[Cfp-interest 2514] Fwd: N2912 review comments JT-024 through JT-069

Tue Aug 9 11:06:16 PDT 2022

This forward is just for historical reference. A subsequent message will show the parts that are still relevant for CFP review of C23 review draft  N3047.

- Jim Thomas

> Begin forwarded message:
> 
> From: Jim Thomas <jaswthomas at sbcglobal.net>
> Subject: N2912 review comments JT-024 through JT-069
> Date: July 8, 2022 at 2:35:11 PM PDT
> To: JeanHeyd Meneide <phdofthehouse at gmail.com>
> Cc: Rajan Bhakta <rbhakta at us.ibm.com>, "Fred J. Tydeman" <tydeman at tybor.com>, David Keaton <dmk at dmk.com>
> 
> Jean-Head,
> 
> The following comments are from my review of the integration of 
> 
> N2601 <https://www.open-std.org/jtc1/sc22/wg14/www/docs/n2601.pdf> 2020/11/01 Thomas, C2X proposal - TS 18661-3 annex update 3
> 
>  into N2912 as Annex H.
> 
> - Jim Thomas
>  
> Comments:
> 
> JT-024 H.1#2: “202011L” will need to be updated.
> 
> 
> 
> JT-025 H.2.1, in the last line of both tables: change two instances of “emax” to  “emin” and change “maximum” to “minimum”.
> 
>  
> 
> JT-026 H.2.3, in the first (header) row of the table: Consider merging the cell with “Extended formats associated with:” and the empty cell to its right. As is, the binary32 column appears to be distinguished from the columns to its right.
> 
>  
> 
> JT-027 H.2.4#1: The first sentence needs to be as in N2601, i.e. remove “as well”. The words for standard floating types and for decimal floating types should be parallel.
> 
>  
> 
> JT-028 H.3#2-3: In both paragraphs, in the text for N and N + 1, I think a comma is needed between “type” and “evaluate”.
> 
>  
> 
> JT-029 H.3#6: Remove the line “The macro”.
> 
>  
> 
> JT-030 H.3#6, 2nd sentence: The spacing around the “-“ symbol looks wrong.
> 
>  
> 
> JT-031 H.4.3: In the bullets are three places where references are to X… instead of H… and the links are missing.
> 
>  
> 
> JT-032 H.9 and throughout: I don’t understand the color scheme, but some things seem inconsistent. For example, in 7.3.5 in cacosfNx the “x” is part of the function name so should have the same color as “cacosf”.  Similarly, in _FloatNx the “x” is part of the type name so should have the same color as “_Float”. Is the coloring still in progress? There are further comments about color, where it seems obviously wrong, but my review of color is spotty.
> 
>  
> 
> JT-033 H.9#3, 7.3.5 functions: Remove the blank line from the box.
> 
>  
> 
> JT-034 H.9#3, 7.3.7 - 7.3.9 functions: blank lines appear in the boxes in inconsistent ways.
> 
>  
> 
> JT-035 H.10#3, 1st sentence: the function names “fegetround” and “fesetround” have the wrong typeface and color.
> 
>  
> 
> JT-036 H.10#4, 2nd sentence: The six function names in the parenthetical text have the wrong typeface and color.
> 
>  
> 
> JT-037 H.10#5, 1st sentence: The three function names in the parenthetical text have the wrong typeface and color.
> 
>  
> 
> JT-038 H.10#5, 2nd sentence: There is a bad line break before the “N” in “strfromencbindN”.
> 
>  
> 
> JT-039 H.11.1#2-8: Most of the macro names have the wrong color.
> 
>  
> 
> JT-040 H.11.2#1: After the line “7.12.4 Trigonometric functions” a large bock of lines is missing, beginning with “_FloatN acosfN(_FloatN x);” through “_DecimalNx ldexpdNx(_DecimalNx value, int exp);”.
> 
>  
> 
> JT-041 H.11.2#1: In the prototypes for the interchange types the parameter x is missing and an x is appended to the type name (changing it from an interchange type to extended type). For example, “long int llogbfN(_FloatN x)” is written as “long int llogbfN(_FloatNx)”. This happens for both _FloatN and _DecimalN types, throughout the subclause.
> 
>  
> 
> JT-042 H.11.2#1:  Blank lines between function families are missing starting with log1p through scalbln.
> 
>  
> 
> JT-043 H.11.2#1: In the box for 7.12.10.2, blank lines to separate the fmod, remainder, and remquo families are missing.
> 
>  
> 
> JT-044 H.11.2#1: In the box for 7.12.11, blank lines to separate the copysign, nan, nextafter, nextup, and nextdown families are missing.
> 
>  
> 
> JT-045 H.11.2#1: In the boxes for 7.12.12, blank lines to separate the function families are missing. (Maybe the omission of blank lines was intentional (?) but I don’t see what criteria were used.)
> 
>  
> 
> JT-046 H.11.2#1: In the box for 7.12.14, 7.12.15, 7.12.16, F.10.12, and F.10.13, blank lines to separate some of the function families are missing.
> 
>  
> 
> JT-047 H.11.2#4: Change “7.12.6.10” to “7.12.6.17”.
> 
>  
> 
> JT-048 H.11.3: In the titles of the subclauses of H.11.3 the type size indication “N” in the function names is not italicized.
> 
>  
> 
> JT-049 H.11.3.1: “unsigned char” has the wrong typeface.
> 
>  
> 
> JT-050 H.11.3.1.1#2 and #3: The “N” in encodefN is not italicized (both paragraphs).
> 
>  
> 
> JT-051 H.11.3.1.2#2 and #3: The “N” in decodefN is not italicized (both paragraphs).
> 
>  
> 
> JT-052 H.11.3.2.1#2: The “M” and “N” in fMencfN are not italicized.
> 
>  
> 
> JT-053 H.11.3.2.1#2: Two instances of “encNptr” have the wrong typeface.
> 
>  
> 
> JT-054 H.11.3.2.2#2: In the first sentence, the “M” and “N” in the two function names are not italicized, and the “and” between the function names is red. In three instances in the paragraph the type size indicator “N” is bold.
> 
>  
> 
> JT-055 H.12.1#1: In the second box, change the first instance of “_DecimalNx” to “_DecimalN”.
> 
>  
> 
> JT-056 H.12: In many instances in the subclauses of H.12 the type size indicator “N” appears to be bold.
> 
>  
> 
> JT-057 H.12.2#3: In two instances in the paragraph “M” is not italicized.
> 
>  
> 
> JT-058 H.12.2#4: The [-]0xh.h…hp±d at the end of the example is intended to represent the form of the string, to help determine its length. Some of the characters in the form (“-“, “0”, “x”, “p”, “.”) are the actual characters in the string and some (“h”, “+-“, “d”) are placeholders. It would be helpful for the actual characters to have the typeface for program output, as in N2601. In the same vein, the space around “+-“ might be misleading.
> 
>  
> 
> JT-059 H.12.3.1: In the heading, there is a space before the “N” in the function name.
> 
>  
> 
> JT-060 H.12.3.1#2: “encptr” has the wrong typeface.
> 
>  
> 
> JT-061 H.12.3.2#1: Remove the “x” from “strfromencbindNx”. This was wrong in N2601.
> 
>  
> 
> JT-062 H.12.3.2#2: “strfromencbindN” and “strfromdN” have the wrong typeface.
> 
>  
> 
> JT-063 H.12.4: In the first sentence “N” is not italicized.
> 
>  
> 
> JT-064 H.12.4.2#2: The “N” in “N/8” in the first sentence is not italicized.
> 
>  
> 
> JT-065 H.12.4.2#2: The “strtoencbindN” in the last sentence has the wrong typeface.
> 
>  
> 
> JT-066 H.13#5: The “f” and “d” prefixes in the second sentence have the wrong typeface.
> 
>  
> 
> JT-067 H.13#5: In “… the macros fM, fMx, …” insert “with” after “macros”.
> 
>  
> 
> JT-068 H.13#5: In the bullet “For prefix dM: …”, in the first sentence “P” and “T” are not italicized.
> 
>  
> 
> JT-069 H.13#6: I don’t think “Declarations” should be capitalized?
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20220809/810461fe/attachment-0001.htm>