KDE Ends Out October With More Fixes, Continued Polishing To Plasma Wayland

[bple2137]

still useless with multi screen for me, keeps placing my icons and tray on the wrong monitor.

The good thing is, KDE devs know that it's bad for now and are addressing the issue. My guess is that it will take few more releases to sort it out.

HiDPI handling for Xwayland apps is what I miss so I'll stick to GNOME on one of my machines for a bit longer even if they'll improve multi-display setups.

[bug77]

Great! Since 5.23, I'm all-in on Wayland now and it's working quite nice.

It's starting to be usable for me, too, but it's not "quite nice" yet. Some maximized windows weirdness (supposedly already fixed, waiting for the next point release), can't drag-and-drop windows in the taskbar, no windows preview in the taskbar (it's all black).
But a few show stopper bugs have been dealt with recently, so kudos for that.

Edit: Scaling also sucks, but that's on Wayland, probably won't be fixed a year from now

[bug77]

Similar for me, but with krunner and full-screen programs.

These multiscreen placing issues have to be solved to make Wayland a real full replacement for Xorg.

It's on the KDE Wayland's list of blockers. It doesn't restore Wayland native windows at all, which is super-annoying since I'm used to having about a dozen tabs open in Konsole, which I now lose on every restart

[TemplarGR]

It is great that KDE has fully embraced Wayland at last. It is improving fast and is quite usable today with only a few major issues left to fix. Now with Nvidia also playing ball with GBM, we can expect Wayland to dominate the Linux desktop soon. Next stop: Pipewire...

[ALRBP]

It is great that KDE has fully embraced Wayland at last. It is improving fast and is quite usable today with only a few major issues left to fix. Now with Nvidia also playing ball with GBM, we can expect Wayland to dominate the Linux desktop soon. Next stop: Pipewire...

From what I heard, Pipewire already works pretty well. While I am sticking to Xorg due to Kwin's Wayland implementation still not managing multiscreen properly, I am seriously thinking about switching to Pipewire.

[er888kh]

So if I start an infinite loop of icon loading at startup, Plasma will asymptotically approach infinite speed and zero memory over time. Brilliant!

It might instead approach a limit, and be still strictly decreasing/increasing. Kind of like Basel problem.

[Random_Jerk]

From what I heard, Pipewire already works pretty well. While I am sticking to Xorg due to Kwin's Wayland implementation still not managing multiscreen properly, I am seriously thinking about switching to Pipewire.

KDE X11 also has issues with multi-monitor support. It still doesn't work with monitors of different refresh rates, it pretty much forces the higher refresh rate monitor to work at 60 Hz or whatever the lower monitor works at. On KDE Wayland, it works perfectly, but is marred by the issues mentioned here. It's a 'pick your poison' type situation right now.

[scottishduck]

It is great that KDE has fully embraced Wayland at last. It is improving fast and is quite usable today with only a few major issues left to fix. Now with Nvidia also playing ball with GBM, we can expect Wayland to dominate the Linux desktop soon. Next stop: Pipewire...

For most regular users, you could go all pipewire right now and not have any issues. Have been doing so for some time.

[Nth_man]

> still useless with multi screen for me, keeps placing my icons and tray on the wrong monitor.

The good thing is, KDE devs know that it's bad for now and are addressing the issue. My guess is that it will take few more releases to sort it out.

Yesterday that was solved:

Bug 296673 - Preserve the window geometry relative to its original screen instead of globally when adding a screen

https://bugs.kde.org/show_bug.cgi?id=296673

https://invent.kde.org/plasma/kwin/-...c23ff95d1a6781

[TemplarGR]

For most regular users, you could go all pipewire right now and not have any issues. Have been doing so for some time.

I know, but what i meant was for it to become the defacto standard of the Linux desktop. That way apps may begin to use it instead of pulseaudio emulation.