P2892 R0 std::simd Types Should be Regular

[brycelelbach]

P2892R0 std::simd Types Should be Regular (Joe Jevnik, David Sankel)

[inbal2l]

2023-06-13 Library Evolution Varna

P2892R0: std::simd Types Should be Regular

2023-06-13 Library Evolution Minutes

Champion: David Sankel

Chair: Bryce Adelstein Lelbach, Fabio Fracassi

Minute Taker: Inbal Levi

Summary

POLL: simd::operator== should be an equality comparison that returns bool, not a mask.

Strongly Favor	Weakly Favor	Neutral	Weakly Against	Strongly Against
2	1	2	5	17

Attendance: 25 (P) + 6 (R)

# of Authors: 2

Author Position: WF (P2892), SA (P1928)

Outcome: Consensus against.

POLL: Remove simd::operator==.

Strongly Favor	Weakly Favor	Neutral	Weakly Against	Strongly Against
3	3	5	5	10

Attendance: 24 (P) + 6 (R)

# of Authors: 2

Author Position: SF (P2892), SA (P1928)

Outcome: No consensus for change.

Does this paper have:

• Examples?
  • Yes.
• Field experience?
  • Implementation experience?
    • Yes, in a Technical Specification with a reference implementation.
  • Usage experience?
    • Yes, in a Technical Specification with a reference implementation.
  • Deployment experience?
    • Yes, in a Technical Specification with a reference implementation.
• Performance considerations?
  • Yes.
• Discussion of prior art?
  • Yes.
• Changes Library Evolution previously requested?
  • Yes.
• Wording?
  • Yes.
• Breaking changes?
  • No.
• Feature test macro?
  • Yes.
• Freestanding?
  • Discussed in the paper.
• Format and/or iostream support?
  • Discussed in the paper.
• Hash support?
  • Discussed in the paper.
• Allocator support?
  • N/A.

Next Steps

The polls suggests that there's no support for removing the comperason operator. The paper did not got support.

[inbal2l]

The change suggested in the paper was not supported. The authors will re-battle this discussion in a separate paper.