limit is described as a may fail in getrlimit) _____________________________________________________________________________ Page: 474 Line: 955 Section: 2.3 Problem: All of the descriptions of file-descriptor-related EMFILE errors explicitly refer to {OPEN_MAX}. However, this error also occurs when the RLIMIT_NOFILE limit is exceeded. Action: Change "maximum number of {OPEN_MAX} file descriptors" to "maximum number of file descriptors". Also for each function that can give a file-descriptor-related EMFILE error, add a new EMFILE entry after the existing one, containing the following text: [EMFILE] The number of open file descriptors would exceed the soft limit set for the process. The text should be XSI shaded except in the case of functions that are themselves XSI extensions. The locations for these additions are: page 546 line 3640 section accept page 626 line 5829 section catopen page 723 line 8782 section dup page 730 line 8980 section endgrent page 738 line 9186 section endpwent page 789 line 10888 section fcntl page 848 line 12616 section fopen page 887 line 14072 section freopen page 964 line 16636 section getgrgid page 967 line 16727 section getgrnam page 979 line 17075 section getlogin page 1009 line 17926 section getpwnam page 1012 line 18020 section getpwuid page 1059 line 19495 section iconv_open page 1308 line 27061 section nftw page 1319 line 27348 section open page 1324 line 27547 section opendir page 1335 line 27833 section pipe page 1353 line 28371 section posix_openpt page 2029 line 47420 section tmpfile The ioctl, mq_open, posix_typed_mem_open, sem_open, shm_open, socket, and socketpair functions have been left off the list because they don't explicitly refer to OPEN_MAX. Perhaps for consistency they should be changed to be like all the others... P1086 L20363, P1268 L25927, P1456 L30935, P1749 L39204, P1824 L41307, P1889 L43289, P1891 L43361. _____________________________________________________________________________ COMMENT Enhancement Request Number 4 gwc@unisoft.com BUG in XSHd6 2.4.1 (rdvk# 17) [gwc sigmask func list] Thu, 26 Apr 2001 19:01:48 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 478 Line: 1140 Section: 2.4.1 Problem: The principal function for controlling a thread's signal mask is pthread_sigmask(), but it is not mentioned here. Action: Change "sigaction()" to "pthread_sigmask(), sigaction()" _____________________________________________________________________________ EDITORIAL Enhancement Request Number 5 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 49) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_6 Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 481 Line: 1244-1249 Section: Signal Problem: This paragraph talks about three different things: SIGCONT, real time signals, and SIGCHLD. Comparing it with the description of SIG_IGN, I guess that two paragraph breaks. Action: Insert paragraph break at the start of the RTS shading and before "When=20 a stopped process...". Don't RTS shade the SIGCHLD paragraph _____________________________________________________________________________ EDITORIAL Enhancement Request Number 6 gwc@unisoft.com BUG in XSHd6 2.4.3 (rdvk# 13) [gwc SIGCHLD shading] Fri, 20 Apr 2001 17:58:17 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 481 Line: 1247 Section: 2.4.3 Problem: The text added here for XSHd5 ERN 23 was supposed to be XSI shaded, but has ended up RTS shaded. Action: Change the shading on "When a stopped ... SA_NOCLDSTOP flag" to XSI. Perhaps a line break is needed to separate the two differently shaded text blocks. _____________________________________________________________________________ COMMENT Enhancement Request Number 7 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 50) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 485 Line: 1429-1433 Section: Interaction Problem: Nothing is said about what happens to a stream when a process terminates abnormally due to a signal, but the description of abnormal=20 program termination in xbd6, p.73 l.2422, section 3.297 does not distinguish between abort() and signals. Action: Change end of first sentence to "exit(), abort(), or due to a signal." _____________________________________________________________________________ COMMENT Enhancement Request Number 8 gwc@unisoft.com BUG in XSHd6 2.5.2 (rdvk# 25) [gwc binary stream] Tue, 1 May 2001 20:25:12 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 487 Line: 1496-1514 Section: 2.5.2 Problem: There is no distinction between text and binary streams in POSIX/SUS. (A "b" character in the f[re]open() mode argument has no effect.) Presumably these references to them found their way in via some additions made for the C99 alignment. Action: Delete lines 1497 and 1498. On line 1496 change "restrictions" to "restriction". On line 1514 delete "both text and binary". _____________________________________________________________________________ COMMENT Enhancement Request Number 9 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 67) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_8 Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 487 Line: 1497-1498 Section: Stream Problem: This item refers to text and binary streams, but in POSIX there is no difference between these. Action: Drop this item. _____________________________________________________________________________ COMMENT Enhancement Request Number 10 gwc@unisoft.com BUG in XSHd6 2.5.2 (rdvk# 26) [gwc wcf on bytestream] Tue, 1 May 2001 20:25:12 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 487 Line: 1509-1510 Section: 2.5.2 Problem: There is a conflict between the statements at lines 1493-1494 and at lines 1509-1510. One says that wide-character input/output functions cannot be applied to byte-oriented streams. The other says, "If a wide-character input/output function is applied to a byte-oriented stream, ..." Action: Delete "If a wide-character input/output function is applied to a byte-oriented stream, the encoding rule used is undefined." _____________________________________________________________________________ COMMENT Enhancement Request Number 11 drepper@redhat.com Bug in XSHd6 Threads (rdvk# 83) {ud-11} Mon, 21 May 2001 22:48:44 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope, please file an interpretation. _____________________________________________________________________________ Page: 501 Line: 2076- Section: Threads Problem: (Draft 5 ERN reference: NA ) One thing I meant to bring up for some time but always forgot: are the functions supposed to be thread-safe? In all implementations I'm aware of the functions are implemented either as macros or inline functions accessing global arrays. If a new locale is loaded these arrays are changed. It will consist in most cases of a simple change of the pointer to the array but there is still the possibility that at one brief moment the pointer value is invalid (just image the pointer consists of ore than one word) or inconsistent (the thread calling setlocale is scheduled in between changing the arrays used for isXXX() and the arrays used for toXXX() which could mean that at some point islower(c) && tolower (c) == c might be violated). Long story, short questions: - are these problems accepted for interfaces declared thread-safe - if not shouldn't the (and perhaps ) interfaces be added to the table on page 501? Action: Depending on the answer given to the question above, either do nothing (or perhaps improve the documentation of the term thread-safe), or add _tolower, _toupper, isalnum, islapha, isascii, isblank, iscntrl, isdigit, isgraph, islower, isprint, ispunct, isspace, isupper, isxdigit, toascii, tolower, toupper, and perhaps iswalnum, iswalpha, iswblank, iswcntrl, iswdigit, iswgraph, iswlower, iswprint, iswpunct, iswspace, iswupper, iswxdigit, iswctype, towctrans, towlower, towupper, wctrans, wctype to the table on top of page 501. The perhaps don't have to be added since most implementations define real functions anyway and requiring them to be thread-safe is not completely out of this world. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 12 drepper@redhat.com Bug in XSHd6 Threads (rdvk# 80) {ud-14} Tue, 22 May 2001 02:08:19 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 501 Line: 2076- Section: Threads Problem: (Draft 5 ERN reference: NA ) The tables lists lgamma() but not the newly introduced lgammaf() and lgammal() functions. Action: Add lgammaf() and lgammal() to the table on top of page 501. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 13 ajosey@opengroup.org Bug in XSHd6 2.9.3 (rdvk# 35) {aj.hc.4} Thu, 10 May 2001 08:22:22 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 501 Line: 2113 Section: 2.9.3 Problem: (Draft 5 ERN reference: NA ) Reference to unknown function Action: Change pthread_mutex_timedwait() to pthread_mutex_timedlock() on the following line 2113 TMO It returns successfully from pthread_mutex_timedwait( ) with m as the mutex argument. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 14 ajosey@opengroup.org Bug in XSHd6 2.10.12 sockatmark (rdvk# 32) {aj.hc.7} Thu, 10 May 2001 09:03:46 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This should be forwarded for input into TC1. (note to AJ, create man page and associated changes for other pages as extra doc) _____________________________________________________________________________ Page: 511 Line: 2513 Section: 2.10.12 Problem: (Draft 5 ERN reference: NA ) This normative text references a sockatmark() function which is not included in the specification. The function does occur in 1003.1g but was not listed in those items brought forward into the revised specification. Action: Either delete the sentence or add a definition of the sockatmark() function defined in 1003.1g 5.4.13.1 If added, the prototype would also need adding to _____________________________________________________________________________ COMMENT Enhancement Request Number 15 gwc@unisoft.com BUG in XSHd6 2.12 Data Types (rdvk# 29) [gwc types list] Thu, 10 May 2001 18:35:05 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 530,531 Line: 3276-3330 Section: 2.12 Problem: The descriptions of some defined types do not match the requirements for the type definitions as stated in the XBD header pages. Action: Change the descriptions of each of following types as indicated. L3276 clock_t: "Arithmetic" -> "Integer or real-floating" L3288 gid_t: "Arithmetic" -> "Integer" L3290 id_t: "Arithmetic" -> "Integer" L3292 ino_t: "Arithmetic" -> "Unsigned integer" L3295 mode_t: "Arithmetic" -> "Integer" L3298 nlink_t: "Arithmetic" -> "Integer" L3299 off_t: "arithmetic" -> "integer" L3300 pid_t: "arithmetic" -> "integer" L3314 rlim_t: "arithmetic" -> "integer" L3323 ssize_t: "Arithmetic" -> "Signed integer" L3324 suseconds_t: "arithmetic" -> "integer" L3326 time_t: "Arithmetic" -> "Integer or real-floating" L3329 uid_t: "Arithmetic" -> "Integer" L3330 useconds_t: "Integer" -> "Unsigned integer" _____________________________________________________________________________ EDITORIAL Enhancement Request Number 16 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 41) [DST-3] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 561 Line: 4040 Section: aio_read Problem: the -> operator is split across 2 lines. Action: Fix so it isn't. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 17 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 51) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 595 Line: 5024-5027 Section: basename() Problem: I would like to see at least one more example showing the treatment of multiple '/' characters. Action: Add two lines to the table: "///" "/" "//usr//lib//" "lib" _____________________________________________________________________________ COMMENT Enhancement Request Number 18 muewi@baumanagement.bremen.de BUG in XSHd6 (rdvk# 73) [muewi-3] Fri, 18 May 2001 16:17:25 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This is covered by the pathname resolution general concept. _____________________________________________________________________________ Page: 643 Line: 6282-6285 Section: chmod() Problem: XRATd6, on p. 3316, l.952-959, lists which interfaces follow symbolic links and which don't. For four interfaces--chmod(), chown(), link(), and utime()--, this information does not appear in the normative text. (But it does appear in the description of stat().) Action: Here an also at page 646 line 6395 section chown() page 1159 line 22540 section link() page 2069 line 48558 section utime() page 2071 line 48633 section utimes() add: When path [path1 for link()] names a symbolic link, the function shall operate on the file named by the link, not the link itself. [Ed recommendation: None Is this covered by pathname resolution?] _____________________________________________________________________________ COMMENT Enhancement Request Number 19 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 52) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope _____________________________________________________________________________ Page: 643 Line: 6287-6290 Section: chmod() Problem: Clearing S_ISGID on non-executables may produce unexpected results since this bit has been abused, e.g., for mandatory locking. Action: Change last line of the paragraph to "successful return from chmod() if any execute bit is set." _____________________________________________________________________________ EDITORIAL Enhancement Request Number 20 gwc@unisoft.com BUG in XSHd6 c99 (rdvk# 9) [gwc confstr V6WRE] Thu, 19 Apr 2001 18:04:27 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 670 Line: 7217 Section: confstr Problem: Missed change from XBDd5 ERN 322. Action: Add after line 7217: _CS_POSIX_V6_WIDTH_RESTRICTED_ENVS [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 21 drepper@redhat.com Bug in XSHd6 confstr() (rdvk# 19) {ud-1} Wed, 25 Apr 2001 01:49:45 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 670 Line: 7234 Section: confstr() Problem: (Draft 5 ERN reference: NA ) The newly introduced _CS_V6_WIDTH_RESTRICTED_ENVS value is missing in the description of confstr(). Action: Add _CS_V6_WIDTH_RESTRICTED_ENVS at the end of the list. [Ed recommendation: Accept] _____________________________________________________________________________ OBJECTION Enhancement Request Number 22 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 42) [DST-4] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: The wording is the best that could be achieved and retain consensus _____________________________________________________________________________ Page: 706 Line: 8181 Section: dbm_clearerr Problem: This is still hand-wavy vague; I'm sure the authors understand it, but it's still not stated in a clearly testable way. As a consequence, it essentially constrains a portable application to not use the dbm interfaces at all (in a useful way). Action: Is this what's meant by the first sentence? If for any _key_ and _content_ pair passed to dbm_store(), if the sum of the dsize fields of _key_ and _content_ exceeds the internal block size, the result is unspecified. The second sentence is unfixable as it stands and makes the function useless, because the application has no way of knowing what the hash algorithm is. A legal hash algorithm is "1". (That is, all keys hash to the single value of "1"). No, that's NOT a quality of implementation issue, because it's imaginable that for SOME dataset that the user chooses, the hash function happens to be degenerate for that dataset. The user can't know whether he happens to have that dataset. Thus this is stating that no database can portably exceed one internal block size, which has a minimum value of 1023. Thus no portable application can use a database of more than 1023 bytes of key/value pairs. I'm not sure what the solution might be, but putting the requirement on the application to avoid a problem it can't determine exists is just not reasonable. _____________________________________________________________________________ COMMENT Enhancement Request Number 23 drepper@redhat.com Bug in XSHd6 dlsym() (rdvk# 82) {ud-12} Tue, 22 May 2001 01:15:40 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 719 Line: 8614 Section: dlsym() Problem: (Draft 5 ERN reference: NA ) RTLD_DEFAULT was added to the Application Usage section in this draft but I forgot to mention the second change which is necessary. Action: Change line 8614 to: The RTLD_DEFAULT and RTLD_NEXT flags are reserved for future use. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 24 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 53) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 752 Line: 9528-9529,9546-9547 Section: errno Problem: The first sentence says that errno must not be used for error checking=20 unless a function's return value indicates that it has definitely been=20 set to a valid value, the second one seems to allow to rely solely on errno for error checking. Action: Delete line 9546-9547. [Ed recommendation: Accept as marked Change "program" to "application", change "that uses errno for error checking" to "that needs to examine the value of errno to determine the error" ] _____________________________________________________________________________ COMMENT Enhancement Request Number 25 curtis.smith@simtrol.com Bug in XSHd6 exec (rdvk# 77) {cms-main} Mon, 21 May 2001 11:54:23 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add new rationale from minutes _____________________________________________________________________________ Page: 754 Line: 9586 Section: exec Problem: (Draft 5 ERN reference: NA ) ISO C says the function called at program startup shall be defined as int main ( void ); OR int main ( int argc, char *argv[] ); or in some other implementation-defined manner. However, here only the second alternative is listed (even though the majority of examples in this standard use the first alternative). Action: Change prototype to int main (void); or int main (int argc, char *argv[]); Possibly also add comment about implementation-defined manners (e.g., using environ as third argument). _____________________________________________________________________________ COMMENT Enhancement Request Number 26 gwc@unisoft.com BUG in XSHd6 exec (rdvk# 18) [gwc exec stdin] Wed, 25 Apr 2001 21:47:11 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_____OPEN Rationale for rejected or partial changes: Refer to Base WG with recommendation they issue a resolution to enable this change. Base WG to respond by June 8 _____________________________________________________________________________ Page: 754 Line: 9606 Section: exec Problem: These lines say that the file is used as standard input to the command interpreter. This is clearly in error. Standard input is inherited from the calling process, as with all other open file descriptors (that aren't close-on-exec). The mistake goes back at least to XPG3. The suggested replacement wording below borrows from the description of the system() function. If this needs a Base Working Group resolution to bring it into scope, hopefully it is not too late to get one. (Since this text is taken from SUSv2, not from POSIX.1, it doesn't need a POSIX.1 interpretation.) Action: Replace lines 9606-9609 with: If the process image file is not a valid executable object, and the system does not recognize it as something that cannot be executed (and thus returns [EINVAL]), execlp() and execvp() shall execute a command interpreter and the environment of the executed command shall be as if the process invoked the sh utility using execl() as follows: execl(, arg0, file, arg1, ..., (char *)0); where is an unspecified pathname for the sh utility, and, for execvp(), where arg0, arg1, etc. correspond to the values passed to execvp() in argv[0], argv[1], etc. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 27 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 54) Also exit() muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Correct as is . _____________________________________________________________________________ Page: 757 Line: 9733 Section: exec Problem: Wrong grammar: "process'" Action: Here and also at page 765 line 10034 section exit() editorial muewi-1 Change to: "process's" or rewrite to avoid possessive case. _____________________________________________________________________________ COMMENT Enhancement Request Number 28 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 55) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This may be right but would need an interpretation request to determine so. _____________________________________________________________________________ Page: 764 Line: 10052-10055 Section: exit() Problem: The state of locks established by other processes is not specified after a call to exit(). Action: Change end of paragraph to "_Exit(), _exit(), or exit()." _____________________________________________________________________________ OBJECTION Enhancement Request Number 29 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 43) [DST-5] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: The application shall have the appropriate privileges or be the owner of the file named by path and have write permission. _____________________________________________________________________________ Page: 776 Line: 10436 Section: fattach Problem: Shallification. Action: Should read: The application shall have the appropriate privileges or shall be the owner of the file named by path and shall have write permission. or at least The application shall have the appropriate privileges and be the owner of the file named by path and have write permission. (At least keep the tense consistent.) _____________________________________________________________________________ EDITORIAL Enhancement Request Number 30 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 2) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 803 Line: 11333 Section: RETURN Problem: Typo suspected: execptions Action: Change "execptions" to "exceptions" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 31 drepper@redhat.com Bug in XSHd6 dlsym() (rdvk# 81) {ud-13} Tue, 22 May 2001 01:15:40 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 815 Line: 11614 Section: feupdateenv() Problem: (Draft 5 ERN reference: NA ) All preprocessordirectives must have the leading # in the first column. Action: Lines 11613 to 11615 would be written { # pragma STDC FENV_ACCESS ON double result; _____________________________________________________________________________ EDITORIAL Enhancement Request Number 32 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 56) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 824 Line: 11858 Section: fgetpos() Problem: The change history says that [EIO] has been added. But I don't see this in the ERRORS section. Action: Either remove item from change history or really add [EIO] and description to ERRORS. [Ed recommendation: Accept as marked Remove EIO from the change history ] _____________________________________________________________________________ COMMENT Enhancement Request Number 33 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 57) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The text was felt to be helpful _____________________________________________________________________________ Page: 861 Line: 13089-13090 Section: fprintf() Problem: All this paragraph says is already expressed by using restrict in the prototypes. Action: Remove paragraph. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 34 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 58) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below__X__ Duplicate_____ Reject_____ Rationale for rejected or partial changes: line 13112 change "each argument... once." to "each conversion specification uses the first unused argument in the argument list" _____________________________________________________________________________ Page: 861 Line: 13113 Section: fprintf() Problem: There may be arguments which will not be used at all, as lines 13096f. say. Action: Change "exactly" to "at most". _____________________________________________________________________________ EDITORIAL Enhancement Request Number 35 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 59) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 862 Line: 13128 Section: fprintf() Problem: The XSI shading starts at the beginning of a line which is not the start of a paragraph, but the XSI indicator in the margin appears on the preceding line. (This is just one place this occurs, it seems to be a general problem with the definition of the typesetter macros generating the margin indication and shading.) Action: Check typesetter macro definitions to ensure marginals and shading appear on same line. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 36 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 60) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 862 Line: 13129 Section: fprintf() Problem: A space is missing between "S" and "conversion". Action: Add space. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 37 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 61) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope, taken from ISO C _____________________________________________________________________________ Page: 863 Line: 13187,13200 Section: Problem: Why are there separate length modifiers for long double and long long? Action: Allow "L" and "ll" to be used as synonyms. _____________________________________________________________________________ COMMENT Enhancement Request Number 38 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 62) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 864 Line: 13249-13250 Section: Problem: The conversion specifier "G" behaves like "E" of "F", not just like "E". Action: Change text in parentheses to: or in style "F" or "E" in the case... _____________________________________________________________________________ COMMENT Enhancement Request Number 39 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 63) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 865 Line: 13255 Section: fprintf() Problem: A radix character must appear as well when the "#" flag is used. Action: Change end of sentence to: ...followed by a digit or a "#" flag is present. _____________________________________________________________________________ OBJECTION Enhancement Request Number 40 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 44) [DST-6] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: add "be" to line 13297 p865 (written->be written) add example on p871 Printing wide characters Suppose that L`@` expands to three bytes wchar_t wz [3] = L"@@"; // zero-terminated wchar_t wn [3] = L"@@@"; // unterminated fprintf (stdout,"%ls", wz); // Outputs 6 bytes fprintf (stdout,"%ls", wn); // Undefined because wn has no terminator fprintf (stdout,"%4ls", wz); // Outputs 3 bytes fprintf (stdout,"%4ls", wn); // Outputs 3 bytes, no terminator needed fprintf (stdout,"%9ls", wz); // Outputs 6 bytes fprintf (stdout,"%9ls", wn); // Outputs 9 bytes, no terminator needed fprintf (stdout,"%10ls", wz); // Outputs 6 bytes fprintf (stdout,"%10ls", wn); // Undefined because wn has no terminator In the last line of the example, after processing three characters nine bytes have been output. The fourth character must then be examined to determine whether it converts to one byte or more. If it converts to more than one byte, the output is only nine bytes. Since there is no fourth character in the array the behavior is undefined. _____________________________________________________________________________ Page: 865 Line: 13294 Section: fprintf Problem: The latter half of the paragraph is the problem. I can't make sense of this sentence. I *THINK* it's trying to say something about the input being null-terminated, but it's unclear. Action: Firstly, the use of "character" here is ambiguous; stick to bytes, because I believe you're talking in terms of storage. If you are talking in units of multibyte characters, then the "(bytes)" on 13293 is just plain wrong. (Note, in on line 13297, "character" *clearly* means multibyte character; this only adds to the confusion if character is taken elsewhere to be synonymous with byte.) As for the last sentence of the paragraph, is it trying to say: If a precision is specified, no more than that many bytes shall be put into the output buffer. If the input array contains fewer wide characters than would be required to use all the bytes specified by the precision, the input array shall be terminated by a wide null. [Ed recommendation: None Response from Clive below Donn Terry said: > @ Page 865 Line 13294 Section fprintf Objection [DST-6] > > Problem: > The latter half of the paragraph is the problem. > I can't make sense of this sentence. I *THINK* it's trying to say > something about the input being null-terminated, but it's unclear. It's saying that the array must be zero-terminated if the precision is large enough that you don't stop short. In other words, suppose that L'@' expands to 3 bytes: wchar_t wz [3] = L"@@"; // zero-terminated wchar_t wn [3] = L"@@@"; // unterminated fprintf ("%ls", wz); // Outputs 6 bytes fprintf ("%ls", wn); // Undefined because no terminator fprintf ("%4ls", wz); // Outputs 3 bytes fprintf ("%4ls", wn); // Outputs 3 bytes, no terminator needed fprintf ("%9ls", wz); // Outputs 6 bytes fprintf ("%9ls", wn); // Outputs 9 bytes, no terminator needed fprintf ("%10ls", wz); // Outputs 6 bytes fprintf ("%10ls", wn); // Undefined because no terminator > Action: > Firstly, the use of "character" here is ambiguous; stick to bytes, > because > I believe you're talking in terms of storage. If you are talking in > units > of multibyte characters, then the "(bytes)" on 13293 is just plain > wrong. No, it's correct. > (Note, in on line 13297, "character" *clearly* means multibyte > character; Indeed; furthermore, the word "be" needs inserting. > As for the last sentence of the paragraph, is it trying to say: > > If a precision is specified, no more than that many bytes shall be > put into the output buffer. If the input array contains fewer wide > characters than would be required to use all the bytes specified by > the > precision, the input array shall be terminated by a wide null. Not quite. Look at the last example above. After 3 characters it's output 9 bytes. It now needs to look at the fourth to determine whether or not that converts to one byte or more. If it converted to more, the output would be only 9 bytes, but since the character doesn't exist there's no way to know, and so the behaviour is undefined. Compare with the precision of 9. That wording was carefully crafted. ] _____________________________________________________________________________ COMMENT Enhancement Request Number 41 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 64) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Wording from ISO C _____________________________________________________________________________ Page: 866 Line: 13313-13314 Section: Problem: "Correctly rounded" is ambiguous. Action: Either specify what rounding mode to use or replace by "..., the value, if not representable exactly, shall be one of the two representable ones nearest to the true value." _____________________________________________________________________________ COMMENT Enhancement Request Number 42 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 65) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: add a see also to strfmon() _____________________________________________________________________________ Page: 871 Line: 13506-13542 Section: fprintf() Problem: This example deals mostly with intricacies of strfmon(); its usage of printf() style functions is quite trivial. Action: Move example to strfmon() description. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 43 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 45) [DST-7] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 929 Line: 15463 Section: fwprintf Problem: spelling. Action: shlal -> shall. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 44 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 3) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 935 Line: 15674 Section: DESCRIPTION Problem: Typo suspected: execut Action: Change "execut" to "execute" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 45 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 46) [DST-8] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 935 Line: 15674 Section: fwscanf Problem: spelling Action: execut -> execute [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 46 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 1) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 959 Line: 16506 Section: RATIONALE Problem: Typo suspected: environmentt Action: Change "environmentt" to "environment" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 47 mccann@zk3.dec.com BUG in XSHd6 (rdvk# 27) [JMD6-2] Tue, 1 May 2001 13:46:38 -0400 (EDT) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Insert the following text, shaded IP6, between lines 17305 and 17306: If the socket address structure contains an IPv4-mapped IPv6 address or an IPv4-compatible IPv6 address, the implementation shall extract the embedded IPv4 address and lookup the node name for that IPv4 address. Note: The IPv6 unspecified address (::) and the IPv6 loopback address (::1) are not IPv4-compatible addresses. If the address is the IPv6 unspecified address (::), a lookup is not performed, and the EAI_NONAME error is returned. Add to Application Usage Given the IPv4-mapped IPv6 address ::ffff:1.2.3.4 , the implementation performs a lookup as if the socket address structure contains the IPv4 address 1.2.3.4. (note to ed:see inet_ntop for examples of how to do the formatting) _____________________________________________________________________________ Page: 986 Line: 17305 Section: 3 Problem: The getnameinfo description does not specify the special handling required for lookups of IPv4-mapped IPv6 addresses, and IPv4-compatible IPv6 addresses. In IETF RFC 2553, Basic Socket Interface Extensions for IPv6, the definition of getipnodebyaddr described special handling for IPv4-mapped and IPv4-compatible IPv6 addresses. Essentially, when given one of these addresses, the implementation is supposed to extract the embedded IPv4 address and perform an IPv4 address lookup for that address. This eliminates the need to store IPv4-mapped and IPv4-compatible addresses in the naming system (e.g. DNS), which would be duplicates of the IPv4 entries. Here is the relevant text from RFC 2553: One possible source of confusion is the handling of IPv4-mapped IPv6 addresses and IPv4-compatible IPv6 addresses, but the following logic should apply. 1. If af is AF_INET6, and if len equals 16, and if the IPv6 address is an IPv4-mapped IPv6 address or an IPv4-compatible IPv6 address, then skip over the first 12 bytes of the IPv6 address, set af to AF_INET, and set len to 4. 2. If af is AF_INET, lookup the name for the given IPv4 address (e.g., query for a PTR record in the in-addr.arpa domain). 3. If af is AF_INET6, lookup the name for the given IPv6 address (e.g., query for a PTR record in the ip6.int domain). 4. If the function is returning success, then the single address that is returned in the hostent structure is a copy of the first argument to the function with the same address family that was passed as an argument to this function. All four steps listed are performed, in order. Also note that the IPv6 hex addresses "::" and "::1" MUST NOT be treated as IPv4- compatible addresses, and if the address is "::", HOST_NOT_FOUND MUST be returned and a query of the address not performed. In the latest version of the Basic Socket Interface Extensions for IPv6 specification (draft-ietf-ipngwg-rfc2553bis-03.txt, work-in-progress), the getipnodebyaddr API was removed, the intent being that getnameinfo would be used for address lookups. However, the getnameinfo API description was not updated to describe handling of IPv4-mapped and IPv4-compatible IPv6 addresses. Action: Insert the following text, shaded IP6, between lines 17305 and 17306: If the socket address structure contains an IPv4-mapped IPv6 address or an IPv4-compatible IPv6 address, the implementation will extract the embedded IPv4 address and lookup the node name for that IPv4 address. For example, given the IPv4-mapped IPv6 address ::ffff:1.2.3.4, the implementation would perform a lookup as if the socket address structure contained the IPv4 address 1.2.3.4. In this example, if DNS were the underlying name service, the implementation would query for a PTR resource record in the 4.3.2.1.in-addr.arpa domain. Note that the IPv6 unspecified address (::) and the IPv6 loopback address (::1) are not IPv4-compatible addresses. If the address is the IPv6 unspecified address (::), a lookup will not be performed, and the EAI_NONAME error will be returned. _____________________________________________________________________________ COMMENT Enhancement Request Number 48 Jon Hitchcock Bugs in XSHd6 (rdvk# 68) {jjh57} Fri, 18 May 2001 10:57:21 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope. The wording at the moment corresponds to the Base document. If the submitter feels strongly an interpretation should be submitted. _____________________________________________________________________________ Page: 1043 Line: 19032 Section: gmtime_r Problem: The draft says gmtime_r() shall return a null pointer if "UTC is not available", but the newly inserted lines 19013-19018 make clear that the conversion only depends on the "seconds since the Epoch" expression which should always be available. Action: Delete the words "or UTC is not available,". _____________________________________________________________________________ COMMENT Enhancement Request Number 49 eggert@twinsun.com Bug in XSHd6 iconv_open (rdvk# 75) {eggert-2001051801} Fri, 18 May 2001 19:31:55 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add to APP USAGE Application developers should consult the system documentation to determine the supported codesets and their naming scheme. _____________________________________________________________________________ Page: 1059 Line: 19509 Section: iconv_open Problem: (Draft 5 ERN reference: NA ) The standard says that codeset names are implementation-defined, which leads to some application portability problems. It's far too late to standardize the names, but to help application developers and implementers see what sorts of names are in common use the rationale should explain some common naming choices. Action: Replace line 19509 with the following text: Not only are codesets implementation-defined, their names are too. SVR4-derived implementations use X11 names in some cases (e.g., "ISO8859-1") and proprietary names in others (e.g., "iso88591"). However, the X11 name registry is no longer maintained. GNU-derived implementations use IANA-registered names, and among those, the preferred MIME name if available (e.g., "ISO-8859-1"); see . Application developers should consult the implementation's documentation to determine the supported codesets and their naming scheme. _____________________________________________________________________________ COMMENT Enhancement Request Number 50 curtis.smith@simtrol.com Bug in XSHd6 localtime (rdvk# 78) {cms-maintime} Mon, 21 May 2001 12:09:54 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1179 Line: 23152 Section: localtime Problem: (Draft 5 ERN reference: NA ) For consistency, main should be defined with explicit return type. Action: Change main() to int main(void) Also on page 1557, line 33602 -- this example also needs a "return 0;" inserted between lines 33614/33615 page 2016, line 46993 (note 2026 changed to 2016 by AJ) _____________________________________________________________________________ EDITORIAL Enhancement Request Number 51 curtis.smith@simtrol.com Bug in XSHd6 localtime (rdvk# 40) {cms.localtime} Fri, 11 May 2001 16:40:50 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change %s%ld to %s%ju And Change (long)result to (uintmax_t)result Also Change the identical example under "time()" on page 2016, line 46993 _____________________________________________________________________________ Page: 1179 Line: 23156 Section: localtime Problem: (Draft 5 ERN reference: NA ) The example format string specified does not correspond to the example output in which the asctime result and time_t value are printed on separate lines. Also, the time_t value is cast to long before printing, but that seems rather a questionable choice since longs may be as short as 32 bits, and one would expect time_t to be longer in any system not trying to maintain binary compatibility with old code. This example is repeated under time() heading on page 2016, line 46993. Action: Change %s%ld to %s\n%ju And Change (long)result to (uintmax_t)result Also Change the identical example under "time()" on page 2016, line 46993 _____________________________________________________________________________ EDITORIAL Enhancement Request Number 52 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 4) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1238 Line: 24921 Section: DESCRIPTION Problem: Typo suspected: gorup Action: Change "gorup" to "group" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 53 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 6) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1238 Line: 24922 Section: DESCRIPTION Problem: Typo suspected: proces Action: Change "proces" to "process" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 54 ajosey@opengroup.org Bug in XSHd6 mmap (rdvk# 38) {aj.hc.1} Wed, 9 May 2001 17:33:39 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1256 Line: 25554 Section: mmap Problem: (Draft 5 ERN reference: na ) The reference to memlockall() in rationale text here should be mlockall() Action: change memlockall() to mlockall() [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 55 mitr@volny.cz Bug in XSHd6 popen (rdvk# 10) {???} Thu, 19 Apr 2001 23:15:45 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1342 Line: 28077-28078 Section: popen Problem: (Draft 5 ERN reference: NA ) The paragraph duplicates lines 28051-28052 in APPLICATION USAGE section. Action: Remove lines 28077-28078. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 56 drepper@redhat.com Bug in XSHd6 posix_trace_attr_getclockres() (rdvk# 79) {ud-15} Tue, 22 May 2001 04:50:06 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The notation shows that you just need trace.h for these funcs below the second trace.h and not time.h _____________________________________________________________________________ Page: 1395 Line: 29457 Section: posix_trace_attr_getclockres() Problem: (Draft 5 ERN reference: NA ) I don't understand the purpose of the second #include line. The following functions are not part of a different profile. When an additional header (like ) was needed in other places this wasn't specially marked either. Action: Unless there is a good reason remove this line. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 57 ajosey@opengroup.org Bug in XSHd6 posix_trace_attr_getinherited (rdvk# 30) {aj.hc.9} Thu, 10 May 2001 10:17:05 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1400 Line: 29599 Section: posix_trace_attr_getinherited Problem: (Draft 5 ERN reference: NA ) The reference to the posix_trace_status_info( ) function should be the posix_trace_status_info( ) structure defined in Action: Change "posix_trace_status_info( ) function" "posix_trace_status_info structure defined in " [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 58 ajosey@opengroup.org Bug in XSHd6 posix_trace_getnext_event (rdvk# 31) {aj.hc.8} Thu, 10 May 2001 09:56:06 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1442 Line: 30640 Section: posix_trace_getnext_event Problem: (Draft 5 ERN reference: NA ) The reference to the posix_trace_event_info( ) function should be the posix_trace_event_info structure defined in Action: Change "posix_trace_event_info( ) function" to "posix_trace_event_info structure defined in " [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 59 curtis.smith@simtrol.com Bug in XSHd6 pthread_create() (rdvk# 72) {cms-pthread_create} Fri, 18 May 2001 14:05:29 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: See austin-group-l mail archive for discussion _____________________________________________________________________________ Page: 1541 Line: 33095-33097 Section: pthread_create() Problem: (Draft 5 ERN reference: NA ) The qualification of the pthread_create's fourth argument (arg) with "restrict" seems inappropriate. It is passed as an argument to start_routine, but without the restrict qualification. One possible resolution would be to change start_routine's sole argument to have the restrict qualifier, but doing so would prohibit the routine from access to the pointed-to memory by using other pointers, possibly obtained by function calls to get a pointer to the same memory. When creating a thread, I typically pass the address of some sort of a thread-data-block to pthread_create. I also would like to provide a function that would return the address of the block, and I believe this is a reasonable expectation. The data block can contain thread-specific information, possibly including the pthread_t element passed as the argument to pthread_create, making me question the legitimacy of having restrict on the argument, as well. I could imagine others wanting to include the pthread_attr_t element in the data block, too. Thus I question the use of restrict on all of the parameters of pthread_create. I believe the chosen usage of restrict was based on the rough rule of thumb, when you have one pointer argument, leave off restrict; with more, put on restrict, but since the nature of start_routine is not well-defined, the rule of thumb cannot apply. Action: Remove all "restrict" keywords from pthread_create function prototype. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 60 curtis.smith@simtrol.comBug in XSHd6 pthread_key_delete rationale (rdvk# 15) {cms-zeroing} Wed, 25 Apr 2001 20:03:51 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: change "zero'ing a reference count" to "setting a reference count to zero" _____________________________________________________________________________ Page: 1565 Line: 33833 Section: pthread_key_delete Problem: (Draft 5 ERN reference: NA ) extraneous apostrophe Action: change "zero'ing" to "zeroing" or change "zero'ing a reference count" to "setting a reference count to zero" _____________________________________________________________________________ COMMENT Enhancement Request Number 61 ajosey@opengroup.org Bug in XSHd6 regcomp (rdvk# 34) {aj.hc.5} Thu, 10 May 2001 08:41:22 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: yes replace regmatch() with regexec() _____________________________________________________________________________ Page: 1698 Line: 37656 Section: regcomp Problem: (Draft 5 ERN reference: NA ) The text in rationale refers to a non-existent function regmatch() (this was also in POSIX.2 rationale this way) Action: replace regmatch() with regexec() ? _____________________________________________________________________________ EDITORIAL Enhancement Request Number 62 ajosey@opengroup.org Bug in XSHd6 regcomp (rdvk# 33) {aj.hc.6} Thu, 10 May 2001 08:49:40 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1698 Line: 37681 Section: regcomp Problem: (Draft 5 ERN reference: NA ) This rationale text talks about a "new type" being introduced, but its no longer new. Action: Change "A new type regoff_t is used" to "The type regoff_t is used" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 63 gwc@unisoft.com BUG in XSHd6 rename (rdvk# 11) [gwc rename ctime] Fri, 20 Apr 2001 18:00:03 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The group felt it better to be nonspecific _____________________________________________________________________________ Page: 1710 Line: 38021 Section: rename Problem: XSHd5 ERN 1137 was "accepted as marked", with the added text being just the first paragraph of what was requested. I assume this was because the rest was considered inappropriate (or perhaps just too long). I still think it would be good to draw attention to the effect that this difference between implementations can have on backups in particular. I suggest a very short addition which does this. Action: Append to line 38021, "In particular this affects backup schemes where st_ctime values are used to determine which files to include on incremental backups." Do the same to XCUd6 at: page 2857 line 25244 section mv comment _____________________________________________________________________________ EDITORIAL Enhancement Request Number 64 curtis.smith@simtrol.com Bug in XSHd6 rmdir errors (rdvk# 23) {cms-rmdir} Tue, 1 May 2001 21:08:50 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1717 Line: 38262-38263 Section: rmdir Problem: (Draft 5 ERN reference: NA ) Line is split improperly, and NAME_MAX is formatted as an error name. Action: Change "longer than NAME_MAX" to "longer than {NAME_MAX}." [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 65 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 5) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1736 Line: 38841 Section: RETURN Problem: Typo suspected: paramenters Action: Change "paramenters" to "parameters" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 66 ajosey@opengroup.org Bug in XSHd6 sigaction (rdvk# 36) {aj.hc.3} Wed, 9 May 2001 17:57:01 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1839 Line: 41768 Section: sigaction Problem: (Draft 5 ERN reference: na ) Typo Action: Change "whnever" to "whenever" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 67 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 7) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1839 Line: 41768 Section: DESCRIPTION Problem: Typo suspected: whnever Action: Change "whnever" to "whenever" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 68 ajosey@opengroup.org Bug in XSHd6 sigaction (rdvk# 37) {aj.hc.2} Wed, 9 May 2001 17:54:30 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1839 Line: 41772 Section: sigaction Problem: (Draft 5 ERN reference: na ) The reference to the sigstack() function should be removed since this was withdrawn in this issue Action: Remove "or sigstack" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 69 Jon Hitchcock Bugs in XSHd6 (rdvk# 69) {jjh58} Fri, 18 May 2001 10:57:21 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: (note accepted since conflicts with ISO C) _____________________________________________________________________________ Page: 1908,1940 Line: 43759,44780 Section: strchr,strrchr Problem: The descriptions of strchr() and strrchr() say that the c argument is converted to an unsigned char. On implementations where "char" is signed, it will be impossible to find bytes in the string s that have negative values. This is not sensible, and it conflicts with the C standard. It would be OK if the description said that bytes in the string s are also interpreted as type unsigned char, but that would seem to be an unnecessary complication. Action: Change "an unsigned" to "a" (unshaded) in each description. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 70 Joseph S. Myers BUG in XSHd6 (rdvk# 21) [JSM-1] Fri, 27 Apr 2001 16:04:23 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1922 Line: 44222 Section: strfmon Problem: The examples added for left-justification [XSHd5 ERN1216] are incorrect: the difference between left justification and right justification is not shown, and the digits to the left of the decimal point are formatted in one too few spaces. They should be as listed in Austin/77r2, which has the examples correct. Action: Change the output for the left-justified example (lines 44222-4) to (with | indicating the left and right of the output): | $ 123.4500 | |-$ 123.4500 | | $ 3,456.7810 | and that for the right-justified example (44225-7) to: | $ 123.4500| | -$ 123.4500| | $ 3,456.7810| Note that these are as shown in the marked resolution to XSHd5 ERN1216, but the PDF does not show them as indicated in that resolution. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 71 Donn Terry Bug in xbdd5 Bug in xsi (rdvk# 47) [DST-9] Tue, 15 May 2001 12:37:25 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1944 Line: 44950 Section: strtod Problem: spelling Action: adn -> and [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 72 curtis.smith@simtrol.com Bug in XSHd6 swprintf synopsis (rdvk# 28) {cms-swprintf} Wed, 2 May 2001 16:32:13 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1964 Line: 45462 Section: swprintf Problem: (Draft 5 ERN reference: NA ) function arguments conflict with original reference at line 15282 and with ISO 9899 Action: Add restrict keywords as follows: int swprintf(wchar_t * restrict ws, size_t n, const wchar_t * restrict format, ...); [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 73 ajosey@opengroup.org Bug in XSHd6 sysconf (rdvk# 14) {aj.apr20.2} Mon, 23 Apr 2001 15:02:01 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1970 Line: 45662 Section: sysconf Problem: (Draft 5 ERN reference: NA ) On backing out the subprofiling options from normative text, as part of addressing the issue in an informative annex, we missed removing the following symbols in sysconf() 45622 _POSIX_BASE _SC_BASE 45623 _POSIX_C_LANG_SUPPORT _SC_C_LANG_SUPPORT 45624 _POSIX_C_LANG_SUPPORT_R _SC_C_LANG_SUPPORT_R 45627 _POSIX_DEVICE_IO _SC_DEVICE_IO 45628 _POSIX_DEVICE_SPECIFIC _SC_DEVICE_SPECIFIC 45629 _POSIX_DEVICE_SPECIFIC_R _SC_DEVICE_SPECIFIC_R 45630 _POSIX_FD_MGMT _SC_FD_MGMT 45631 _POSIX_FIFO _SC_FIFO 45632 _POSIX_FILE_ATTRIBUTES _SC_FILE_ATTRIBUTES 45634 _POSIX_FILE_SYSTEM _SC_FILE_SYSTEM 45644 _POSIX_NETWORKING _SC_NETWORKING 45645 _POSIX_PIPE _SC_PIPE 45655 _POSIX_SIGNALS _SC_SIGNALS 45656 _POSIX_SINGLE_PROCESS _SC_SINGLE_PROCESS 45661 _POSIX_SYSTEM_DATABASE _SC_SYSTEM_DATABASE 45662 _POSIX_SYSTEM_DATABASE_R _SC_SYSTEM_DATABASE_R 45682 _POSIX_USER_GROUPS _SC_USER_GROUPS 45683 _POSIX_USER_GROUPS_R _SC_USER_GROUPS_R Action: Remove the symbols as noted above. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 74 curtis.smith@simtrol.com Bug in XSHd6 system() rationale (rdvk# 48) {cms.system} Tue, 15 May 2001 18:39:44 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 1978 Line: 45958 Section: system() Problem: (Draft 5 ERN reference: NA ) typo Action: Change "statu" to "status" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 75 curtis.smith@simtrol.com Bug in XSHd6 tcsendbreak description (rdvk# 24) {cms-brk} Tue, 1 May 2001 21:31:07 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation for other issues to look at _____________________________________________________________________________ Page: 1999 Line: 46491 Section: tcsendbreak Problem: (Draft 5 ERN reference: NA ) commas used for decimal points Action: Change "0,25 seconds" to "0.25 seconds"; Change "0,5 seconds" to "0.5 seconds" [Editors recommendation: none. Also raised in email on the reflector. Since I brought it up, I'll mention places where I've seen the comma as a thousands spearator. If I were the editor, I'd change all these to spaces, but I don't think it's worth filing a formal request they are valid English: XBD p. 320 l. 11484 TMP_MAX at least 10,000 (XSI) XBD p. 324 l. 11611 RAND_MAX at least 32,767 XSH p. 498 l. 1996 # of ns in second (1,000 million) XSH p. 977 l. 17023 1,000,000 XSH p. 1183 l. 23302 10,000 XSH p. 2024 l. 47265 1,000 million XSH p. 2024 l. 47272 1,000 Hertz [should be lowercase "hertz" in my questionable editorial opinion--another instance on line, too] XSH p. 2805 l. 23225 100,000 bytes (may be more???) ] _____________________________________________________________________________ COMMENT Enhancement Request Number 76 muewi@baumanagement.bremen.de BUG in xhs6 (rdvk# 66) muewi-1 Wed, 16 May 2001 15:37:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_____OPEN Rationale for rejected or partial changes: Fwd to Fred Tydeman, default will be to reject if do not hear by June 8. _____________________________________________________________________________ Page: 2015 Line: 46955 Section: tgamma() Problem: While this statement is mathematically correct, a floating point implementation really incurring this problem is highly unusual. Action: Change sentence to: Overflow may appear near negative integers for unusual floating point representations. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 77 gwinn@res.ray.com Bug in XSHd6 time() (rdvk# 74) {JMG6-1a} Sat, 19 May 2001 02:20:53 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2016 Line: 46989 Section: time() Problem: (Draft 5 ERN reference: NA ) For consistency, "since January 1, 1970 0:00 UTC" should be "since the Epoch". This editorial comment replaces JMG6-1, which was filed against the wrong volume. Action: Fix as shown above. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 78 curtis.smith@simtrol.com Bug in XSHd6 errors (rdvk# 22) {cms-a} Tue, 1 May 2001 21:35:33 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2024 Line: 47260 Section: errors Problem: (Draft 5 ERN reference: NA ) missing word "fail" Action: change "shall if" to "shall fail if" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 79 curtis.smith@simtrol.com Bug in XSHd6 times (rdvk# 39) {cms.times} Fri, 11 May 2001 16:57:50 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Insert before line 47355 /* this examples assumes that clock_t is an integer type */ change 47539 printf(msg) to fputs(msg,stdout); make the change below in the action _____________________________________________________________________________ Page: 2027 Line: 47360 Section: times Problem: (Draft 5 ERN reference: NA ) The example prints out the differnece of two clock_t values using a long format, but that may not be correct in some implementations. Action: Change printf("Real Time: %ld, User Time %ld, System Time %ld\n", en_time - st_time, en_cpu.tms_utime - st_cpu.tms_utime, en_cpu.tms_stime - st_cpu.tms_stime); to printf("Real Time: %jd, User Time %jd, System Time %jd\n", (intmax_t)(en_time - st_time), (intmax_t)(en_cpu.tms_utime - st_cpu.tms_utime), (intmax_t)(en_cpu.tms_stime - st_cpu.tms_stime) ); _____________________________________________________________________________ EDITORIAL Enhancement Request Number 80 Jens.Schweikhardt@marconi.com BUG in XSHd6 (rdvk# 8) [Schweikhardt] Thu, 19 Apr 2001 08:54:16 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2092 Line: 49194 Section: RATIONALE Problem: Typo suspected: processs Action: Change "processs" to "process's" or maybe to "process'", you are the expert... [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 81 Jon Hitchcock Bugs in XSHd6 (rdvk# 70) {jjh59} Fri, 18 May 2001 10:57:21 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2171-2200 Line: 20-1507 Section: index Problem: Many functions (for example _Exit()) have entries in the index both with and without the pair of brackets. Action: Standardize on one form. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 82 Jon Hitchcock Bugs in XSHd6 (rdvk# 71) {jjh60} Fri, 18 May 2001 10:57:21 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2172 Line: 67 Section: index Problem: XSH ERN 12 asked that _POSIX_SOURCE be mentioned in the text and in the index. It has been added at line 533 on page 464, but it is not in the index. Action: Add an entry to the index for _POSIX_SOURCE. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 83 gwc@unisoft.com BUG in XSHd6 write (rdvk# 12) [gwc write EFBIG] Fri, 20 Apr 2001 17:59:22 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2159 Line: 51220 Section: write Problem: The requested action in XSHd5 ERN 1450 was not clear about what to do with the shading on the EFBIG description. (Mea culpa). The intention was that the added text would not be shaded. Action: Remove the shading from "and there was no room for any bytes to be written".