Document Number: AUSTIN/86 Title: XCUd6 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 XCU Draft 6. Aardvark Summary Table ______________________ ERN 1 Accept as marked ERN 2 Accept as marked ERN 3 Accept as marked ERN 4 Reject ERN 5 Accept as marked ERN 6 Duplicate of 5 ERN 7 Accept ERN 8 Accept ERN 9 Duplicate of 8 ERN 10 Accept as marked ERN 11 Accept ERN 12 Reject ERN 13 Reject ERN 14 Reject ERN 15 Accept as marked ERN 16 Accept ERN 17 Accept ERN 18 Reject ERN 19 Reject ERN 20 Reject ERN 21 Accept as marked ERN 22 Reject ERN 23 Accept as marked ERN 24 Duplicate of 23 ERN 25 Accept ERN 26 Accept ERN 27 Reject ERN 28 Reject ERN 29 Accept ERN 30 Accept as marked ERN 31 Reject ERN 32 Reject ERN 33 Accept ERN 34 Accept ERN 35 Accept as marked ERN 36 Reject ERN 37 Accept ERN 38 Accept ERN 39 Accept ERN 40 Reject ERN 41 Accept ERN 42 Reject ERN 43 Duplicate of 42 ERN 44 Reject ERN 45 Reject ERN 46 Reject ERN 47 Accept ERN 48 OPEN ERN 49 Reject ERN 50 Accept ERN 51 Accept ERN 52 Reject ERN 53 Accept as marked ERN 54 Accept ERN 55 Accept as marked ERN 56 Accept as marked ERN 57 Accept ERN 58 Accept as marked ERN 59 Accept as marked ERN 60 Reject ERN 61 Accept ERN 62 OPEN ERN 63 Accept ERN 64 Reject ERN 65 Accept ERN 66 Reject 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 Reject ERN 76 Reject ERN 77 Accept ERN 78 Reject ERN 79 Accept ERN 80 Accept ERN 81 Duplicate of 80 ERN 82 Reject ERN 83 Accept as marked ERN 84 Accept as marked ERN 85 Reject ERN 86 Reject ERN 87 Accept ERN 88 Accept as marked ERN 89 Accept ERN 90 Accept as marked ERN 91 Reject ERN 92 Reject ERN 93 Accept as marked ERN 94 Accept ERN 95 Reject ERN 96 Accept ERN 97 Accept as marked ERN 98 Accept as marked ERN 99 Reject ERN 100 Reject ERN 101 Accept as marked ERN 102 Accept ERN 103 Accept ERN 104 Reject ERN 105 Accept ERN 106 Accept ERN 107 Reject ERN 108 Accept as marked ERN 109 Reject ERN 110 Accept ERN 111 Accept ERN 112 Reject ERN 113 Reject ERN 114 Accept ERN 115 Reject ERN 116 Accept as marked ERN 117 Accept as marked ERN 118 Accept ERN 119 Reject ERN 120 Accept as marked ERN 121 Accept ERN 122 Accept as marked ERN 123 Reject ERN 124 Accept ERN 125 Accept ERN 126 Accept as marked ERN 127 Duplicate of 128 ERN 128 Accept _____________________________________________________________________________ COMMENT Enhancement Request Number 1 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 13) [Schweikhardt] Fri, 20 Apr 2001 08:53:46 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: In line 512 Change "Keyword" to "Preprocessing Directive" and add an entry for keyword: "C-Language Keyword return" _____________________________________________________________________________ Page: xiv Line: 512 Section: Typographical Problem: The table says "C-Language Keyword #define" ISO 9899 enumerates all C keywords (auto ... while). #define is not one of them, it's a preprocessing directive. We should use the ISO terms for consistency. Action: In line 512 Change "Keyword" to "Preprocessing Directive" and if we really want an entry for keyword, add after that line "C-Language Keyword return" [Ed recommendation: None We need a term that covers its use in the document, so "Preprocessing directive and keywords" might be better, and we could add "union" to the example to make it clearer. ] _____________________________________________________________________________ OBJECTION Enhancement Request Number 2 gwc@unisoft.com BUG in XCUd6 1.7.1.4 (rdvk# 29) [gwc table 1-1] Wed, 16 May 2001 18:15:13 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Accept the proposed changes below. All to check whether these changes will cause you to object, if so why, respond by June 8. _____________________________________________________________________________ Page: 2204,2205 Line: 133-169 Section: 1.7.1.4 Problem: Adding a table to cover all the combinations of file types (table 1-1) was a good idea. Unfortunately many of the details in the table, and the codes it uses, are currently incorrect. The actions specified for some of the combinations that were already covered by the draft 5 text are different (from draft 5), and some of the newly covered combinations have actions that are not those that would be expected based on the behaviour of the functions in XSH that create the "new" file type. These are the problems I have spotted: 1. The row for existing type "None" does not belong here. This part of section 1.7.1.4 is concerned with attempts to create a file when a file with that name already exists. The requirements for the case where no file exists and a command creates a new file are already stated elsewhere. The CF code is just a reference to one of those places (which happens to be the first part of the same section). The NF code is worse because it bypasses those places and refers directly to the functions in XSH instead. 2. The mkdir(), mkfifo(), mknod() and symlink() functions always fail if the named file already exists, so all the "-"s in the B, C, D, F and L columns should be "F". They also fail if a symlink exists, so the "FL" entries in the B, C, D, F and L columns should be changed to "F". (Note that the command "ln -s file existing_dir" does not result in an attempt to create a symlink with the name "existing_dir", so the row D column L entry does not apply to this case. It applies to things like pax extracting symlinks from an archive.) 3. open() with O_CREAT fails when a directory exists, it does not open the directory, so the "OD" in row D column R should be "F". (This was item 3 in the draft 5 text.) 4. Since the behaviour of shm_open(), sem_open() and mq_open() when O_CREAT is set and a symlink exists is unspecified, I don't think that there is any justification for the table to indicate that the symlink is followed. Thus the "FL"s in columns M, P and Q should be "-". For typed memory, the method used to create these files is implementation- defined, so whether or not it would follow symlinks is also implementation-defined, and the "FL" in column T should also be "-". 5. For sockets, the behaviour of bind() when a file exists in the location given by sockaddr_un.sun_path[] is unspecified, so again I think that column S should be entirely "-"s. 6. fattach() attaches a STREAM to an *existing* file, it does not create a new file. So files of type A are never "created" as such, and therefore column A does not belong in the table. The entry for row A under "Function Creating New" should also be updated to reflect this. (I suggest using "N/A" here, as it is already in use elsewhere in the document.) Once the table has been updated to correct these problems, there are some consequential changes needed to the code descriptions. I believe that once the changes requested below have been applied, the requirements specified in the new table and code descriptions will be the same as in the old draft 5 text for the cases that the old text covered. For the cases that the old text did not cover, I hope that these changes mean that the requirements are now reasonable and do not make any unjustified assumptions. Finally, all of this stuff needs to be covered by an "unless specified otherwise" clause, just as for the no-existing-file case (see first paragraph of section 1.7.1.4). I have suggested a rewording of lines 133-135 which does this. Action: Replace lines 133-135 with: When an attempt is made to create a file that already exists, the utility shall take the action indicated in Table 1-1 (on page 2205) corresponding to the type of the file the utility is trying to create and the type of the existing file, unless the utility description states otherwise. Replace the contents of table 1-1 with the following: _______________________________________________________________________ New Type Function Existing Type B C D F L M P Q R S T Creating New _______________________________________________________________________ A fattach()-ed STREAM F F F F F - - - OF - - N/A B Block Special F F F F F - - - OF - - mknod()** C Character Special F F F F F - - - OF - - mknod()** D Directory F F F F F - - - F - - mkdir() F FIFO Special File F F F F F - - - O - - mkfifo() L Symbolic Link F F F F F - - - FL - - symlink() M Shared Memory F F F F F - - - - - - shm_open() P Semaphore F F F F F - - - - - - sem_open() Q Message Queue F F F F F - - - - - - mq_open() R Regular File F F F F F - - - RF - - open() S Socket F F F F F - - - - - - bind() T Typed Memory F F F F F - - - - - - * _______________________________________________________________________ Delete the descriptions of the (now unused) CF, NF and OD codes, on lines 153, 162 and 169. Replace the description of code F on lines 154-157 with the following: F Fail. The attempt to create the new file shall fail and the utility shall either continue with its operation or exit immediately with a non-zero exit status, depending on the description of the utility. On line 158 change "symbolic links" to "symbolic link". _____________________________________________________________________________ COMMENT Enhancement Request Number 3 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 30) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: This text summarises the reqts for rmdir() and unlink(). It is possible using the unlink() function for item 4 to occur. We will add rationale to make this clearer (Cathy has notes). _____________________________________________________________________________ Page: 2206 Line: 196 Section: System Problem: Unlike the rmdir() function, trying to delete a non-empty directory is supposed to update st_ctime. Action: Change item 4 to: If the file is a directory that is not empty, the requested operation shall fail. _______________________________________________________________________________________ OBJECTION Enhancement Request Number 4 eggert@twinsun.com Bug in XCUd6 1.7.2.1 Arithmetic Precision and Operations (rdvk# 17) {eggert-2001050301} Thu, 3 May 2001 23:02:23 +0100 (BST) _______________________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Despite the change bars this wording is unchanged from what appeared in POSIX.2 . An interpretation has been filed, if the response indicates that is appropriate to make a change, it can be made in corrigenda. _______________________________________________________________________________________ Page: 2207 Line: 254 Section: 1.7.2.1 Problem: (Draft 5 ERN reference: unknown ) The text as stated seems to require the support for ISO C syntax for constants, e.g. integer suffixes like 'L' or hexadecimal floating-point constants. Surely this is not intended. It would require, for example that 'diff -C 0L a b' ignore the 'L', and no 'diff' implementation that I know of does this. Also, the text says that "Integer variables and constants ... shall be implemented as equivalent to the ISO C standard signed long data type". This confuses the notions of variables and constants with the notion of data types. There is similar confusion with the wording used for floating point. Action: Replace the sentence in XCUd6 lines 254-257 with: Integer arithmetic used by the standard utilities listed in this volume of IEEE Std 1003.1-200x, including the values of operands and option-arguments, shall be implemented in conformace to the ISO C requirements for the signed long data type. Similarly, floating point arithmetic shall be implemented in conformance to the ISO C requirements for the double type. _____________________________________________________________________________ COMMENT Enhancement Request Number 5 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 14) [Schweikhardt] Fri, 20 Apr 2001 08:53:46 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: In line 260 change "functions" to "control flow keywords" In line 262 Change "ISO C Standard Operators and Functions" to "Selected ISO C Standard Operators and Control Flow Keywords" _____________________________________________________________________________ Page: 2209 Line: 262 Section: Table Problem: The table heading says "ISO C Standard Operators and Functions" but there's not a single function mentioned in the table (but some control flow keywords.) The list of operators is also incomplete: postfix -> and . and function call and the cast operator are not listed. I assume they have been left out for a reason. Action: In line 262 Change "ISO C Standard Operators and Functions" to "Selected ISO C Standard Operators and Control Flow Keywords" _____________________________________________________________________________ COMMENT Enhancement Request Number 6 muewi@baumanagement.bremen.de BUG in XCUd6 (rdvk# 121) [muewi-4] Fri, 18 May 2001 23:11:50 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_5 Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2209,2243 Line: 294-303,1707 Section: 1.7.2.1 Problem: C language flow control statements are neither functions nor operators and are not required for anything in this volume (besides c99, of course). Action: Remove these lines. [Ed recommendation: None They are referenced in the document] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 7 ajosey@opengroup.org Bug in XCUd6 1.11 (rdvk# 1) {aj.apr19.2} Thu, 19 Apr 2001 10:57:08 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2225 Line: 994 Section: 1.11 Problem: (Draft 5 ERN reference: XCUd5 ERN 224 ) The text in STDERR for individual utilities was updated from "Used only for diagnostic messages" to "The standard error shall be used only for diagnostic messages." but this front matter was not. We need to make the change here also. Action: Change "Used only for diagnostic messages" to "The standard error shall be used only for diagnostic messages." [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 8 ajosey@opengroup.org Bug in XCUd6 1.11 (rdvk# 114) {aj.1.11} Fri, 18 May 2001 11:17:47 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2228 Line: 1115 Section: 1.11 Problem: (Draft 5 ERN reference: NA ) This section should be consistent with XSH Action: Replace lines 1115-1117 with XSH p462, lines 473-474 [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 9 Jon Hitchcock Bug in XCUd6 CHANGE HISTORY (rdvk# 116) {jjh56} Fri, 18 May 2001 10:57:02 +0100 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_8 Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2228 Line: 1116 Section: CHANGE Problem: The introduction claims that the change history subsections list the functional differences between Issues 4 and 6, but in fact they only list changes since Issue 4, Version 2. Action: Change "Issues 4 and 6" to "Issue 4, Version 2 and Issue 6". [Ed recommendation: Dup of 8 ] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 10 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 31) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2232 Line: 1240-1241 Section: Double-Quotes Problem: The subject is "The input characters..." (plural), so the predicate should be plural, too. (Unless we speak ancient Greek and the subject is neuter.) Action: Change "is not" to "are not". [Ed recommendation: Accept as marked shallification Change "is not affected by" to "shall not be affected by" ] _____________________________________________________________________________ COMMENT Enhancement Request Number 11 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 32) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: [Notes Staying compatible with the Bourne Shell was not in the scope of the project, however, the grammar was updated in this revision to allow for x do ..... since linebreak can be empty So the change is needed (to complete the change in the grammar) Do the change below in Action] _____________________________________________________________________________ Page: 2235 Line: 1353 Section: Reserved Problem: To stay compatible with the historical Bourne shell, the third word in a for command may be "do". Action: Split item into two new items: \item The third word in a "case" command (only "in" is valid in this case) \item The third word in a "for" command (only "in" and "do" are valid in this case) _____________________________________________________________________________ COMMENT Enhancement Request Number 12 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 33) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: See the time utility, p3113-3115 _____________________________________________________________________________ Page: 2235 Line: 1357 Section: Reserved Problem: Implementations which support timing of complete pipelines (e.g., bash), may have "time" as a reserved word, too. Action: Add "time" to list of possibly reserved words, maybe with a note that it still has to conform to semantics of the time utility. _____________________________________________________________________________ COMMENT Enhancement Request Number 13 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 34) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This text is from a base document, labels are reserved for future use, quoting is unspecified at present, changing it may reduce consensus _____________________________________________________________________________ Page: 2235 Line: 1358 Section: Reserved Problem: A quoted colon should be a character as any other one. Action: Change to ...a name and an unquoted colon... _____________________________________________________________________________ COMMENT Enhancement Request Number 14 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 35) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This is presently unspecified, to change this may reduce consensus _____________________________________________________________________________ Page: 2235 Line: 1366-1367 Section: Positional Problem: What does ${00} mean? Action: Add a paragraph after this one: ${00} (and similar parameter names) may either be identical to $0 or constitute an error. (Rationale: ksh and bash use $0, a Bourne shell I've access to produces an error. Any other result would be unexpected and irritating.) _____________________________________________________________________________ COMMENT Enhancement Request Number 15 dwc@spartan.eng.sun.com Bug in XCUd6 (rdvk# 125) [DWC-3] Mon, 21 May 2001 23:28:00 -0700 (PDT) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: we rejected 14 _____________________________________________________________________________ Page: 2235 Line: 1366-1367 Section: 2.5.1 Problem: (Meaning of ${00}) I disagree with the change to these lines proposed by Wilhelm Mueller. The second sentence in this paragraph: The digits denoting the positional parameters shall always be interpreted as a decimal value, even if there is a leading zero. clearly requires that ${00} and $0 be treated the same. Making the change Wilhelm suggests could break existing strictly conforming applications. If Wilhelm wants to allow ${00} to generate an error (as it was in some versions of the Bourne shell), he needs to file an interpretation request to see if the interpretations committee believes that this current standard specification of historic ksh behavior was intended to be required. Action: Do not add the paragraph Wilhelm suggests that would allow ${00} to be treated as an error. ------------------------------------------------------------------------------ _____________________________________________________________________________ EDITORIAL Enhancement Request Number 16 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 36) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2236 Line: 1396 Section: Special Problem: Two commas are missing. Action: Following the ')', the sentence should read: as specified on invocation, by the set special builtin command, or implicitly... _____________________________________________________________________________ COMMENT Enhancement Request Number 17 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 37) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2237 Line: 1438 Section: Shell Problem: LC_ALL is not the default value for the LC_* variables---that is LANG. LC_ALL is an override. Action: Change description up to first comma to: The value of this variable overrides the LC_* variables and LANG,... _____________________________________________________________________________ EDITORIAL Enhancement Request Number 18 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 38) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2237 Line: 1450 Section: Shell Problem: LC_MESSAGES is part of XSI Action: Add XSI shading and marginal. [Ed recommendation: Reject LC_MESSAGES is in 1003.2-1992] _____________________________________________________________________________ COMMENT Enhancement Request Number 19 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 39) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: The proposed change was not acceptable, and the group was unable to achieve consensus on wording to add this here at this late stage. _____________________________________________________________________________ Page: 2237 Line: 1459-1460 Section: Shell Problem: OLDPWD is missing. Action: Add: OLDPWD value of PWD before last successful cd _____________________________________________________________________________ COMMENT Enhancement Request Number 20 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 40) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: It is out of scope for us to consider this addition. There may be shells that do not support it. _____________________________________________________________________________ Page: 2239 Line: 1522-1538 Section: Tilde Problem: The expansion of ~- to the contents of OLDPWD is quite common and useful. Action: Add a sentence, maybe before "If the login name is null..." (l.1533): If the login name is "-", the tilde prefix is replaced by the value of the variable OLDPWD; if OLDPWD is empty or does not exist, the results are unspecified. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 21 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 41) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change 1589 parameter -> null [3rd col] change 1589 assign null -> assign word [4th col] _____________________________________________________________________________ Page: 2240 Line: 1584-1593 Section: Parameter Problem: In the third column of this table (heading "parameter Set But Null"), "null" is used three times, "parameter" once. Although the result is the same, either one should be used consistently. Action: (Preferably) change the three occurrences of "null" in the third column to "parameter". _____________________________________________________________________________ COMMENT Enhancement Request Number 22 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 42) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The example has an asterix between two quoted strings. _____________________________________________________________________________ Page: 2242 Line: 1655-1656 Section: Parameter Problem: Add another example with nested quotes. Action: Add a line "${x#"*"}" The asterisk is not special, the result of the variable expansion is the same as with ${x#"*"}, but no word splitting is performed on the result of the substitution, _____________________________________________________________________________ EDITORIAL Enhancement Request Number 23 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 43) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Join l 1680 into line 1679. delete l 1681. _____________________________________________________________________________ Page: 2242 Line: 1680-1681,1689-1692 Section: Command Problem: The second item refers to a quite short paragraph a few lines further down. Action: Replace text of second item with paragraph l1689-1692 _____________________________________________________________________________ COMMENT Enhancement Request Number 24 dwc@spartan.eng.sun.com Bug in XCUd6 (rdvk# 126) [DWC-4] Mon, 21 May 2001 23:28:00 -0700 (PDT) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_23 Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2242 Line: 1680-1681,1689-1692 Section: 2.6.3 Problem: (Command Substitution change request) I disagree with the change to these lines proposed by WIlhelm Mueller. This section is talking about valid shell constructs that can't used inside a command substitution construct. The text Wilhelm wants to move into this list does not represent a valid shell construct (when it is not being used as an arithmetic expansion). Action: Do not make the change to P2242, L1681 Wilhelm suggests. ------------------------------------------------------------------------------ _____________________________________________________________________________ EDITORIAL Enhancement Request Number 25 ajosey@opengroup.org Bug in XCUd6 2.6.4 (rdvk# 2) {aj.apr19.1} Thu, 19 Apr 2001 08:37:27 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2243 Line: 1705 Section: 2.6.4 Problem: (Draft 5 ERN reference: NA ) New text uses wrong terminology Action: Change "Subclause" to "section" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 26 gwc@unisoft.com BUG in XCUd6 (rdvk# 119) [gwc muewi-1 response] Fri, 18 May 2001 17:58:27 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2243 Line: 1706,1707 Section: Arithmetic Problem: The following changes requested by Wilhelm Mueller are incorrect. > Problem: > The cast is not listed as not required. > > Action: > Change to: > The sizeof() operator, type casting, and the prefix... > Problem: > C language statements aren't supported at all. What was probably meant > is that the ?: operator doesn't have to be supported. > > Action: > Change to > Support of the ?: operator pair is not required. Lines 1703-1707 list exceptions to the rules in 1.7.2.1. Those rules already do not require casts to be supported, so listing non-support of casts as an exception would be wrong. The table in 1.7.2.1 does include selection, iteration and jump statements, and that is why they are listed here as not required. And just for completeness: support of the ?: operator is required and removing it would be out of scope. Action: Do not make the changes requested by Wilhelm Mueller that are quoted above. _____________________________________________________________________________ COMMENT Enhancement Request Number 27 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 44) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: See ERN 26 _____________________________________________________________________________ Page: 2243 Line: 1706 Section: Arithmetic Problem: The cast is not listed as not required. Action: Change to: The sizeof() operator, type casting, and the prefix... _____________________________________________________________________________ COMMENT Enhancement Request Number 28 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 45) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: See ERN 26 _____________________________________________________________________________ Page: 2243 Line: 1707 Section: Arithmetic Problem: C language statements aren't supported at all. What was probably meant is that the ?: operator doesn't have to be supported. Action: Change to Support of the ?: operator pair is not required. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 29 dwc@spartan.eng.sun.com Bug in XCUd6 (rdvk# 127) [DWC-5] Mon, 21 May 2001 23:28:00 -0700 (PDT) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2244 Line: 1749 Section: 2.6.7 Problem: (Quote Removal quote characters) The backslash character is denoted by the C character constant '\\'. Action: Change "'\'" on P2244, L1749 to "'\\'". ------------------------------------------------------------------------------ _____________________________________________________________________________ COMMENT Enhancement Request Number 30 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 46) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add after "is closed." on line 1854,1862 Attempts to close a file descriptor that is not open shall not constitute an error. _____________________________________________________________________________ Page: 2246 Line: 1853-1854 Section: Duplicating Problem: What happens when a file descriptor already closed is to be closed again? Since the close() call fails, a redirection error might be expected. But all (Bourne like) shells I found allowed to close any descriptor, no matter what their status was. Action: Here and also at page 2247 line 1861-1862 section Duplicating an Output File Descriptor comment Add a sentence: Closing a file descriptor already closed (or never been opened) does not constitute an error. _____________________________________________________________________________ COMMENT Enhancement Request Number 31 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 47) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The change is out of scope. If the submitter feels strongly that this wording should be added they should submit an interpretation request. _____________________________________________________________________________ Page: 2247 Line: 1865-1869 Section: Open Problem: Are the following redirections valid? echo text 3<>f 4>&3 5<&3 6<&4 1>&5 (It seems they are, as all shells I tried just dup()ed the fd without checking if it will be used for reading or writing.) Action: Add a sentence: If a file descriptor open for reading and writing is duplicated using the <& or >& operators, the duplicate will still be open for reading and writing. (Rationale: otherwise, an operator <>& would become necessary, too.) _____________________________________________________________________________ EDITORIAL Enhancement Request Number 32 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 48) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The wording is per 1003.2 and the editors are happy with the current wording. _____________________________________________________________________________ Page: 2248 Line: 1901-1906 Section: Exit Problem: The first sentence looks very strange, grammatically. Action: Change to sth. like ... the shell shall retrieve the exit status by the equivalent of calling the wait() function and then applying the WEXITSTATUS() macro to its parameter. [Ed recommendation: None wording as per 1003.2] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 33 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 49) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2253 Line: 2120 Section: Compound Problem: The list of grouping command uses "compound-list", this line just "list". Action: Change "list" to "compound-list". _____________________________________________________________________________ EDITORIAL Enhancement Request Number 34 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 50) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2255 Line: 2169-2176 Section: Compound Problem: There's no "fi". Action: Add a line containing just "fi" following l.2176 [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 35 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 51) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: 6. [Third word of for and case] 6a. [case only] When the TOKEN is exactly the reserved word "in", the token identifier for "in" shall result. Otherwise, the token WORD shall be returned. 6b. [for only] When the TOKEN is exactly the reserved word "in" or "do", the token identifier for "in" or "do" shall result, respectively. Otherwise, the token WORD shall be returned. (As indicated in the grammar, a linebreak precedes the tokens "in" and "do". If s are present at the indicated location, it is the token after them that is treated in this fashion.) 2447 /* Apply rule 6 */ _____________________________________________________________________________ Page: 2258 Line: 2296-2300,2400 Section: Shell Problem: Due to historical practice, the reserved word "do" has to be treated like "in", but only after "for" (and "select", but that's out of scope). Action: Split 6. into two points, 6a being the current one, 6b being the same for "do" instead of "in" and only applying to "for" (not "case"). On line 2400, the comment has to be changed to "Apply rule 6a". _____________________________________________________________________________ COMMENT Enhancement Request Number 36 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 52) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope to change the grammar in such a way _____________________________________________________________________________ Page: 2260 Line: 2394,2447 Section: Shell Problem: While the Bourne shell did not allow "for x; do" (but did allow "for x do" or "for x do", modern shells have no problem with this, they allow all of these forms. Action: Split line 2394 into two lines: for_clause: For name do_group /* Apply rule 6b */ | For name sequential_sep do_group _____________________________________________________________________________ EDITORIAL Enhancement Request Number 37 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 53) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2260 Line: 2395 Section: Shell Problem: The last word on that line is "sequential_sep_do_group". Action: Change to: sequential_sep do_group [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 38 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 54) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2260 Line: 2415,2417 Section: Shell Problem: The consecutive occurrence of "linebreak", though not actually wrong ("linebreak" may expand to nothing) is redundant and, if actually used for a parser generator, would produce funny shift-reduce conflicts. Action: On both lines, delete the second "linebreak". _____________________________________________________________________________ EDITORIAL Enhancement Request Number 39 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 55) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2260-2261 Line: 2421-2423 Section: Shell Problem: Preceding the "linebreak", there should be a "DSEMI". Action: On each of these three lines, Insert "DSEMI" before last (or only) "linebreak". _____________________________________________________________________________ COMMENT Enhancement Request Number 40 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 56) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: This is out of scope to make this change. _____________________________________________________________________________ Page: 2261 Line: 2417,2418,2422,2423,2425 Section: Problem: Rule 4 does not apply if the parser needs the pattern for '(' pattern ')' (in lines 2417, 2418, 2422, 2423). Action: Insert new rule: ppattern: WORD /* Do not apply rule 4 */ | pattern '|' WORD ; and change the other four lines to start with "'(' ppattern ')'" _____________________________________________________________________________ EDITORIAL Enhancement Request Number 41 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 57) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2261 Line: 2426 Section: Shell Problem: All other comments referring to a rule just use the rule number, this one puts parentheses around it. Action: Change '(4)' to '4' for consistency. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 42 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 58) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: An interpretation will be filed against the base document. At the present time this change is out of scope. We think that a Change to "Then compound_list else_part" might be correct. It also appears that there needs to be an empty else_part substitution which has influence on the if_clause. _____________________________________________________________________________ Page: 2261 Line: 2431 Section: Shell Problem: There seems to be not very much allowed following that "Then". Action: Change to "Then compound_list else_part Fi". _____________________________________________________________________________ COMMENT Enhancement Request Number 43 dwc@spartan.eng.sun.com Bug in XCUd6 (rdvk# 128) [DWC-6] Mon, 21 May 2001 23:28:00 -0700 (PDT) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_42 Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2261 Line: 2431 Section: 2.10.2 Problem: (Shell Grammar Rules) I strongly disagree with the change to this line proposed by Wilhelm Mueller. If the change Wilhelm proposes making to P2261, L2431 is made, the following sequence: if cmd1 then cmd2 elif cmd3 then cmd4 fi fi becomes a valid if_clause. (The second "fi" is not a typo.) This comes from plugging the modified else-part on P2261, L2431 into the if_clause on P2261, L2428. Action: Do not make the change Wilhelm suggests. _____________________________________________________________________________ COMMENT Enhancement Request Number 44 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 59) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This change seems unnecessary. If the submitter wishes to pursue this an interpretation request should be filed. It is presently out of scope. _____________________________________________________________________________ Page: 2264 Line: 2585-2591 Section: Patterns Problem: Even when pattern matching is used where shell quote removal is performed (such as in the argument to the find name primary when find is being typed in by the user), special characters can be escaped to remove their special meaning, but some kind of shell quoting will have to be applied, too. Action: Delete everything preceding "special characters can be escaped..." (and capitalize "special"). Add a sentence: "When using patterns with quoted special characters when shell expansions and quote removal will be performed, such as in interactive commands or shell scripts, two levels of quoting will be necessary, one to prohibit the shell from expanding special characters and remove quoting characters, one to escape the special character itself. For example, to find file names containing a (literal) '*' (and not starting with '.'), one of these can be used: find . -name \*\\\*\* find . -name '*\**' find . -name "*\**" find . -name "*\\**" _____________________________________________________________________________ COMMENT Enhancement Request Number 45 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 60) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The man page clearly states that there is no output on stderr _____________________________________________________________________________ Page: 2267 Line: 2680 Section: break Problem: It should be explicitly stated that break does never print anything to STDERR, even when n is not a number at all. Action: Here and also at page 2271 line 2797 section continue comment Add "Not even for n not being a number". _____________________________________________________________________________ COMMENT Enhancement Request Number 46 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 61) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject__X__ Rationale for rejected or partial changes: This is unnecessary -- See RATIONALE _____________________________________________________________________________ Page: 2268 Line: 2712 Section: break Problem: Since names followed by a ':' have been reserved, labels may be mentioned as a future direction. Action: Here and also at page 2272 line 2820 section continue comment Add "Labeled commands and breaking/continuing a labeled command may be introduced." _____________________________________________________________________________ COMMENT Enhancement Request Number 47 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 62) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2273 Line: 2851 Section: dot Problem: The description specifies diagnostic messages. Action: Change "None." to "The standard error shall be used only for diagnostic messages." [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 48 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 63) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The standard developers felt it unnecessary to elucidate further. _____________________________________________________________________________ Page: 2274 Line: 2873-2874 Section: dot Problem: Passing parameters to dot scripts differs somewhat from passing them to functions. Action: It should be mentioned that, as with shell functions, these optional parameters must be sort of local to the dot script and that, as opposed to real functions where "$@" &c. become empty when the function is called without parameters, when no parameters are specified in the dot command, "$@" &c. of the caller must be available to the dot script. _____________________________________________________________________________ COMMENT Enhancement Request Number 49 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 64) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: It is out of scope to add this. _____________________________________________________________________________ Page: 2281 Line: 3054-3092 Section: export Problem: There is no simple way to remove the export attribute from a name without unsetting it completely. (No, writing "var= cmd" is not the same: it puts var with an empty value in the environment of cmd.) Action: At least under `future directions', add an option -u (or -n as in bash) for unexport. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 50 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 65) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2287 Line: 3254 Section: set Problem: There should be a space between "--" and "[argument...]" Action: Add space. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 51 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 66) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2287 Line: 3262-3264 Section: set Problem: The reference to shell quoting appears too late---context has just switched to read only variables. Action: Reorder, e.g. thus: The value string shall be written with appropriate quoting; see the description of shell quoting in Section 2.2 (on page xxxx). The output shall be suitable for re-input to the shell, setting or resetting, as far as possible, the variables that are currently set; read-only variables cannot be reset. _____________________________________________________________________________ COMMENT Enhancement Request Number 52 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 67) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Thanks for the comment. There is a need for an interpretation as +n appears in the synopsis. _____________________________________________________________________________ Page: 2288 Line: 3323-3324,3341 Section: set Problem: What does "set +n" (or "set +o noexec") do? Action: Either add that behaviour is unspecified or require that implementations which do not ignore "set -n" in interactive shells shall support it to turn of check only mode. _____________________________________________________________________________ COMMENT Enhancement Request Number 53 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 68) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change "unless" on L3358 to "unless stated otherwise in the description of the option or unless". add to CH PASC Interpretation 1003.2 #167 is applied, clarifying that the options default also takes into account the description of the option. _____________________________________________________________________________ Page: 2289 Line: 3358-3359 Section: set Problem: The default for -m is on in interactive shells. Action: Change to: The default for all these options, with the exception of -m in interactive shells, shall be off unless the shell was invoked with them on. The default for -m shall be on in interactive shells unless the shell was invoked with it explicitly turned off. See sh (on page xxx). _____________________________________________________________________________ EDITORIAL Enhancement Request Number 54 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 69) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2289 Line: 3365 Section: set Problem: A space is missing in "set--". Action: Change to "set --". [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 55 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 70) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2290 Line: 3402 Section: set Problem: The example talks about a variable $x, but uses an option -x. Action: Change to: Set $1 to the value of the variable x, even if it begins with '-' or '+': [Ed recommendation: Accept as marked Set $1 to the value of x, even if it begins with '-' or '+': ] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 56 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 71) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2291 Line: 3433 Section: set Problem: Here's an "x-" which does not belong there. Action: Change to set -x [Ed recommendation: Accept as marked Change to set -k ] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 57 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 72) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2297 Line: 3597 Section: trap Problem: The two " characters look a bit far apart--at first, I read " ". Action: Change font or move second " to the left to decrease amount of visible white space. _____________________________________________________________________________ COMMENT Enhancement Request Number 58 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 74) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change l 3605 commencing "Implementations..." Implementations may permit names with the SIG prefix or ignore case in signal names as an extension. _____________________________________________________________________________ Page: 2297 Line: 3605 Section: trap Problem: trap may permit lowercase signal names, kill must ignore case completely. Action: For sake of consistence, either make them accept the same textual representations for signal names: Either make ignoring case optional for kill or require it for trap, too. In any case, change "may permit lowercase signal names" to "may ignore case distinction in signal names". _____________________________________________________________________________ EDITORIAL Enhancement Request Number 59 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 73) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: eval action (italic action) _____________________________________________________________________________ Page: 2297 Line: 3610 Section: trap Problem: "action" is not a shell variable, but a meta-variable of the grammar. Action: Change to eval "action" with the word action in the slanted font used for meta-variables. _____________________________________________________________________________ COMMENT Enhancement Request Number 60 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 75) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The wording is from the base document. If the submitter feels strongly an interpretation request should be filed. _____________________________________________________________________________ Page: 2297 Line: 3620-3624 Section: trap Problem: The intended behaviour described here has never been implemented by any shell. Consider code like this: f () { save=$(trap) trap 'do_recover' INT do_something eval "$save" } trap - INT f trap There's an additional trap handler showing up now... Action: Add after "...trapping results": ...for traps caught or ignored at the time the output was generated. Traps with a default action may not be reset to the default action when handlers have been established between the time the output was generated and the time it is used to re-establish the trap handlers. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 61 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 76) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2300 Line: 3721 Section: unset Problem: The two " look a bit too far apart (like " "). Action: Change font or spacing to produce less visible white space. _____________________________________________________________________________ OBJECTION Enhancement Request Number 62 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 18) [DST-10] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_____OPEN Rationale for rejected or partial changes: refer to Base WG with the following recommendation: Check the structure of the SCCS file and compare the newly computed checksum with the checksum that is stored in the SCCS file. If the newly computed checksum does not match the checksum in the SCCS file, a diagnostic message shall be written. Base WG to respond by June 8 _____________________________________________________________________________ Page: 2330 Line: 4848 Section: admin Problem: The format of a SCCS file is unspecified. Here you specify a small piece of it, by saying that there's a line (at all). It might be that the file HAS no lines (but contains them in some other representation, such as (e.g.) counted strings). Unspecified is unspecified. Action: Remove the parenthetical remark. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 63 ajosey@opengroup.org Bug in XCUd6 ar (rdvk# 115) {aj.ar.shade} Fri, 18 May 2001 06:20:16 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2336 Line: 5025 Section: ar Problem: (Draft 5 ERN reference: NA ) The shading in the synopsis is incorrect (due to the switch to MAN and back at one point), some of the sS's should be coded as sT's? Action: Look at previous XCU5 ar man page to see how the last 5 lines in the SYNOPSIS should be coded Also consider adding at the start of the DESCRIPTION to make it clear: The ar utility is part of the Software Development Utilities option. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 64 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 77) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope. The wording is from a base document and if the submitter feels strongly an interpretation request should be filed _____________________________________________________________________________ Page: 2347 Line: 5440-5442 Section: at Problem: Using -q and -t (or timespec) together, the results are unspecified. But the same line say that a missing -q operand is the same as "-q a", so writing "at 3pm" is ok and means "at -q a 3pm", but writing it that way produces unspecified results. And the description of the batch utility simply states that it is identical to "at -m -q b now", which is unspecified, as well. Action: Change to: ... results are unspecified unless the queue name is "a" or "b". _____________________________________________________________________________ EDITORIAL Enhancement Request Number 65 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 78) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2352 Line: 5681-5684 Section: at Problem: An eof indicator is missing. Action: As at other places, insert a line EOT between lines 5682 and 5683 [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 66 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 79) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This is an example and the group did not see the merit to change it given the comments in the example. _____________________________________________________________________________ Page: 2352 Line: 5698-5700 Section: at Problem: Even with a small amount of processing, the start time of the at job will move through the day. Action: To make the amount of this move as small as possible, the at-command resubmitting the job should come first, with an appropriate comment. _____________________________________________________________________________ COMMENT Enhancement Request Number 67 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 80) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2355,2356 Line: 5793-5796,5849-5850 Section: awk Problem: I understand from the description of the -f option that the concatenation of all progfiles must be a valid awk program, but the description at INPUT FILES suggests to me that every progfile must be a valid program itself. Action: Change wording under INPUT FILES: If -f progfile is specified, the application shall ensure that the files named by each of the progfile option-argument are text files and their concatenation, in the same order as they appear in the arguments, is an awk program. _____________________________________________________________________________ COMMENT Enhancement Request Number 68 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 19) [DST-11] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2364 Line: 6158 Section: awk Problem: Since I'm not really happy with the outcome of the "8 bit byte" debate, I'm submitting this as a comment. The discussion of byte sizes greater than 8 bits is no longer relevant. Action: Remove it. (I'll let someone who thinks the 8-bit byte decision is a good idea do that.) _____________________________________________________________________________ COMMENT Enhancement Request Number 69 ajosey@opengroup.org Bug in XCUd6 rationale (rdvk# 123) {aj.p2382.6972} Mon, 21 May 2001 10:25:41 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2382 Line: 6972 Section: rationale Problem: (Draft 5 ERN reference: NA ) The wording is not right for this issue Action: Change "The ISO POSIX-2 standard description" to "This description" Also change line 6993 from IEEE Std 1003.1-200x has made a few relatively minor changes toward making the language more compatible with the C language as specified by the ISO C standard; to " A few relatively minor changes toward making the language more compatible with the ISO C standard were made;" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 70 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 81) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2382 Line: 6974 Section: awk Problem: "functin" Action: should be "function" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 71 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 9) [Schweikhardt] Wed, 18 Apr 2001 14:10:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2382 Line: 6974 Section: 1 Problem: Typo suspected: functin Action: Change "functin" to "function" [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 72 ajosey@opengroup.org Bug in XCUd6 rationale (rdvk# 122) {aj.p2383.7023} Mon, 21 May 2001 10:31:14 +0100 (BST) _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2383 Line: 7023 Section: rationale Problem: (Draft 5 ERN reference: NA ) The wording is not right for this issue Action: Change lines 7023-7024 to: This has been deprecated by the authors of the language, who asked that it not be specified. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 73 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 82) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2394 Line: 7424 Section: batch Problem: The here-document appears on the same line as its delimiters. Action: insert newline after "<From the required options, I would conclude that the system consists conceptually of three parts: preprocessor (invoked alone when -E), compiler, and linker (not invoked at all when -c). Action: Add preprocessor to the conceptional phases. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 76 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 85) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2415 Line: 8292-8294 Section: c99 Problem: TMPDIR must be either XSI shaded or not, but not both. Action: Delete XSI shaded part. [Ed recommendation: None. The non shaded text is a "should". on XSI systems this requirement is a "shall". We may be able to reword. ] _____________________________________________________________________________ COMMENT Enhancement Request Number 77 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 3) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2418 Line: 8393 Section: EXTENDED Problem: The text states: "Using return from main() shall be equivalent (other than with respect to language scope issues) to calling exit() with the returned value." This wording rules out recursive calls to main (e.g. programs wanting to start all over which do not want to use longjmp or similar techniques, or programs calling main with different argc and argv.) The ISO C Standard (unlike C++, I'm told) allows recursive calls to main and uses slightly different wording when describing the effect of returning from main: From C99 5.1.2.2.3, Program Termination "..., a return from the initial call to the main function is equivalent to calling the exit function with the value returned by the main function as its argument; reaching the } that terminates the main function returns a value of 0." This incompatibility with the ISO C Standard shall be removed and I propose the following Action: In lines 8393 and 8395 Change "main" to "the initial call to main" _____________________________________________________________________________ COMMENT Enhancement Request Number 78 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 86) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The standard is silent, and thus this is unspecified. _____________________________________________________________________________ Page: 2418 Line: 8428-8433 Section: c99 Problem: When "c99 -E" is used, an object file will never be created, no matter if the preprocessor was successful or not. Action: Change first sentence: When c00 encounters a compilation error that causes (or would cause, when the option -E was specified) an object file not... _____________________________________________________________________________ COMMENT Enhancement Request Number 79 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 87) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2419 Line: 8465-8467 Section: c99 Problem: When -c is not specified, it is unspecified if foo.o will exist at any point of time. Action: Change last sentence to: It may also create and leave foo.o. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 80 gwc@unisoft.com BUG in XCUd6 c99 (rdvk# 15) [gwc c99 space] Thu, 19 Apr 2001 18:03:44 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2420 Line: 8514 Section: c99 Problem: An unwanted space was inserted when applying XBDd5 ERN 322. Action: Remove the space from the middle of "${CENV} _LIBS" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 81 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 88) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_80 Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2420 Line: 8514 Section: c99 Problem: There's an extra space in "${CENV} _LIBS". Action: Remove space. [Ed recommendation: Dup of 80] _____________________________________________________________________________ COMMENT Enhancement Request Number 82 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 89) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: the standard is silent, and does not require the behavior _____________________________________________________________________________ Page: 2428 Line: 8734,8808-8810 Section: cd Problem: The given syntax does not allow "cd -P -". From the description of the operand "-", I don't see why this should be so. (In ksh and bash, it works as expected.) Action: Add options -P and -L on line 8734. Add after line 8810: Any specified option will be passed to 'cd "$OLDPWD"'. _____________________________________________________________________________ COMMENT Enhancement Request Number 83 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 90) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: replace sentence beginning on line 8881 "This behavior..." with This behavior is now the default. It is not consistent with the definition of dot. _____________________________________________________________________________ Page: 2431 Line: 8873-8885 Section: cd Problem: Since -L is the default, the historical ksh behaviour is now the correct one, that is, the commands "cd .. && ls" and "ls .." will now produce different output when the last component of $PWD is a symbolic link. Action: Delete text from line 8873 to the end of the sentenc in line 8883. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 84 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 11) [Schweikhardt] Wed, 18 Apr 2001 14:10:09 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 2438 Line: 9136 Section: NAME Problem: Actually the problem is above that line, in the top left corner, where the utility name "chmod" is in bold. The final "d" of chmod is partially obscured by the frame with the "IEEE P1003.1 ... Issue 6" text. This is true at least when viewing with acroread4 and when printed from within acroread4. Action: For the final document, make sure that any header (or footer) frame doesn't obscure any other text on a page. [Ed recommendation: Accept as marked the doc header is applied as an acrobat field and will be fixed in the final publication] _____________________________________________________________________________ COMMENT Enhancement Request Number 85 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 113) [muewi-1] Thu, 17 May 2001 15:57:38 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: It is out of scope to add this. _____________________________________________________________________________ Page: 2438 Line: 9139-0155 Section: chmod Problem: While chown and chgrp are required (for security reasons, at least) to support option -h (and options -LRP) to control the behaviour when finding a symlink, no matter if ownership is supported for symlinks or not, this option is missing for chmod. Action: Change synopsis to chmod [-h] mode file... chmod -R mode file... Add description of -h option, similar to that of chown -h. _____________________________________________________________________________ COMMENT Enhancement Request Number 86 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 91) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope, but we will refer to Base WG _____________________________________________________________________________ Page: 2440 Line: 9254,9292 Section: chmod Problem: The S_ISVTX permission for directories is part of XSI. Action: Add "1000 S_ISVTX" with XSI shading to the table. Add 't' to the terminals for perm. Add somewhere an extended description with XSI shading: The perm symbol t shall apply to directories only, the effect when using it with any other file type is unspecified. I shall be used with the who symbol o, a, or no who symbol at all and specify the S_ISVTX permission bit. It shall not be an error to specify a who symbol of u or g in conjunction with the perm symbol t; it is ignored for u and g. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 87 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 92) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2441 Line: 9267 Section: chmod Problem: The non-terminal "section" does not occur anywhere else. Action: Probably, "clause" is correct. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 88 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 93) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: take option 1 _____________________________________________________________________________ Page: 2442 Line: 9326-9327 Section: chmod Problem: Simply "=g" sets the owner, group, and other bits to match group bits as modified by the current umask. Action: Either change the Mode column to "uo=g" or change the Result column to describe the correct result. _____________________________________________________________________________ COMMENT Enhancement Request Number 89 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 94) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2444 Line: 9394-9395 Section: chown Problem: The new chown() function supports (gid_t)(-1) to request no change of the group ownership. Action: Change to: ...; otherwise, the group ownership shall not be changed. (also add CH , the "otherwise.." text is changed in item 3 in the DESCRIPTION to "otherwise, the group ownership shall not be changed." _____________________________________________________________________________ COMMENT Enhancement Request Number 90 muewi@baumanagement.bremen.de BUG in XCUd6 (rdvk# 117) [muewi-3] Fri, 18 May 2001 16:00:55 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: We have fixed up the rationale _____________________________________________________________________________ Page: 2445 Line: 9426-9427 Section: chown Problem: XRATd6, on p. 3318, l.1044-1061, explains in detail why the default must be a physical walk, but the description of chown (and chgrp) say the default is unspecified. Action: Here an also at page 2435 line 9063-9064 section chgrp replace the sentence starting with "Unless..." to: When none of the options -H, -L, -P are specified, the behaviour specified for -P shall be assumed. _____________________________________________________________________________ COMMENT Enhancement Request Number 91 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 95) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope. The standard is silent, _____________________________________________________________________________ Page: 2448 Line: 9552-9553 Section: cksum Problem: A few cksum implementations I tried use a file name argument of '-' to indicate stdin. As cksum supports a several file names as arguments, this makes sense. Action: Change second sentence to: If no file operands are specified or a file operand is '-', the standard input shall be used. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 92 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 96) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: We believe this is covered by pathname resolution if the submitter feels strongly submit an interpretation _____________________________________________________________________________ Page: 2468 Line: 10277-10300 Section: cp Problem: There are a few cases where the name of a symbolic link to a directory (or to a symlink to a directory or...) must behave like the name of the directory itself. Action: On lines 10277-10278, 10284, 10291, 10300 change "of type directory" to "of type directory or of type symbolic link, where the symbolic link resolves to a file of type directory". _____________________________________________________________________________ OBJECTION Enhancement Request Number 93 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 20) [DST-12] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Change "undefined results" to "unspecified results" on line 10985 to match the FORTRAN _____________________________________________________________________________ Page: 2486 Line: 10985 Section: ctags Problem: The standard (as written) requires that ctags do something undefined if C isn't supported. Granted, "undefined" *CAN* mean that it's the same result as you would get if C was supported, it doesn't naturally read that way, given the implicit "shall" of a sentence that seems to be imperative to me. Action: change to "ctags may produce undefined".... _____________________________________________________________________________ EDITORIAL Enhancement Request Number 94 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 6) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2487 Line: 11036 Section: RATIONALE Problem: Typo suspected: patrerns Action: Change "patrerns" to "patterns" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 95 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 97) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: Out of scope. _____________________________________________________________________________ Page: 2496-2498 Line: 11344-11421 Section: date Problem: The list of of conversion specifiers differs from the list in xsh6, strftime(), e.g., %g %G %V are missing. Action: Ensure the lists are the same. (Maybe this can be done automatically?) _____________________________________________________________________________ EDITORIAL Enhancement Request Number 96 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 4) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2515 Line: 12024 Section: CHANGE Problem: Typo suspected: APPLICAION Action: Change "APPLICAION" to "APPLICATION" No I did not read the whole d6 to find this and the following typos. I simply made a wordcount of d6text.txt with this perl script (which I put in the public domain) and used ispell on all words with a wordcount of 1. Of course, this doesn't find repeated typos and I was unsure for a number of words ispell warned about; because I'm not a native English speaker I would like to ask the editors to apply an ispell check to the complete wordlist to shake out most of the remaining typos in the final document. Orthography is an integral part of sanity :-) #!/usr/bin/perl -w # # usage: wordcount textfile ... while (<>) { chop; s/[^-A-Za-z0-9]+/ /g; foreach $word (split) { next if $word =~ /^\d+$/; next if $word =~ /^\d+-\d+$/; next if $word =~ /^-/; next if $word =~ /^0x/; next if $word =~ /^[A-Z][A-Z][A-Z]$/; # skip TLAs next if (length ($word) < 2); ++$count{$word}; } } foreach (sort keys %count) { printf "%5d %s\n", $count{$_}, $_; } # End of Perl Script. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 97 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 98) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add "assuming that IFS has its standard value or is unset" to both to line 12743 in parentheses _____________________________________________________________________________ Page: 2535 Line: 12746-12754 Section: echo Problem: The expansion of "$*" separates the arguments by the first character if IFS, not necessarily a space. Action: Add "assuming that IFS has its standard value or is unset" to both items. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 98 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 21) [DST-13] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_____OPEN Rationale for rejected or partial changes: We will send an email to Keith Bostic. If no response is received before we go to press no change will be made. _____________________________________________________________________________ Page: 2577 Line: 14446 Section: ex Problem: This sentence reads (in effect): "Pretend it didn't happen", but in a very obscure way. Since I'm not at all sure what this is really trying to say.... Action: Ask a vi expert to try again with this paragraph, with particular emphasis on getting rid of "in this fashion" or other pronoun-like objects. [Ed recommendation: None Change from Editorial->Comment, there are no editorial instructions here] _____________________________________________________________________________ COMMENT Enhancement Request Number 99 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 99) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: See line 17030 _____________________________________________________________________________ Page: 2647,2649 Line: 17100-17102,17214-17217 Section: file Problem: There is at least one text file which is a valid shell script, C source, and FORTRAN source (an entry at the Obfuscated C Code Contest a few years ago). So at least for the last lines of the table, a "shall" requirement is not feasible. Action: Add a note immediately following that table that the classification is sometimes a good guess, but nothing more. _____________________________________________________________________________ COMMENT Enhancement Request Number 100 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 100) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope _____________________________________________________________________________ Page: 2648 Line: 17137-17139 Section: file Problem: There is no way to require a specific number of bytes, starting at a specific offset to exist, no matter what their value might be. Action: Allow an optional unsigned decimal integer following a c specification. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 101 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 101) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: The mask value shall be AND'ed with the value of the input file before the comparison with the value field of the line is made. _____________________________________________________________________________ Page: 2648 Line: 17163-17166 Section: file Problem: It is not clear to what value the three occurrences of this word in the second sentence refer. Action: Change second sentence to: The value from the inspected file shall be AND'ed with the mask value before it is compared to the value following on that line of the `magic' file. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 102 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 102) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2649 Line: 17198-17201 Section: file Problem: The description talks about "of the bits in the value field", but only all set bits are meant. Action: Change twice to: "of the set bits in the value field" _____________________________________________________________________________ EDITORIAL Enhancement Request Number 103 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 103) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2650 Line: 17235-17236 Section: file Problem: Symbolic links are now always supported. Action: Delete paragraph. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 104 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 104) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope to invent _____________________________________________________________________________ Page: 2653 Line: 17344-17364 Section: find Problem: Sometimes it is useful to look for files where any subset of a given set of permissions is set (like, any x permission and any s bit). To specify this with the current form becomes quite clumsy. Action: Since simply using "+" in front of mode clashes with a symbolic mode starting with "+", some other means must be chosen: I would propose "-perm ^mode", with the ^ reminding of the C language XOR operator (as in file's magic files). _____________________________________________________________________________ EDITORIAL Enhancement Request Number 105 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 22) [DST-14] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2654 Line: 17387 Section: find Problem: I think this is just a typo... the ";" here seems to have been carried over from the prior synopsis. Action: Remove the ";" in the second -exec synopsis line. [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 106 gwc@unisoft.com BUG in XCUd6 find (rdvk# 28) [gwc find -exec] Wed, 16 May 2001 18:16:04 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2654 Line: 17387 Section: find Problem: As well as removing the ";" as requested by Donn Terry, the formatting of "{} +" could be improved. Action: Either add more space between the "{}" and the "+" or change the "{} +" to a constant-width font. [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 107 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 105) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope to invent _____________________________________________________________________________ Page: 2660 Line: 17672-17673 Section: fold Problem: As fold supports a several file names as arguments, it makes sense to support '-' for stdin. Action: Change first sentence to: If no file operands are specified or a file operand is '-', the standard input shall be used. _____________________________________________________________________________ COMMENT Enhancement Request Number 108 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 106) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add to DESC para 2 When a file contains less than number lines, it shall be copied to standard output in its entirety. This shall not be an error. _____________________________________________________________________________ Page: 2702 Line: 19198-19200 Section: head Problem: What happens when the file contains less than number lines? Action: Add here (or under STDOUT or DESCRIPTION?): When a file contains less than number lines, it is produces on STDOUT in its entirety. This is not an error. _____________________________________________________________________________ COMMENT Enhancement Request Number 109 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 107) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope _____________________________________________________________________________ Page: 2702 Line: 19204-19205 Section: head Problem: As head supports a several file names as arguments, it makes sense to support '-' for stdin. Action: Change first sentence to: If no file operands are specified or a file operand is '-', the standard input shall be used. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 110 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 5) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2729 Line: 20202 Section: OPTIONS Problem: Typo suspected: SYMOPSIS Action: Change "SYMOPSIS" to "SYNOPSIS" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 111 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 108) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2729 Line: 20202 Section: kill Problem: SYMOPSIS Action: SYNOPSIS [Ed recommendation: Accept] _____________________________________________________________________________ COMMENT Enhancement Request Number 112 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 27) [DST-20] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The output of ls is specified, even if not predictable for this case. No change is required. The review group did not see the need to make this unspecified, since translating filemodes with bits unknown to ls is covered, unknown uid or gid is covered, etc. So, while the output may not be predictable, it will always be according the specification here. The application will not crash, nor will applications that depend on the output of ls in this case. _____________________________________________________________________________ Page: 2773 Line: 21919 Section: ls Problem: The lstat() function is permitted to return an unspecified value when reading the owner, group, and mode of a symbolic link (because they are not required to have one). (1207/23977). In the case of a symbolic link, what is ls to print if that information is "random trash"? Action: Something needs to be said, even if it's "The owner, group, and access bits shall correspond to the values returned by lstat(); if the value returned by lstat() is unspecified, ls shall print an unspecified value for each of these fields." (Or however you want to say "ls won't crash".) (Do we want to explicitly disallow blanks in the fields in this case?) _____________________________________________________________________________ COMMENT Enhancement Request Number 113 muewi@baumanagement.bremen.de BUG in XCUd6 (rdvk# 118) [muewi-3] Fri, 18 May 2001 16:28:10 +0200 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope Refer to Base WG, we may reverse the decision if the Base WG indicate we should _____________________________________________________________________________ Page: 2774 Line: 21976-21979 Section: ls Problem: XSI supports S_ISVTX on directories Action: Add, with XSI shading: T In `other permissions': the file is a directory, search permission is not granted to others, and the restricted deletion flag is set. t In `other permissions': the file is a directory, search permission is granted to others, and the restricted deletion flag is set. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 114 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 23) [DST-15] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 2776 Line: 22058 Section: ls Problem: "is currently" is obsolete. "Was historically" is better. Action: As above. [Ed recommendation: Accept] _____________________________________________________________________________ OBJECTION Enhancement Request Number 115 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 26) [DST-19] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: This issue had been discussed before, and the authors of this part of the specification have indicated that the wording is intentional. This change would reduce consensus (AJ hold this objection on file for the final recirc) _____________________________________________________________________________ Page: 2780 Line: 22240 Section: m4 Problem: (Not accepted previously) The specification of 32 bit signed arithmetic is inconsistent with every other place in the standard where aritmetic is discussed (in both the XCU and XSI) In XCU, it consistently simply says something about "a C signed long type" (see 2243/1709, where discussing shell aritmetic, which is very permissive, I'm not asking for that much). Action: Use either wording like that for the shell, or "using arithmetic corresponding to the C-language signed long type". _____________________________________________________________________________ COMMENT Enhancement Request Number 116 curtis.smith@vsin.com Bug in XCUd6 pax (size) (rdvk# 12) {cms-pax} Thu, 19 Apr 2001 15:09:36 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Add leading 7 to the octal limit on 27532 _____________________________________________________________________________ Page: 2915 Line: 27531-27532 Section: pax Problem: (Draft 5 ERN reference: NA ) The size above which an extended header is required is specified as a decimal number and its octal equivalent, yet the two values are not the same. The decimal value is 2^33-1; the octal value is 2^30-1. Action: verify correct value and fix incorrect one(s) _____________________________________________________________________________ COMMENT Enhancement Request Number 117 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 109) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: delete ", which.... directory." on line 31937 Add as a second sentence For each dir operand the rmdir utility shall perform actions equiv to the rmdir() function defined in the SI Vol of ...., with the dir operand as its only argument _____________________________________________________________________________ Page: 3029 Line: 31936-31941 Section: rmdir Problem: When $PWD is an empty directory, "rmdir ." will not succeed. ("rmdir .." won't, either, but either ".." is not empty, since $PWD still exists, or it is empty since $PWD was deleted by some other action, but then ".." does not exist anymore.) Action: Change end of first paragraph to: ...empty directory besides '.' or '..'. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 118 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 110) [muewi-1] _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 3048 Line: 32666 Section: sh Problem: The option +i is missing. Action: Insert 'i' at its correct place. _____________________________________________________________________________ COMMENT Enhancement Request Number 119 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 111) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: The wording is as per the base document, if the submitter feels strongly an interpretation request should be submitted. _____________________________________________________________________________ Page: 3048 Line: 32702-32704 Section: sh Problem: Since sh shall support the Utility Syntax Guidelines, a "--", and any argument not starting with a '-', will terminate any special processing, thus forcing a subsequent occurrence of '-' to be treated as a regular operand. Action: Change second sentence to: If operands beginning with a '-' follow the single hyphen, results are undefined. _____________________________________________________________________________ COMMENT Enhancement Request Number 120 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 24) [DST-16] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 3063 Line: 33353 Section: sh Problem: "This standard" doesn't ALWAYS mean the standard we're drafting, and certainly the "earlier versions..." clause doesn't apply to a standard that never did and never will (presumably) exist. Action: "earlier versions of IEEE...." -> "to such a hypothetical standard". [Ed recommendation: Accept as marked: l 33351 "be included in earlier versions of IEEE...." -> "be standardized" l 33353 "conforming strictly to earlier versions of IEEE....but lacking" -> "strictly conforming versions lacking" l 33358 "from earlier versions.... 1003.1" -> "" l 33362-33364 "at this time, rather .... 1003.1." -> "." ] _____________________________________________________________________________ COMMENT Enhancement Request Number 121 gwc@unisoft.com BUG in XCUd6 test (rdvk# 16) [gwc test -L] Thu, 26 Apr 2001 18:59:45 +0100 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 3106 Line: 34914 Section: test Problem: The exception for -h applies to -L as well. Action: Change "the -h file primary" to "the -h file and -L file primaries" _____________________________________________________________________________ COMMENT Enhancement Request Number 122 gwinn@res.ray.com Bug in XCUd6 touch (rdvk# 120) {JMG6-2} Sat, 19 May 2001 02:09:58 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 3120 Line: 35439 Section: touch Problem: (Draft 5 ERN reference: NA ) There are no "double leap seconds", and there never were, so one need not allow for them, and ISO/IEC 9899: 1999 standard has nothing to do with it. Action: Change "The range for double leap seconds is changed from [00,61] to [00,60] to align with the ISO/IEC 9899: 1999 standard" to "The range for seconds is changed from [00,61] to [00,60] to align with the ISO/IEC 9899: 1999 standard, and to allow for positive leap seconds." [Ed recommendation: Accept as marked we will need to check for other similar occurrences in XSH and XBD] _____________________________________________________________________________ COMMENT Enhancement Request Number 123 muewi@baumanagement.bremen.de BUG in xcu6 (rdvk# 112) [muewi-1] _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_____ Reject_X___ Rationale for rejected or partial changes: out of scope for us to be adding such a thing at this late stage _____________________________________________________________________________ Page: 3138 Line: 36080-36155 Section: ulimit Problem: Since getrlimit() and setrlimit() are part of XSI, the corresponding ulimit options should be supported, too. Action: New text: SYNOPSIS ulimit [-H|-S] [blocks] ulimit [-H|-S] -a ulimit [-H|-S] -[cdfnstv][value] ... DESCRIPTION The ulimit utility shall set or report resource limits. These limits apply to the shell and its child processes. OPTIONS The following options shall be supported. Every option letter besides H, S, and a takes an optional value argument. When the option letter is not the last character if a parameter, the remaining characters of the parameter provide the argument. Otherwise, when the next argument does not start with a '-', it provides the value. All values are either unsigned decimal integers or the string "unlimited". When no value is provided, the corresponding limit is reported, otherwise it is changed. When no option besides -S or -H appears, -f is assumed. -S Set or report the soft limit. The soft limit is the limit actually used to check if a limit is exceeded. It cannot be increased above the hard limit. -H Set or report the hard limit. Only a process with appropriate privileges can increase the hard limit. When neither -S nor -H are specified, both limits are set and soft limits are reported. When both are specified, the result is undefined. -a Report all limits -c Set or report the maximum size of core files in units of 512 bytes -d Set or report the maximum size of data segment in units of 1024 bytes. -f Set or report the maximum file size in units of 512 bytes -n Set or report the largest number allowed for a file descriptor plus 1, that is, the maximum number of open files. -s Set or report the maximum size of stack segment in units of 1024 bytes. -t Set or report the maximum CPU time a process may use in seconds. -v Set or report the maximum memory size of a process in units of 1024 bytes. OPERANDS blocks same as "-f blocks" STDIN Not used. INPUT FILES None. ENVIRONMENT VARIABLES (same as before) ASYNCHRONOUS EVENTS Default. STDOUT When, besides -H or -S, neither an option nor an operand---in which case -f is assumed---or exactly one option with no corresponding value is specified, the corresponding limit is written in the following format: "%d\n", If there is no current limit, in the POSIX locale the following format shall be used: "unlimited\n" When, besides -H or -S, more than one option without a corresponding value is specified, or when the option -a is specified, for each limit requested, one line of output is written in the following format: "%s: %d\n", , or, when there is no current limit, "%s: unlimited\n", The is implementation defined, must not contain newline characters and shall contain an indication of the units used for the limit. The order in which the limits are reported is unspecified. (All subsequent sections remain besides SEE ALSO, where a reference to getrlimit() and setrlimit() have to be added, and, of course the CHANGE HISTORY. But maybe another example might be useful: Report current hard limits: ulimit -Ha time(seconds) unlimited file(blocks) unlimited data(kbytes) unlimited stack(kbytes) unlimited coredump(blocks) unlimited nofiles(descriptors) 1024 vmemory(kbytes) unlimited Report current soft limit for coredump size and number of open files: ulimit -s -c coredump(blocks) 0 nofiles(descriptors) 64 Report current cpu time limit: ulimit -t unlimited Note that, when exactly one limit is requested, only the value is written, which simplifies its usage in shell scripts. [Ed recommendation: None Looks very much out of scope for a review doing the final sanity check] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 124 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 8) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 3159 Line: 36840 Section: EXAMPLES Problem: Typo suspected: repreated Action: Change "repreated" to "repeated" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 125 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 7) [Schweikhardt] Thu, 19 Apr 2001 08:53:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 3177 Line: 37439 Section: CHANGE Problem: Typo suspected: reaponse Action: Change "reaponse" to "response" [Ed recommendation: Accept] _____________________________________________________________________________ EDITORIAL Enhancement Request Number 126 Donn Terry Bug in xbdd5 Bug in xcu (rdvk# 25) [DST-17] Tue, 15 May 2001 12:37:07 -0700 _____________________________________________________________________________ Accept_____ Accept as marked below_X___ Duplicate_____ Reject_____ Rationale for rejected or partial changes: Item 1 is a comma replace , with replace 'character with and `character replace 0 with reorder as per editor _____________________________________________________________________________ Page: 3187 Line: 37791 Section: vi Problem: Much better... However it still has problems: 1) ' (right single quote) appears both on the same line as and on the line containing "? W l". 2) Both ' and ` should be NAMED by their proper name (as well as printed) (both here and at 38184 and 38195.) (The names used for these characters in XBD P136/137 are ' = and ` = , but if the 10646 names are different, use them. I don't have a 10646 to look at.) 3) Rearrange to put $ and % in one of the other columns. You have plenty of room to do that, and it will make the table a little shorter. Action: 1) Remove duplication. 2) Name the ambiguous characters (for identically the same reasons we do it with l (ell) and 1 (one) as command line options). 3) Rearrange. _____________________________________________________________________________ EDITORIAL Enhancement Request Number 127 ajosey@opengroup.org Bug in XCUd6 yacc (rdvk# 124) {aj.yacc} Mon, 21 May 2001 14:16:36 +0100 (BST) _____________________________________________________________________________ Accept_____ Accept as marked below_____ Duplicate_of_128 Reject_____ Rationale for rejected or partial changes: See Ed recommendation _____________________________________________________________________________ Page: 3261 Line: 40636 Section: yacc Problem: (Draft 5 ERN reference: NA ) parameter "file" should be "grammar" to match the synopsis Action: change "file" to "grammar" [Ed recommendation: Dup of 128] _____________________________________________________________________________ COMMENT Enhancement Request Number 128 Jens.Schweikhardt@marconi.com BUG in XCUd6 (rdvk# 10) [Schweikhardt] Wed, 18 Apr 2001 14:10:09 +0200 _____________________________________________________________________________ Accept_X___ Accept as marked below_____ Duplicate_____ Reject_____ Rationale for rejected or partial changes: _____________________________________________________________________________ Page: 3261 Line: 40636 Section: DESCRIPTION Problem: The text mentions "file" whereas the SYNOPSIS, OPERANDS, INPUT FILES and STDERR mention "grammar" as the operand. Action: Change "file" to "grammar" [Ed recommendation: Accept]