P3148 R1 Formatting of chrono Time Values

[wg21bot]

P3148R0 Formatting of chrono Time Values (Alan Talbot)

[inbal2l]

2023-12-20 Library Evolution Telecon

P3148R0 Formatting of chrono Time Values

2023-12-19 Library Evolution Telecon Minutes

Champion: Alen Talbot

Chair: Inbal Levi

Minute Taker: Ben Craig

Summary

POLL: We would like the author to come back with a revision exploring the non-breaking change (use “%.0s”, “%.0S”, “%.0T”) proposed in P2945R0

SF	F	N	A	SA
4	8	3	0	0

Attendance: 18

# of Authors: 1

Author Position: SF

Outcome: Strong consensus in favor

POLL: We support applying the breaking change against C++20 and C++23 (repurpose "%S", "%T") proposed in P2945R0

SF	F	N	A	SA
0	1	0	4	8

Attendance: 18

# of Authors: 1

Author Position: Abstain

Outcome: Consensus against

WF: The 99% option with seconds as two digits, lots of code to be written, shouldn't be afraid of breaking changes. I don't think this is the most important change, so only WF.

Conclusion:
The format specifiers were supported, but there were concerns about the breaking option and requests for more exploration of the potential issues that may come up.
No consensus for the breaking change, the room would like to see an additional revision of the paper with an exploration of the non-breaking change. Alan Talbot volunteered to assist with the new revision.
The usage of %f brought up concerns from the author of ::fmt, author to provide a revision with:

• Proposing the non-breaking change
• Addresses the concerns which came up in regards to %f
• Please make sure the author of P3015 and the implementor of ::fmt give feedback on the new revision before coming back to LEWG

From last meeting: Needs a feature test macro

Next Steps

Author to come back to LEWG with R1 with the non-breaking change (make sure to capture and address concerns regarding usage of %f from ::fmt author).

[inbal2l]

2024-11-18 Library Evolution Wroclaw Meeting

P3148R0 Formatting of chrono Time Values

2024-11-18 Library Evolution Wroclaw Meeting Minutes

Champion: Alan Talbot
Chair: Inbal Levi / Fabio Fracassi
Minute Taker: Robert Leahy

Summary

POLL: We see the need for additional time_point specifiers along the lines of what is described in [P2945R1] or [P3148R0]

SF	F	N	A	SA
5	9	2	2	0

Attendance: 17 (IP) + 5(R)
Author's Position: SF
Outcome: Consensus in favor

A: I believe that the motivation is very weak. In general I'm not against adding more specifiers but need sufficient motivation.

A: An libc++ implementor, I don’t see enough motivation
POLL: We see the need for defining precision on time_point specifiers along the lines of what is described in [P2945R1] or [P3148R0]

SF	F	N	A	SA
8	9	1	1	0

Attendance: 17 (IP) + 5(R)
Author's Position: SF
Outcome: Consensus in favor

A: I think that both papers address it in the wrong way but can be fixed in the current wording and may write a paper

Next Steps

Both the need for additional time_point specifiers and the need for precision got supported.
During the next LEWG meeting, we will go over each of the specifiers and consider which we would like to add.