Distributions quotes of the week

> Red Hat is working on resolving the remaining problems and gaps in the Wayland session.

Thanks a lot, I am looking forward to see these improvements trickle down through the ecosystem!

After coping for +100 days with the user experience regressions in Xwayland[1] (none of the bugs reported to Ubuntu (I know, not Red Hat) seems to have been fixed), I switched back to Xorg.
After 3 weeks, I don't regret switching back.

[1] https://lwn.net/Articles/910027/

Eh, removing X.org is not a great idea, given there are still significant useability regressions in Xwayland.

Maybe this is seen as a way to push so that the Wayland/XWayland issues are finally fixed. Many bugs seen using a Wayland session may ultimately be in the toolkits, the desktop environments and the application themselves, so forcing everything to Wayland may also make urgent the fixing. However, I expect it to be a big pain for the users, something bringing back memories of the KDE 3->4 transition, when distros stopped shipping KDE 3 forcing everyone on 4 "when it was obviously too early for the switch".

As of today I am still personally better served by Xorg. In my usage case there is really nothing "visible" that is so broken in Xorg to make it hard to use in the day to day experience (yes, I know about the invisible, security, etc), while there are many little and larger broken things that make using Wayland not impossible but certainly more painful, so the choice is obvious. I may be wrong, but I expect this to be the same for all those users not having special needs that make them crash against insurmountable problems of Xorg and I expect them to be still the majority. Maybe the problem is that, notwithstanding the criticism, Xorg still works too well ;-).

Again I may be wrong, but I get the impression that at least some of the real issues (as well as perception problems) of Wayland are, at least to some extent, self inflicted. From the beginning, Wayland gave the impression of delimiting a surface for the project much smaller than what X11 was covering and not proposing "one obvious way" to do all the rest that was needed (to say it in a Pythonic way). So, at least at the beginning, one either remained: with no way at all to do something; with one way of doing something reimplemented in 10 different ways in different projects, often in incomplete, incompatible ways; with things that had to be done relying on "pieces" of X11; with things abandoned altogether.

As a few examples, think:
- of the management of a session (application prompting to save work before logout, restoring of applications on login (gnome does it in one way, KDE does it only for X11 applications, some other DE, boh?);
- of the way to access individual applications remotely (until not long ago, the only reliable thing was to start apps in X11 mode and use the X11 wire protocol);
- of deciding whether to start an app in X11 or Wayland mode (every toolkit has its own different environment variable for this... or has this changed recently?)
- of deciding when more computer power is needed and the GPU frequency should be scaled up (which seems to be the reason for very weak performance on low end CPU/GPU combos, in comparison to Xorg)
just to mention a few things...

Wayland doesn't support color management, and won't for the foreseeable future.

Lack of an X11 display option makes a distro a non-starter for anyone needing
accurate display color, i.e. photography, desktop publishing etc.