Don't drop X11 support after Plasma 6.7 (last release Feb 2027)

[Piotr H. Dabrowski]

Hi,

This is a followup thread from
https://mail.kde.org/pipermail/plasma-devel/2026-January/123858.html
Quoting myself again:

> with Plasma 6.7 being the last version with X11 support
> 6.7.7 2027-02-02

I think dropping support for an ACTUALLY WORKING, VERY STABLE, and
BATTLE-TESTED environment like Xorg/XLibre in favor of a solution where
even the underlying protocol design is still effectively in an alpha state,
valid use cases are broken, and basic core features are still missing and
still actively debated among developers, is pretty insane. Literally.

I would like to hear a MERITORIOUS reply to these concerns:

https://www.reddit.com/r/linux/comments/1pxectw/wayland_is_flawed_at_its_core_and_the_community/

https://www.semicomplete.com/blog/xdotool-and-exploring-wayland-fragmentation/

https://vas.neocities.org/wayland_shitware

https://news.ycombinator.com/item?id=44301194

Namely:
- no mouse-grabbing protocol; pointer-warp protocol is still broken
(required by virtually every game)
- screen sharing is still broken
- color management is still broken
- automation tools are broken (xdotool)
- global hotkeys broken
- accessibility features can't be properly implemented
- no sane capabilities system
- no standard way to communicate with the window manager (like EWMH)
- fragmentation hell: no DE is required to implement any protocol, no API
can be considered reliable
- every new feature requires a new protocol and multiple, DE-specific
implementations, which will take years to become even remotely useful
...and much more.
READ the actual articles; it's 10 minutes of work spent on actually
listening to the community.

Wayland is unusable for the majority of Linux users, based on the use cases
described and the numerous comments above.
Are you willing to take our voices into consideration, or will you continue
to ignore us?

Power-users, in particular, will be hurt by you ignoring our needs. KDE has
always been the desktop environment that supported us.
Do you remember "Simple by Default, Powerful when Needed"? With Wayland, it
feels more like "Broken by Default, Powerful Never".

I am fully in favor of replacing old X11 with something modern and better,
but Wayland is either not that solution, or at the very least, it is not
ready yet.

That said, there should obviously be NO END DATE SCHEDULED FOR X11 SUPPORT.
Maybe we can have this discussion again in 5 years or so, when Wayland is
(hopefully) ready and actually usable.

Once again: stop ignoring valid concerns and provide meritorious responses
instead of avoiding the topic. Otherwise, this will end badly for the
entire Linux community.

On Sun, 18 Jan 2026, 21:00 Nate Graham <nate at kde.org> wrote:

> It's a trend bigger than KDE, but we're a part of it too
> [...]
> The larger context here is that after decades of service, the world is
> moving on from X11.

It's not a trend at all.
It's an artificial and FORCED adoption of the practically unusable Wayland
protocol by dropping support for working X11 solutions and
breaking/removing perfectly stable and functional X11 software.
And still, despite all this sick effort of shoving Wayland down users'
throats, Wayland usage remains roughly on par with X11, both holding about
50% of the market share.

> almost no Plasma developers are still regularly using or developing the
X11
> session.

Maybe it's worth asking the ones who still use X11 why they do so? It's
potentially a very important question and a source of valuable insight.

> Upstream X11 development also largely ended years ago

X11 was intentionally killed. Fortunately, we now have XLibre, which is
under active development.

> However, I can appreciate the stress of losing something that from your
> perspective is working fine - especially if the successor isn't working
> as well for your use cases.

The problem is that the successor doesn't work at all. Even if KDE and
Plasma do their best to work around the inherent flaws of the Wayland
protocol, it's still "just KDE". It's not a generic solution. No user or
team will rely on KDE-specific workarounds, and no application will adopt
KDE-only APIs.

> So now is the time to articulate what's still painful about Wayland for
> you *that's not already mentioned on
> https://community.kde.org/Plasma/Wayland_Known_Significant_Issues *,
> because we have a year or so to address those concerns. And addressing
> as many of them as possible is the goal!

I think you should include some real-life use cases here.
Honest warning: if you ask enough people, that list will likely grow so
large it may take years to address everything.

Some real-life use cases crucial for us:

We currently do automation with `xdotool` - fast, trivial, and reliable.
How can we send synthetic events (keyboard, mouse) to applications in KDE
under Wayland?
I need a server/compositor-based solution, not an uinput hack. Or can
uinput device be attached only to a specific Wayland session (without
hacking the entire OS)?
It must work from a Bash script - no user interaction, no "allow virtual
keyboard" dialogs.

How do we run a headless Wayland session?
I need a minimalistic session with all the automation features available
but without the full Plasma desktop. Something like an empty Xorg session:
no DE, no WM, or a minimalistic WM like IceWM. Maybe just KWin?
I also need to run multiple (up to 10) of these sessions for a single user
on a single machine.

I need to attach VNCs to that headless Wayland sessions (or start them as
VNC sessions).

I need to take screenshots from a Bash script for each
WAYLAND_DISPLAY=wayland-N.
In fact everything should work from a Bash script, because that's quick and
easy to prototype.

All of this is TRIVIAL UNDER X11. Under Wayland it seems impossible, and
even things that are supported are fragile and fragmented.

And now the tricky question: can we make all of this DE-agnostic?
Because:

On Sun, 18 Jan 2026, 22:07 David Edmundson <david at davidedmundson.co.uk>
wrote:

> In addition, the Reddit thread linked has a lot of things that are simply
> not true in Plasma. We have gone to a lot of efforts for compatibility in
> ways that's aren't true for all desktops.

That is the core issue: incompatibility and fragmentation.
If I develop an automation workflow that somehow actually works under KDE,
do I then have to tell my team to switch to KDE, or else it won't work for
them?
Should an entire team be forced to switch DEs because Wayland is (not) so
great?

> I don't want generic Wayland rants, but things that are actually vetted
> and relevant to us

Dropping X11 support is EXTREMELY RELEVANT to most KDE users, including the
silent majority - the ones you'll only hear from AFTER their stable
environments are broken.

Best Regards,
Piotr H. Dabrowski
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20260202/f31b9f8c/attachment.htm>