Ubuntu 25.10 "Questing Quokka" removes Xorg-based sessions, fully embracing Wayland as the new default for GNOME
So, what does this mean for X11 going forward? Let’s be honest—this is likely the end of the road for X11. With Ubuntu now joining other major Linux distributions in dropping X11 for Wayland, the aging display server has almost no mainstream users left. And without users, there’s little reason for further development.
Love it? Hate it? Or are you indifferent?
Frankly, I do not care until it ruins my
Since i rely 99.99% on remote desktop access to my Ubuntu server, changing to Wayland may break how things work. I run xrdp on my server to achieve my remote access. Guess what the 'x' stands for...
So far it's not Ubuntu dropping support for X11/Xorg; just Ubuntu Desktop as GNOME, Mutter & gdm have opted to drop support in newer versions.
X11/Xorg will still be available for 25.10 & is expected for 26.04 LTS if you're using other desktop/WMs. Other desktops may also opt to drop X11/Xorg, but we'll have to wait and see, as Ubuntu 26.04 is still some time in the future.
My reaction was stunned silence though when I read the discourse post.. though to be honest it was very much expected; the question really for me was was when the news would drop & how I'd feel about it when it dropped.
Evaluating my reaction; I'm feeling loss, the stunned silence I was describing being denial, or the common first stage of grief; and knowing it was coming didn't really help that much.
Being retired, I no longer use remote access terminals.
Do you know of any "workable" substitutes for xrdp, xterm, etc, that exist out there to operate on a Wayland-based DM?
If we can be assured that UbuntuMATE 26.04 will continue support for Xorg/X11, then at least it will offer the Community with much needed breathing room for a non-panic response to identify and migrate various necessary components that need to be re-targetted for Wayland-based operations.
Eugene, I guess this means that the Community needs some clear communication from the UM Development Team regarding how they plan to respond to the underlying changes.
Specifically, identifying
will form the foundations for the obviously-implied component rebuild, where such is required.
Their own sense of timeline for that migration would also help alleviate the big unknown, and what may become a growing concern, regarding individual/organizational operational continuity.
Also, to give us a sense of the current exposure/risk from this decision, has anyone compiled a report of which applications are dependant upon the X11 technologies?
[Edit:]
Apparently, we can look for dependencies on our own computers using the following command:
apt-cache rdepends --installed ${packagename}
I just don't know which is the most basic package that controls any X11 installation. Looking at what Synaptic reports for installed X11 references (UM 22.04.5), I have the following:
Not knowing enough about the X11 architecture, my guess would be to look for all uses of each of
Opinions, anyone? Search on the first offers this:
libx11-6
Reverse Depends:
arctica-greeter
argyll
at-spi2-core
audacious-plugins
ayatana-indicator-keyboard
bamfdaemon
blender
brltty-x11
caja
clementine
compton
conky-all
dbus-x11
eom
firefox
freeglut3
ghostscript-x
gimp
glew-utils
gmemusage
gnome-session-bin
gnome-settings-daemon
gnuplot-qt
graphviz
gstreamer1.0-clutter-3.0
gstreamer1.0-gl
gstreamer1.0-plugins-bad
gstreamer1.0-plugins-good
gstreamer1.0-vaapi
gstreamer1.0-x
gtk-3-examples
hwloc
inkscape
kinit
lib* (104 packages)
lightdm-gtk-greeter
marco
mate-applet-brisk-menu
mate-applets
mate-control-center
mate-indicator-applet
mate-netbook
mate-notification-daemon
mate-panel
mate-power-manager
mate-screensaver
mate-session-manager
mate-settings-daemon
mate-terminal
mate-utils
mesa-utils-bin
mpv
mupdf
numlockx
octave
onboard
openjdk-11-jre
openjdk-17-jre
pango1.0-tools
pcmanfm
perl-tk
picom
plasma-framework
pluma
pulseaudio
pulseaudio-utils
qt5-gtk2-platformtheme
qt5-gtk-platformtheme
recollcmd
redshift
rhythmbox
scrot
slick-greeter
spacenavd
spice-vdagent
texlive-binaries
thunderbird
tilda
tk8.6-blt2.5
tkblt
tkdnd
tk-html1
tk-html3
tk-table
vim-gtk3
vlc-plugin-qt
vlc-plugin-video-output
vulkan-tools
w3m-img
winff-gtk2
wmctrl
x11-apps
x11-session-utils
x11-utils
x11-xkb-utils
x11-xserver-utils
xauth
xbrlapi
xdotool
xinit
xinput
xkb-data
xli
xserver-xorg-input-wacom
xserver-xorg-video-intel
xserver-xorg-video-vmware
xterm
zenity
The article mainly concerns GNOME's plans, and Ubuntu aligning with that. Nothing worth a panic! 
Wayland will inevitably be the default, but I reckon X11 will still be around for many more years in maintenance mode.
KDE, the other big desktop environment, will support X11 session until Plasma 7. They split their compositor into two earlier this year, since one caused bugs for the other.
Since Wayland itself is just a specification, each implementation happens in the compositor (KDE = KWin, GNOME = Mutter) and I think the underlying toolkits need to support it too (GTK 3+, Qt 5+).
For MATE, I think it's going to be mate-wayland-session (based on Wayfire, inspired by Compiz)
Wayland isn't really a 'drop in' replacement for X11. Sure, it's modern, bit more secure, smoother with screen switching, but not perfect (yet). KDE keep track of "significant issues" affecting them.
I use X11 at home and Wayland at work. I notice subtle differences like:
I do personally prefer X11. 
The old timer's done so well.
I have no special knowledge in this regard, I'm not a Ubuntu MATE developer, in fact on launchpad you won't find reference to me and Ubuntu MATE in anyway (just other Ubuntu teams).
I am aware of the following though
One benefit of the lack of progress though is less and slower change; so that probably means Xorg is a safer bet than Wayland, but 26.04 is still ~1.5 cycles away.
The more the day of switching things to wayland approaches the more I wonder is that focus on "security" in this case a worthy trade-off since I more often than not interpret the call-for-security as a subtle whistle of a "less freedom" approach that takes away the choice from the user and strips some important layers that we all got used to in favour of minimalist sterility that can be tracked/managed more easily because it offers less flexibility more than because it's sturdier (at least I see it as such)
I do hope that when MATE does integrate the wayland we won't lose options that we love to use, as we could all see in other recent threads how many people still prefer old themes that were replaced with yaru and old software boutique/greeting screen as well...
Changes can be good when they respect the foundations they are built into, not when they just smash the old and force in the new as we sometimes can see happen
I really doubt X11 will suddenly disappear from Ubuntu any time soon — the project is still supported upstream and still releasing updates — it's not like it's abandoned or suddenly a huge liability risk.
X11 is just a little old, but currently faces this 'new kid on the block' wanting to take over its duties. Yeah... but can the new kid run GUI programs over SSH?? Not sure, but X11 forwarding was pretty neat!
Maybe in 5-10+ years, Wayland might reach feature parity and serve the modern display needs (VRR, HDR) while older incompatible programs remain working through xwayland.
There is another….
Here's an article addressing remote connections to Wayland written last October (2024):
I love it! 
Well, maybe not love, but I do think there needs to be some balancing out of sentiment.
For those who missed the negative sentiment, there's always the anger-transformed-into-a-GitHub gist: Think twice about Wayland. It breaks everything! · GitHub ... and more recently, the introduction to the X11libre xserver README - which, makes me feel a little bad for the person who wrote it, because they clearly feel quite outcast.
I don't want to over-simplify, but much of the strongest negative sentiment reads to me like "Those corporations should be contributing to X11 instead of pushing Wayland on us. 
"
I disagree.
Those corporations (e.g. RedHat, others) have poured considerable resources over a long period of time into maintaining the X Window System. An employee of RedHat sits on the dang X.org board. Developers within those corporations had to make a business case (likely not easily) in order to green-light resources for Wayland; they probably wanted badly to do this for all the reasons that they outline in the motivation for Wayland. Those motivations seem relatively sound.
Yeah, I certainly prefer not-for-profit organisations, but I can't see anything particularly egregious about the corporations' involvement in this instance.
Where I do agree with the critics is that end-users (like myself) are entitled, at least, to a fair bit of noise-making when these changes disrupt the day-to-day activities of our desktops or servers etc. If features that we rely on are disabled, without plans to close those gaps, we are within our rights to advocate loudly for re-implementation. I'm glad to see a lot of the 'gaps' are being plugged, e.g. network transparency efforts like Waypipe seem to be hitting their stride.
Finally, for me at least, Wayland (compositors) look about ready for use. Maybe some things will break, and maybe I'll find things I like better, as other X-apologists have.
I'm happy for X11/X Windows to take another step towards retirement. It had a great run and we accomplished a lot with it!
I'm just saying that just because the parent distro is choosing to cut us off despite our desktop not yet being fully ported to Wayland doesn't mean we have to suffer through old or bot rotted versions of the x-window display. If the X11 libre manages to prove itself a suitable replacement with fixed bugs and working code, why shouldn't we switch to it?
Why stick with old unsupported code?
Thats an enormous "if"; and X11Libre, while it has one very capable contributor, is currently less supported than the Xorg edition. It has no community nor organisation behind it (yet). It needs to be packaged on Debian and then make its way through Canonical, too. My guess is MATE desktop will be running on Wayland and will be benefitting from the ongoing development and testing (of Wayland) that millions of users supply before even 1 user has tested X11Libre.
X11Libre is possibly only useful in extreme corner cases... maybe someone with particular hardware (probably older?) which is locked in to drivers that are specific to X11 AND also experiencing some problem with the existing Xorg X window server that is fixed in X11Libre?
Well, indifference certainly didn't top the list!
As one who likes Mate for its simplicity and traditional menu system so “everything you know is wrong” doesn’t happen very often, I can do without wiz-bang “cool” features. And don’t really care about the engine underneath except:
Since in retirement my main use for a powerful desktop is learning more about AI and improving the AI add-on system I’ve made for my home security DVR/NVR systems (I have 16 channel DVR, 8 channel NVR, and a couple of standalone POE IP netcams) If Cuda doesn’t work on Wayland it is 100% useless to me!
See my Git Hub for some examples of it in action and give it a try if you’ve got Ubuntu and a USB camera (/dev/video0 etc.) or an IP netcam that suports RTSP streams or Onvif snapshots.
https://github.com/wb666greene/AI-Person-Detector-with-YOLO-verification-Version-2
I know installation is not for beginners or the feint of heart, but if you try and have problems, raise issues on the GitHub (or message me here) and I’ll try to help out and hopefully improve the installation instructions. At present I only support 20.04 and 22.04 because of Coral TPU driver issues, but I’ve recently found references to getting the TPU to work on 24.04 so I’ll be installing 24.04 asdual boot on my i7 laptop with RTX 3070 and M2 A+E key Coral TPU (I don’t really like WiFi if there is an option, so I removed it to install the TPU).
This is classic Open Source, I scratched my itch and am trying to share my results.