20 July 2006 J3/06-203 Sub ject: Comments on Clause 10 From: Van Snyder 1 Edits -- and comments without editorial suggestions Edits refer to 06-007. Page and line numbers are displayed in the margin. Absent other instructions, a page and line number or line number range implies all of the indicated text is to be replaced by associated text, while a page and line number followed by + (-) indicates that associated text is to be inserted after (before) the indicated line. Remarks are noted in the margin, or appear between [ and ] in the text. [259:5,7]----------------------------------------------------------------------- [A PRINT statement or a READ statement without a control list doesn't have a FMT= specifier; all it has is . Editor: "FMT= specifier" "" twice.] [264:19-20]----------------------------------------------------------------------- [How many times is it necessary to say, on one page, that the A edit descriptor might cause file posi- tioning? Once is enough. Editor: delete "; during . . . positioning".] [264:37]----------------------------------------------------------------------- [Editor: "\bits{bits, }" "\bits{bits,}".] [266:19-267:1]----------------------------------------------------------------------- [Copy "A lower-case . . . specification" to be an additional sentence within the paragraph at [267:20], which is where it belongs. Then delete "an . . . or".] [267:28]----------------------------------------------------------------------- [Is the plus sign optional if the SIGN mode is PLUS?] [267:31-33]----------------------------------------------------------------------- [Is a sign allowed for NaN output? One is allowed for input. Is a plus sign required for a positive NaN if SIGN mode is PLUS?] [267:35]----------------------------------------------------------------------- [Is the plus sign optional if the SIGN mode is PLUS?] [268:2]----------------------------------------------------------------------- [Editor: Insert "(10.7.2.3.7)" after "rounded".] [268:17-20]----------------------------------------------------------------------- [Do ISO typesetting rules require bullets?] [268:19]----------------------------------------------------------------------- [Editor: Insert "(10.7.2.3.7)" after "rounding".] [269:8-14]----------------------------------------------------------------------- [Do ISO typesetting rules require bullets?] [269:10]----------------------------------------------------------------------- [Editor: Insert "(10.7.2.3.7)" after "rounding", then make the parenthesized phrase a separate nonparen- thesized sentence.] [270:3-8]----------------------------------------------------------------------- [Do ISO typesetting rules require bullets?] [270:5]----------------------------------------------------------------------- [Editor: Insert "(10.7.2.3.7)" after "rounding".] [270:22,24]----------------------------------------------------------------------- [Editor: Anticipating IEEE 754R, delete "(typically in binary form)", "exact" "finite", "on most processors" "in binary form" .] [271:24-27]----------------------------------------------------------------------- [If these two paragraphs say the same thing, delete one of them. Otherwise, delete one of them.] [273:4-8]----------------------------------------------------------------------- [Editor: Replace the comma after "positions" by a semicolon, insert "for real or complex data" before 20 July 2006 Page 1 of 4 20 July 2006 J3/06-203 "the fractional", delete "of which", delete "When . . . effect.", delete the extraneous quote at the end of the paragraph.] [273:14]----------------------------------------------------------------------- [Editor: "Iw" "Iw.m where the value of m is specified by d if d appears and is considered not to appear if d does not appear".] [274:6]----------------------------------------------------------------------- [Editor: "Zw" "Zw.m where the value of m is specified by d if d appears and is considered not to appear if d does not appear".] [277:8-12]----------------------------------------------------------------------- [Editor: Move item (2) before item (1), delete the commas after "input" (twice), replace "(provided that no exponent exists in the field)" at [274:8-9] by a comma.] [277:14,18]----------------------------------------------------------------------- [Editor: Delete the commas after "output" twice.] [277:20]----------------------------------------------------------------------- [Editor: Delete the comma after "input".] [277:22]----------------------------------------------------------------------- [Editor: Delete the comma after "output".] [277:35]----------------------------------------------------------------------- [Editor: What causes the extraneous blank after "editing"? The TEXsource looks fine. Is something in the conditional macros adding non-rubber space?] [278:2]----------------------------------------------------------------------- [Editor: Add "10.6" to the cross references.] [278:31-37]----------------------------------------------------------------------- [The term "contiguous blanks" is confusing. What are discontiguous blanks? Blanks with blanks between them? If it means that all of the blanks before and after a comma, semicolon or slash, or all of the blanks between values, constitute a single separator (so null values are not manufactured by sequences of blanks), it would be clearer to say so.] [279:21-22]----------------------------------------------------------------------- [Is the separator in a complex value different from a value separator? In particular, is it prohibited simply to separate the real and imaginary parts by blanks? It is confusing to use the term "separator" to mean something different from "value separator" so soon after defining "value separator" (and then not so far using it for anything).] [280:15]----------------------------------------------------------------------- [The cited cross reference (7.4.1.3) refers to a paragraph about intrinsic assignment. We don't really have a term "character assignment statement". Editor: "a character" "an intrinsic".] [280:27]----------------------------------------------------------------------- [[280:11-16] says that character input works like character assignment, but so far there hasn't been anything that says input for other types works like assignment. So what does "after the assignment of the previous value" mean?] [280:29]----------------------------------------------------------------------- [The phrase "is defined" sounds as though the input takes over control of the manipulation of the from the input list. If it really means that the value of the after execution of the input statement is the same as it would be if enough values had been supplied, it would be clearer to say "becomes defined" instead of "is defined".] [281:6-8]----------------------------------------------------------------------- [The phrase "except within character constants" sounds like it's OK to begin new records or insert blanks within the real or imaginary part of a complex constant. Editor: "for complex constant and character sequences" "for character sequences, or after the sepaator within a complex constant as explained below", second "for complex constants" " after the sepaator within a complex constant as explained below".] 20 July 2006 Page 2 of 4 20 July 2006 J3/06-203 [283:0+1]----------------------------------------------------------------------- [The term "exactly" implies that the length may be specified approximately, but, since it isn't, the reader who goes looking for the approximate value will be sent on a wild-goose chase. Editor: Delete "exactly".] [283:6]----------------------------------------------------------------------- [Again, what does "contiguous" mean, especially given the last sentence of the paragraph: "Any sequence of two or more blanks is treated as a single blank. . . ."] [283:14]----------------------------------------------------------------------- [What list? Editor: "list item" "effective item", although this won't become clear until [284:24-26].] [284:11+]----------------------------------------------------------------------- [If a subob ject designator is specified for a group ob ject item, is the definition status of subob jects for which no subob ject designatot is specified affected?] [284:12]----------------------------------------------------------------------- [Subclause 10.11.1.2 (except for its first paragraph) is more about group ob ject list items, and Subclause 10.11.1.3 is more about input values. Editor: Exchange the Subclause headings.] [284:16]----------------------------------------------------------------------- [Editor: Replace the cross reference "(10.8.8)" by "(10.6)", since (10.8.8) doesn't actually say anything relevant to the present discussion.] [284:22-31+4]----------------------------------------------------------------------- [Would be clearer if first in the subclause. Editor: Move the parfagraphs and Note 10.35 to [284:12+].] [284:33]----------------------------------------------------------------------- [I couldn't find a specification that namelist works "as if " by assignment. Is "assignment" the best word her?] [284:36-38]----------------------------------------------------------------------- [The material of the paragaph belongs in 10.11.1.6. But don't bother to to move it -- delete it instead -- because it's partly redundant and partly nonsense. The nonsense part is that a comment is not permitted to appear after a slash and that a comment is not permitted to appear within a character literal constant. The former is contradicted by [284:39-40] -- "the remainder of the record is ignored" -- while the latter is nonsense because if something that looked like a comment appeared within a character literal constant it would be part of the character literal constant. It is possible, but perhaps fatuous, to add a Note 10.38a "If text that appears to be a namelist comment appears within a character literal constant, it is not a comment, it is part of the constant." The remainder is redundant to 10.11.1.6.] [285:1+]----------------------------------------------------------------------- [The paragraphs at [283:11-16] and [284:13-21] belong in this subclause. Move them here after making the changes within those paragraphs listed above. Ten replace the reference to this subclause (10.11.1.3) in the paragraph moved from [284:13-21] by "this subclause".] [285:16]----------------------------------------------------------------------- [Editor: There is no list. Delete "list".] [285:29-30]----------------------------------------------------------------------- [The cited cross reference (7.4.1.3) refers to a paragraph about intrinsic assignment. We don't really have a term "character assignment statement". Editor: "a character" "an intrinsic".] [285:34]----------------------------------------------------------------------- [Blank is a value separator. Editor: Insert "nonblank" before "value".] [287:11-288:5]----------------------------------------------------------------------- [Subclause 10.11.2.1, with the exception of Note 10.40, appears to be identical to material in 10.10.2 -- probably intentionally so. Editor: Replace the subclause title by Editing of namelist output values, then replace [287:12-34] by "Values in namelist output records are edited as for list-directed output (10.10.2)." Then delete [288:1-5] and the last sentence of unresolved technical issue note 48.] [288:7]----------------------------------------------------------------------- [For list-directed output, the "r*c" form is allowed for two consecutive identical values, even if they don't 20 July 2006 Page 3 of 4 20 July 2006 J3/06-203 correspond to an array. Editor: "in an array" "for the same namelist group item".] 20 July 2006 Page 4 of 4