Re: [PATCH 0/2] MR9573: win32u: Initialize surface with white colour on creation.

[Paul Gofman (＠gofman)]

I hope in this form (with explicit DWM bits) it should be reliable. That succeeds on all the Windows Testbot machines starting from Win8 (while the colours on Win7 suggests that the issue there might be is it still does transition but doesn't have dwmapi to to control that). I am afraid that in case of some regressions with this change figuring out neater conditions for this check starting from scratch, without having a base test will be harder. Can we give it a chance? And in worst case if it does prove to be flaky update or just remove later?