GNOME developer proposes removing the X11 session
The suggested change is the first step in desktop environment becoming Wayland-only
The two changes are just proposals at present, but GNOME's Wayland-only future is on the horizon. Whether that's a good or bad sign is less clear.
Jordan Petridis' two merge requests in the GNOME's source code in Gitlab are the most definite signs yet that GNOME is getting close to dropping support for running on what its developers consider the legacy X11 display server. The first change removes the
gnome-xorg.desktop file. As Petridis says:
X11 has been receiving less and less testing. We have been defaulting to the wayland session since 2016 and it's about time we drop the x11 session completely.
The support would still be present, so if a user replaced the file, they would still be able to start a GNOME-on-X.org desktop session. However, his second proposed change is more sweeping. While still marked as a "Draft," and probably aimed at the following release of GNOME, it proposes the removal of the code for running GNOME on X11 at all.
To put this in context, the Fedora project is considering a comparable change: removing or hiding the GNOME on X.org session from the login menu, which is already the plan for the Fedora KDE spin when it moves to KDE version 6, which is still in development. These changes are under discussion for Fedora 40, the future version which will follow Fedora 39, currently scheduled for release next week.
It's not really a question if this is going to happen. The big question is when… and the smaller one is, it seems to us, why.
Currently Wayland still has issues in a number of areas, including accessibility tools (such as screen readers and magnifiers), DRM leasing, headless remote desktop (as this still needs a monitor attached), as well as the variable refresh rate issues we talked about last year, among others.
Yes, Wayland does offer some handy new features, although for this particular vulture, there is only one that is of any use whatsoever: the ability to run two (or more) displays at different scaling factors. This article is being written on a machine with two 27 inch displays, but one has a resolution of 5120 × 2880 while the other is 2560 × 1440. That means that one has four times more pixels than the other: twice both vertically and horizontally. For this to work without things suddenly and dramatically changing size when you move them from one screen to another, one screen must run at double the pixel density of the other.
Although X.org can happily run two or three displays at different resolutions without a problem, it can't handle per-screen pixel densities: for now, that's a global setting.
This change will have both positive and negative ramifications.
For most people, it would almost certainly be fine, and make little to no visible difference that most people would ever notice.
There are some GNOME – and KDE – users for whom it would be bad news though. One large and quite vocal group are people who use Nvidia cards with the proprietary Nvidia binary drivers. Another are folks who need accessibility tools to use their computer.
For now, Wayland only works fully on Linux and FreeBSD. There are experimental projects to port it to OpenBSD, and indeed to NetBSD and to DragonflyBSD. These are in their early stages, though, so this will mean the end of both KDE and GNOME on NetBSD and other OSes, at least for now.
As for other Unix variants, especially the various proprietary ones, we predict that GNOME will never return to them. Long ago, Solaris adopted GNOME, even making it the default, although even then it was controversial.
Today, though, the age of proprietary UNIX workstations is gone, and the various FOSS Unix-a-likes are UNIX. None of the BSDs have desktops and laptops as their primary goal, and their users would probably favor simple window managers over full-fat desktops anyway.
There are over 20 Linux desktop environments out there, and a large majority don't support Wayland. Some of these are based on GNOME components, including Mint's Cinnamon, Solus OS's Budgie, Elementary OS's Pantheon, and the recently reinvigorated Unity, and this change will ultimately pose problems for all of them – although of course it may also cause their development teams to prioritize adding Wayland support. Budgie is expressly working on it, and Xfce is close behind.
So, conversely, if the big two go Wayland-only, that puts more distance between them and the other couple of dozen desktops that remain X11-only. This change could be good news for the smaller players: they will end up having better cross-platform support.
We have no usage numbers, but there are also many – maybe millions – of Chinese users with either Kylin's UKUI or Deepin's DDE. Both of these only run on X.org. If China's 3-5-2 program has gone as planned, it could be that GNOME and KDE are not nearly as dominant as they look in the West.
(Tilting the balance the other way, ChromeOS has its own display server, Exo – and for the record, it is Wayland-based. We suspect ChromeOS's Aura Shell has more users than all the other Linux GUIs put together. And of course ChromeOS is in turn dwarfed by Android's billions of users; it also has its own display server, SurfaceFlinger.)
This plan seems to The Reg's FOSS Desk to be strong-arming people into adopting Wayland. To us, it's reminiscent of the sort of moves proprietary software vendors execute to compel users to upgrade, such as file format changes.
As an example, Microsoft has done it twice with Word alone. From MS Word 1.0 in 1983 up until Word 6 and the almost entirely cosmetic change of Word 95 in Office 95, it used one file format, called
.DOC. Then, the buggy and troubled Word 97 brought in a new one, to muddy the waters still called
.DOC, which remained until Office 2003. There was little new functionality in Office 97 aside from Outlook and, of course, the universally loved Office Assistant, by default the famous Clippy.
Even if you were — like the author – perfectly happy with Office 95, if the people you worked with – family, co-workers, or clients – had the new version, you pretty much had to upgrade, because you would otherwise no longer be able to open any files they sent you. The tactic worked, so Microsoft did it again, with the new
.DOCX format in Office 2007… presumably realizing that its new "ribbon" UI would hinder adoption.
- Red Hat retires mailing list, leaving Linux loyalists to read between the lines
- Researcher bags two-for-one deal on Linux bugs while probing GNOME component
- Ubuntu's 'Mantic Minotaur' peeks out of the labyrinth
- GNOME 45 formalizes extensions module system
Intentionally creating incompatibility and interoperability difficulties with older versions or entrenched products is a time-honored way of getting reluctant users to adopt new ones. That's why we wrote that you cannot buy software
For all that the GNOME communities talk about the lack of testing on X11, or the exciting new abilities of Wayland compared to X.org, even so, as a type of display server, it's been around for some 13 years now. If it were as compelling as its advocates think it is, surely more people – and desktops – would have adopted it. They haven't, and so now, it seems that it has become necessary to force them. ®