P1779 ABI isolation for member functions

[wg21bot]

P1779R0 ABI isolation for member functions (S. Davis Herring)

[opensdh]

If I understand the process correctly, this should have the "modules" label.

[wg21bot]

P1779R1 ABI isolation for member functions (S. Davis Herring)

[jfbastien]

This is potentially for C++20, but SG2 should see it first.

[opensdh]

This is potentially for C++20, but SG2 should see it first.

SG2 did look at it in Cologne. It recommended it for C++23, but then other things happened. Ville said that EWG chose the ship vehicle, and so EWG also looked at it in Cologne. The discussion was inconclusive; my memory is that Ville said something about "I'll make it a DR". Later, Richard pointed out that it would be a silent performance regression in C++23.

[jfbastien]

Ah indeed, SG2:

Forward D1779r1 to EWG for C++20
SF F N A SA
4 3 0 8 0
Forward D1779r1 to EWG for C++23
SF F N A SA
5 3 5 2 0

Conclusion in EWG was "Now it is time to go away with this and come with a DR later. Once C++23 opens"