Experimental Wayland Support For Wine Now Sees More Functionality Working

[phoronix]

Phoronix: Experimental Wayland Support For Wine Now Sees More Functionality Working

Back in December there was an experimental driver for native Wayland support within Wine published by Collabora developer Alexandros Frantzis. A new version of the Wayland patches for Wine have now been published...

http://www.phoronix.com/scan.php?pag...e-RFC-Feb-2021

[Alexmitter]

I already played around with the original commit state from the last announcement and it was already working very well, can't wait to see it being merged.

The people at collabora are just amazing.

[duby229]

I just would like to point out that this argument that Wayland support somehow will lead to improved performance simply isn't true. Improved performance hasn't been demonstrated at all, not even once in all these years. It's just not true.

[RahulSundaram]

I just would like to point out that this argument that Wayland support somehow will lead to improved performance simply isn't true. Improved performance hasn't been demonstrated at all, not even once in all these years. It's just not true.

I haven't seen anyone actually working on Wayland claiming it has better performance or even that it was a goal. Have you?

[Sethox]

I haven't seen anyone actually working on Wayland claiming it has better performance or even that it was a goal. Have you?

I know right, people tend to miss the point after explicitly saying "security reasons". For convenience sake, people tend to push security to the side just so they can have that working matter/something on their screen.

[Zan Lynx]

I just would like to point out that this argument that Wayland support somehow will lead to improved performance simply isn't true. Improved performance hasn't been demonstrated at all, not even once in all these years. It's just not true.

If you have decided to use Wayland for any reason then removing the jump through XWayland will improve performance. Maybe not by a lot. But by some.

I find it remarkably silly when there's a bunch of code which has to make several context switches through the kernel simply to pass a message along a chain of file descriptors. It happens a lot. But it is still silly.

[Drol]

Even in that case not the acceleration with NVIDIA?

[aufkrawall]

If you have decided to use Wayland for any reason then removing the jump through XWayland will improve performance. Maybe not by a lot. But by some.

This is completely irrelevant for any dGPU, as Xorg fullscreen, XWayland fullscreen and Windows fullscreen are equally fast.
When you run Doom (2016) in Wine Xwayland with amdvlk-pro driver, it's still a few percent faster than on native Windows. That's the scale this is about in terms of performance.

[Azrael5]

Excellent!

[Alexmitter]

I just would like to point out that this argument that Wayland support somehow will lead to improved performance simply isn't true. Improved performance hasn't been demonstrated at all, not even once in all these years. It's just not true.

You have fallen into the trap of thinking "improved performance == better FPS".
This is about things like not piping around the picture 3 times before it hits the framebuffer, proper vblank timing, improved CPU usage, faster frame presenting. Those are all things that suck on X11 and are unfixable.
Playing a fast pace rhythm game on Xorg? Forget it.
Want to use a low power device like the Pinephone or Pinebook Pro with Xorg? Forget it, it sucks, its unusable while wayland desktops run fine.