As the sun continues to set on the X11 display protocol, X.Org - the premier implementation - has been forked by a former developer who accuses its maintainers of “abandoning the project, and letting it rot forever.”
He’s not exactly wrong. X.Org is essentially mothballed. It is an enormous, complicated piece of deprecated infrastructure, with a very limited amount of resources and experienced maintainers. The corporations which sponsor Free Software development don’t particularly care about desktop end-users, and the resources which are being spent on desktop experience are largely being spent on Wayland compositors. On the other hand, it appears many of his commits on X.Org were reverted for sloppy management of licensing / attribution, as well as some regressions which were introduced.
It is worth noting that when Wayland was introduced in 2008, X.Org developers were among its biggest advocates and contributors. The writing has been on the wall for a long time now, and the work of building an alternative is mostly complete.
That said, Wayland is not at all a 1 to 1 replacement for X, and like with the introduction of Systemd, there are a lot of people with strong feelings about this, a lot of conspiracy mongers cranking out YouTube slop. People throwing out accusations about how “they” are trying to ruin Linux yet again.
I personally have fond memories of X. Especially in the later days when the whole “unix porn” phenomenon bloomed and there was a sort of renaissance of customization. I miss herbstluftwm terribly. That said, I’ve been running Wayland for something like 6 years now and I do not really get why people hate it. It works fine, and it actually has a future.
Update:
It’s also worth noting the author of this fork is a chud. Some excerpts from the README
This fork was necessary since toxic elements within Xorg projects, moles from BigTech, are boycotting any substantial work on Xorg, in order to destroy the project, to eliminate competition of their own products. Classic “embrace, extend, extinguish” tactics.
This is an independent project, not at all affiliated with BigTech or any of their subsidiaries or tax evasion tools, nor any political activists groups, state actors, etc. It’s explicitly free of any “DEI” or similar discriminatory policies.
Together we’ll make X great again!
Mfs would rather fork a decades old codebase which had its origins in mainframe computers than contribute to Wayland.
My opinion is that xorg needs to be ripped away and removed from package repos so that the collective desktop community is forced to support Wayland. Distros ripping out xorg is part of why Nvidia dropped their strong arming and actually started to work with the desktop community.
Most of Wayland’s recent success is when people start using it, realize “oh shit, this needs fixing” and new protocols are created, or that people actually start correctly supporting Wayland (using portals, pipewire) instead of assuming that xorg will do everything for them.
I fully second this. I’ve been using Linux as my main desktop OS since about 1998. I cut my teeth on Slackware 3.4. I’ve seen distros and important system software come and go. Wayland certainly needs work, and isn’t a 1:1 replacement for the X protocol, but it’s designed for the modern world. The X protocol (any implementation) is a relic of a bygone age. It was designed for running programs remotely over a trusted connection in an academic setting. It was a poor choice for a desktop OS where web browsers can facilitate malware coming in. X was only adopted for Linux back in the day because there wasn’t much else mature when desktop Linux at home started being a thing. We need to let the dumb thing finally die.
I think what forced Nvidia’s hand ultimately wasn’t the transition to Wayland as a default, but the increasing gap between Nvidia’s proprietary driver, proprietary EGLStream API, and proprietary OpenGL/Vulkan implementations, vs. what all their competitors got “for free.” AMD and Intel have drivers in the kernel, EGL is co-maintained by AMD and Intel among others, and OpenGL/Vulkan are implemented by Mesa3D. These were all able to evolve in tandem with one-another, while Nvidia was determined to (fail at) doing all of this in-house. It manifested in shit Wayland support, shit XWayland performance, and other problems, but this probably would have never changed without there being a vastly superior alternative.
IMO, the status quo is actually fine. X.org essentially exists to implement XWayland as a compatibility layer for old software. It is maintained, but there is no interest in developing new features because nobody in their right mind is developing new software targeting X11, let alone X11 extensions introduced in 2025! Even the toolkits, GTK and Qt are moving in the direction of deprecating X11 completely. GTK 5 will not support X11, and it is likely that Qt will also draw the line sometime soon. Within a couple years, the only way to develop X11 software will be by forking deprecated versions of these toolkits, and it will be about as sensible as developing an application with GTK+ 2 or Qt 3 today.
It has been a drawn out process, but I think it has been wise for Wayland to kind of let implementers cook on the protocols, rather than shit out a thousand pages of spec before any prototype has even been created. Some unique use cases like drag and drop between different applications, mouse pointer grabs/warps, etc. have suffered from this, but essential functions DO get implemented, if not universalized right away.
GNU imp developers in shambles.
fun fact, GNU Imp 2.xx was the sole reason why python2 was being kept in Fedora when every other python2 package was either upgraded to python3 or removed.
But even Wilbur got their ass to Wayland.
It’s finally been ported to GTK 3 as of earlier this year
What bothers me is that Wayland is at this point the sunk cost alternative.
We spent so long now trying to paper over the bad decisions in Wayland that we’ll now be forced to stick with it for another 30 years until it finally rots like X11. We missed the opportunity to say “real use cases need too many xdg-purple-monkey-dishwasher extensions, let’s wipe and restart knowing from day 1 users will want screen sharing and global media keys”.
I’d probably be less hostile towards Wayland if I could find a compositor I liked. It feels like my choices are:
I just want my desktop to look and work like a 1996 workstation running mwm or CDE (menu and icons on the root window, big chonky bevels everywhere, no client-side decoration BS) and I don’t see a convenient way to it with current compositor choices.
Fedora Workstation’s (GNOME) system reqs is 20 GB for recommended (whole OS), GNOME and KDE are lean and resource efficient and don’t “ape” windows and macos tendencies, they invented them while Bill Jobs was looking to squeeze more profit margins out of their users in a market they monopolized.
I understand your point about a smaller range of available desktops with Wayland, but that just comes with the territory of a newer and developing standard that requires more effort than the
#include <X11/Xlib.h>that came before it.You can probably get that CDE-style desktop with a well-designed KDE Plasma global theme. This image I got off unixporn shows the Aero design being used.
aero’s about a decade too modern for OP lmao
I’m baby
windows 7 was how far I go back.
I feel you on that last paragraph. Some time ago I came across this project which seems to be worth following. I’ve been running a hacked together Sway+Swaysome+Swaylock+Waybar combo for the past few years that is efficient for my workflow but it looks incredibly bland and can be a PITA to maintain. I really want a self-contained wl compositor that is customizable, has style to it and maybe even comes with some cool screensavers.