https://bugs.kde.org/show_bug.cgi?id=415262

--- Comment #6 from tempel.jul...@gmail.com ---
(In reply to Roman Gilg from comment #5)
> Thanks for testing! Early feedback like yours is super helpful. Would be
> great of course if you could stay on master in the future to continue with
> feedback if something breaks and by that shield inexperienced users from
> bugs they would not able to recover from on their own. :)
I think it would be helpful if it was less hassle due to a lot of git master
dependencies. It's really painful to compile kwin-git on Arch, while e.g.
gnome-shell-git or xfwm4-git are easy to compile. There are so many KDE
packages, I'd suppose that there is a lot of potential for combining them into
fewer.


> I need to look into the code some more for a final patch. The above one was
> just done rather quickly to test the assumption.
I'll stick with the git packages until 5.18 is released and report back in case
of oddities, no big deal.

> I understand your sentiment but from my perspective I would need to dig into
> how exactly unredirect is supposed to be implemented and even then to my
> knowledge there are certain structural constraints of unredirect that make
> it a subpar experience.
> 
> On the other side we have the Wayland session where in the future we will be
> able to do an optimal direct scanout of client buffers and on the other side
> still have lots of other things to do to make it work on every user install
> without issues. So I would like our devs to concentrate on improving that
> rather.

I totally understand your situation.
Perhaps suspend of compositing could be improved instead? Maybe my thinking is
too simple, but wouldn't it work if KWin polled e.g. every 0.5s if there was a
fullscreen window covering everything else, and in case it isn't (e.g. after
alt + tabbing out of a game), resume compositing (and vice versa)?
The result basically should be the same of unredirecting, without the need to
actually implement it.

I think I also found a bug of KWin compositing in conjunction with Proton
fullscreen hack patch for Wine, where the fullscreen window just stays black
after alt + tabbing out of it and back into it. This does not happen when I
disable compositing before. Gonna open a new report for it.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to