P0959 A Proposal for a Universally Unique Identifier Library

[jensmaurer]

P0959R2 A Proposal for a Universally Unique Identifier Library (Marius Bancila, Tony van Eerd)

https://issues.isocpp.org/show_bug.cgi?id=434

[brycelelbach]

Kona 2019-02 LEWGI Minutes

P0959R2 std::uuid

Champion: Corentin Jabot

Minute Taker: Steve Downey

Start Overview: 9:03

Security concerns: Don’t allow unsecure variants, e.g. anything prior to version 4.

Start Discussion: 9:10

Start Polling: 9:40

POLL: We should promise more committee time to standardizing a UUID library, knowing that our time is scarce and this will leave less time for other work.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
5	8	1	0	0

Attendance: 16

CJ (N): Concerned about the concept of standardizing UUIDs, given the potential security concerns.

POLL: Add an istream operator for std::uuid.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
2	2	7	2	1

Attendance: 16

POLL: Replace overloads taking char const and std::string with one overload that takes string_view, and make the new string_view overloads constexpr.__

NO OBJECTION TO UNANIMOUS CONSENT

Attendance: 16

__POLL: Make as_bytes noexcept.__

NO OBJECTION TO UNANIMOUS CONSENT

Attendance: 16

Question to Author: Why doesn’t this just work with char or char8_t? Why is it templated on CharT?

Question to Author: Comparisons?

POLL: std::uuid_random_generator’s should use a “see-below” UniformRandomNumberGenerator (instead of std::mt19937), whose semantics WG23 will help define.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
4	5	1	1	0

Attendance: 16

Corentin (A): Concerned this will make it unimplementable on some platforms. We should solve this problem generally, not just for this class.

__POLL: Remove std::uuid_random_generator and just have std::basic_uuid_random_generator<UniformRandomNumberGenerator> (e.g. don’t have a default).__

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
0	6	3	2	0

Attendance: 16

Volunteers to Help Revise: Steve, Corentin

End: 9:50

[cor3ntin]

@brycelelbach you remember why this has an ewgi tag?

[klappdev]

Hi. A very useful feature. What are the chances that will be added to C++26? Thank you.

[tahonermann]

@klappdev, these issues are used for administrative tracking purposes and are not intended for community discussion. Please feel free to share your thoughts on the std-proposals mailing list.