Scheduled for 2024-Oct-08 LEWG Telecon.

2024-09-17 Library Evolution Telecon

P3348R0: C++26 should refer to C23 not C17

2024-09-17 Library Evolution Telecon Minutes

Champion: Jonathan Wakely
Chair: Fabio Fracassi
Minute Taker: Eddie Nolan

Summary

No polls were taken, author will apply the minor fixes and the revised paper will be seen again by LEWG.

Next Steps

This paper was seen on 2024-09-17 by LEWG, but lacked quorum for forwarding (due to CppCon). Some non-binding feedback has been given to the author. This paper will be re-scheduled. A new version that incorporates the feedback would be preferred.

2024-10-08 Library Evolution Telecon

P3348R1: C++26 should refer to C23 not C17
(looked at D revision)

2024-10-08 Library Evolution Telecon Minutes

Champion: Jonathan Wakely
Chair: Fabio Fracassi
Minute Taker: Guy Davidson

Summary

POLL: We want to rebase C++26 onto C23 (as drafted in D3348R1 once the outstanding issues are incorporated)

Next Steps

• We approved the general direction of the paper, and the resolutions proposed
• There are topics that still require resolution in SG6 (numerics) and SG22 (liaison group for alignment with C).
  • For numerics:
    • FLT_HAS_SUBNORM, DBL_HAS_SUBNORM and LDBL_HAS_SUBNORM macros marked obsolescent. (should do the same?)
    • deprecate INFINITY and NAN macros (define in <float.h> and <match.h> headers)
    • additions to <fenv.h>, <math.h>
  • For liaison: C macros added

P3348R1 C++26 should refer to C23 not C17 (Jonathan Wakely)

P3348R1: C++26 should refer to C23 not C17

2024-11-18 SG6 Minutes

Chair: Matthias Kretz

Champion: Jonathan Wakely

Summary

On the floating-point macros, just do the same in C++ (except that apply only to new <cmath> functions).

We want all the new <cmath> functions (including iscanonical) in a new paper, so that this paper can go quicker.

We want to ignore everything about decimal floating-point.

P3348R2 C++26 should refer to C23 not C17 (Jonathan Wakely)

2025-02-10 Library Evolution Hagenberg Meeting

P3348R2: C++26 should refer to C23 not C17

2025-02-10 Library Evolution Hagenberg Minutes

Champion: Jonathan Wakely
Chair: Inbal Levi / Andreas Weis
Minute Taker: Bronek Kozicki

Summary

POLL: Add the functions from “math.h” header to the rebasing paper: nextup, nextupf, nextupl, nextdown, nextdownf, nextdownl

Attendance: 26 (IP) + 5 (R)
Outcome: No objection to UC

POLL: We approve SG6’s recommendation: drop the math addition for rounding and exceptions.

Attendance: 26 (IP) + 5 (R)
Outcome: No objection to UC

POLL: For “realloc” modify when receiving 0 argument for new size, from UB to unspecified EB.

Attendance: 26 (IP) + 5 (R)
Author's Position: SF
Outcome: Strong consensus in favour

A: Compatibility with C is my reason for voting against

ACTION: Update wording to include the C macros into the paper.
ACTION: Have SG22's input on this decision.

POLL: Apply the changes above and forward the updated P3348R3 to LWG for C++26.

Attendance: 26 (IP) + 5 (R)
Author's Position: SF
Outcome: Strong consensus in favour

Next Steps

Apply changes above and forward P3348R3 to LWG for C++26.

P3348R3 C++26 should refer to C23 not C17 (Jonathan Wakely)