We saw this paper in yesterday's EWG telecon:

POLL: we are interested in pursuing a way to specify that multiple/all class members are constexpr, either as suggested in this paper or through another mechanism, considering that time is limited and there is only so much work we can do.

SF	F	N	A	SA
2	5	2	2	0

Result: consensus.

POLL: support class-level consteval at the same time or before supporting constexpr, to inform the design of class constexpr.

SF	F	N	A	SA
2	8	3	0	0

Result: consensus.

POLL: use access specifier constexpr instead of class-level constexpr.

SF	F	N	A	SA
0	1	3	4	5

Result: no consensus.

POLL: class-level constexpr should also affect friend functions.

SF	F	N	A	SA
0	1	2	5	1

Result: no consensus.

POLL: it should be ill-formed to put constexpr on a class which can’t be entirely constexpr (e.g. because of its base or data members not being constexpr).

SF	F	N	A	SA
0	4	2	4	1

Result: no consensus.

POLL: static data members should also be constexpr in a constexpr class.

SF	F	N	A	SA
3	4	0	1	0

Result: consensus.

POLL: allow constexpr and final to appear in either order.

SF	F	N	A	SA
2	6	4	0	0

Result: consensus.

P2350 constexpr class #1028

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions