4 ms·
"Was it really worth it?" - is better than nothing, at least something happens. But things like only providing "just core" protocol and then functionality depen
by mrktf 3y ago
"Was it really worth it?" - is better than nothing, at least something happens. But things like only providing "just core" protocol and then functionality depends on compositor I would say is less than ideal architecture.
There are some more criticism on https://itvision.altervista.org/why.linux.is.not.ready.for.the.desktop.current.html https://itvision.altervista.org/why.linux.is.not.ready.for.t... go to "Wayland and its compositors", I think is pretty accurate for wayland (sorry, I use xfce environment on X11 and do not use wayland everyday).
Probably nothing new from me, but I imagine that replacement should come from team with same attitude which created pipewire (obviously nothing is perfect, but for me it really felt as modern replacement) and tough about following functionality:
* multi-monitor, 30bpc, variable refresh rate, fractional scaling and etc;
* in architecture foresee RDP, accessibility use cases;
* requiring session manager to implement common desktop services: clipboard, hook on screensaver/lock, global keyboard shortcut integration, light/dark whatever theme controller, battery saving mode and everything that today is required for common desktop (maybe mobile) functionality (I would take inspiration from Android, IOS, Windows);
* have minimal widget and dialog set (it can be session manager requirement): message box, file open/save, simple window with buttons;
* allow server side widget processing/display – obviously it would decouple GUI interfaces even further, but with smart RPC programming it will have seamless RDP integration and in limited cases allows to have modern widgets displayed with only server side glue library.