To: J3                                                     J3/26-131
From: Malcolm Cohen
Subject: Editor's report for 26-007r1
Date: 2026-March-28


1. Introduction

This is the editor's report from the creation of 26-007r1.


2. Papers from 2024 (edits against 24-007)

(meeting 233) 24-134r1 The unbearable vagueness of being CFI_setpointer

NOTE Defective edit instructions did not say to delete "of", but the result
     shown did not contain "of". I deleted "of" to get the result shown.
[528:27-28] 18.5.5.9 The CFI_setpointer function, p3 Description

Done as intended.

(meeting 234) 24-189 Editorial improvement in CO_REDUCE

Done without change.


3. Papers from meeting 239

26-123r4 US19 more functions

DIFF Deleted duplicate "were added".
     Omitted parenthetical mention of IEEE-754.
NOTE I hard-coded the reference, as there will be a revision soon, and it
     might add more functions, change the functions, etc. For the purposes
     of F203z, this being hard-coded will retain the proper history.
NOTE It is likely that IEEE-754 will be revised before our publication, but
     I think it is unlikely that the ISO version will be updated before then.
[xv] Introduction

DIFF No, after LOGICAL.
NOTE And yes, all the other new ones starting LOG alphabetically precede LOGP1.
[430] after LOG add one line:

COMMENT Not quite convinced that the technical "radical function" is better
        than "Nth root function". Why would we not call it RADICAL if it were
        the radical function?
[431] after RESHAPE add one line:

COMMENT Not sure why "exponential growth" is better than "compound interest"
DIFF ">=-1" -> prose.
DIFF You spelled it "1+X" in one place and "X+1" in another.
     Changed the second to "1+X".
DIFF Used maths not pseudo-Fortran, as we nearly always do in this clause.
UTI 032 The N argument should be integer according to IEEE.
[464:34+] insert one new intrinsic {note: after COMPLETE and after its NOTE}

DIFF Used maths not pseudo-Fortran, as we nearly always do in this clause.
[476:8+] add these three new intrinsics {note: after EXP}

DIFF Wrong insert location - I inserted after LOGICAL.
     "-1" -> "minus one".
     "X+1" -> "1+X" (to match the description).
[497:25+] insert one intrinsic: {note: after LOG}

DIFF "-1" -> "minus one".
     Maths not pseudo-Fortran.
     Used 1+X to match the descriptions.
UTI 033 Inconsistency with LOG10P1, etc.
COMMENT: Why do we not just recommend that if the user writes LOG(1+X) or
         EXP(x)-1, the processor should use a more accurate but equally
         efficient approximation? That would
         (a) be far more profitable,
         (b) benefit everyone not just those inserting kludgey EXPM1 etc.,
         (c) retain Fortran's mathematical bent,
         (d) a similar amount of implementation work.
         Having these separate intrinsics are a waste of valuable implementer
         resource.
[498:11+] insert three intrinsics:  {note: after LOG10)

DIFF ", and" -> ". Its value".
     Maths and prose not pseudo-Fortran that assumed N was exactly
     representable in default real.
[525:17+] insert one intrinsic {note: after RESHAPE}

COMMENT The argument for this function accepting complex is pretty weak as
        there is no hardware support for complex division or square root, let
        alone reciprocal square root. Obviously, as IEEE don't have any complex
        arithmetic at all, IEEE is not a reason for RSQRT(complex).
DIFF Maths and prose not pseudo-Fortran. I would have used just maths, only we
     weirdly use prose (and no maths) for SQRT.
[525:27+] insert one intrinsic {note: after RRSPACING}

Done with changes.

26-127 editorial

Done without change.

26-128r2 C23 enums

DIFF Used the correct ISO reference (blank, not a hyphen, before 9899).
[xv] Introduction in Data declaration,

COMMENT Wording reads a little clunky, but seems ok.
[109:18-20] Same subclause, p2,

DIFF spurious hyphen in this ISO reference too.
[110:15+14+] Between NOTE 4 "Example of an interoperable..."

Done with changes.

26-129r1 US11 source location

DIFF "in" -> "from". (that's how we usually say it, 124 times vs 11 times).
[xv] Introduction

DIFF "Shall be" -> "shall be".
     ".FALSE." -> "false".
     Rewrote first set of cases to use otherwise, to reduce weasel-wording.
     ".TRUE." -> "true".
     Many other rewordings.
     Added UTI 031 about PROCEDURE_NAME and GENERIC TEMPLATES
     Added UTI 034 about PROCEDURE_NAME in an interface body.
[554:35+] add one intrinsic (after PARENT_TEAM)

DIFF Rewordings.
     Added trivial example for SOURCE_LINE.
[555:13+] Add two new intrinsics (note: after SINGLE_PRECISION)

Done with changes.

"editorial cleanup"
    Delete empty bullets in Introduction.
        Source form
    and
        Execution control
    Delete par nums in NORM2
        done.
    Reviewed UTI 32/33 text for correctness.

Done.

25-124r1 editorial fixes for subobjects.

Done without change.

===END===