P1228 A proposal to add an efficient string concatenation routine to the Standard Library

[jensmaurer]

P1228R1 A proposal to add an efficient string concatenation routine to the Standard Library (Jorg Brown)

See also P1479 ostringstream wrapper #235.

[tahonermann]

SG16 reviewed this paper at our February 13th, 2019 telecon. Meeting notes are available at https://github.com/sg16-unicode/sg16-meetings#february-13th-2019

We conducted one poll to give guidance to the author.

Poll: Do we want to restrict arguments to strings, characters, and types convertible to strings?

SF  F  N  A SA
 3  4  1  0  1

Chair's poll interpretation: SG16 advises restricting concatenation to string and character types initially with the option to add support for numeric and other types later. The chair advises caution regarding these poll results; the SA vote reflects the author's well-informed long term experience with the feature.

Removing the SG16 label.

[brycelelbach]

Discussed at the Kona 2019 meeting by LEWGI

P1479R0 ostringstream wrapper

Champion: Mike Spertus

Minute Taker: Ben Craig

Start Overview: 16:00

End: 16:17

P1228R1 std::concat

Champion: Jorg Brown

Minute Taker: Ben Craig

Start Overview: 16:17

Start Polling: 16:45

POLL: We should promise more committee time to pursuing P1479R0 ostringstream wrapper, knowing that our time is scarce and this will leave less time for other work.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
0	1	8	6	1

Attendance: 17

That has no consensus.

Peter Bindels (SA): Easy to do yourself, not sure it’s a good use of time.

POLL: We should promise more committee time to pursuing P1228R1 efficient string concatenation, knowing that our time is scarce and this will leave less time for other work.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
2	9	3	2	0

Attendance: 17

That has consensus.

Peter Bindels (A): Not sure this is a good fit for std::, use case may be too niche.

Bryce (A): Too many of these things; feel like fmt is good enough, and constexpr fmt may be a thing in the future.

POLL: We want to see a revision of P1228R1 that just concatenates things currently convertible to string_view and chars.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
0	5	4	2	3

Attendance: 15

That has no consensus.

End: 17:00

CONSENSUS: We will not pursue P1479R0.

CONSENSUS: Bring a revision of P1228R1 to LEWGI for further design review.