Question about the future of Xorg

[Lyude Paul]

Most of the people working on Wayland also worked on X11, a lot of lessons
from X11's development specifically influenced Wayland's design. I think
it's very much worth noting this because people talk about X vs. Wayland as
if a separate project popped up next to X11 out of nowhere, but the reality
is a lot of us see Wayland as the natural progression to X11 because it
allowed us to be free of some of the constraints that come from how the X
server is designed, like heavy round trips in its protocol, the fact that
most modern compositing on X is just extension on extension on top of the
actual X11 server which no longer does half of the things it's supposed to
handle.
You know X server used to draw its own widgets? It also has font rendering,
a HAL for video drivers and input drivers, and more. And almost all of this
stuff was out of use well before Wayland came onto the scene. Client
toolkits do their own rendering (easier, faster, much more flexible), and
font rendering. Save for nvidia's video drivers, we had pretty much stopped
using the HAL for new input and video drivers as well
(xf86-video-modesetting and xf86-input-libinput work for p much all
hardware. There's a huge amount of stuff like this baked into the server,
but very little of it makes sense on modern operating systems and much of
it is really constraining.
X wasn't "badly designed" per-say, for what it was I'll absolutely say it
was a wonderful piece of software and it did its job well. But it's also
designed for an era of computing that is much different than how most
modern desktops work, so for a lot of the functionality we wanted to see in
Wayland the only way to have implemented it in X would have required
breaking people's setups. So, technically speaking splitting the
development off was kind of a given in some sense anyway. Even if we didn't
move work to Wayland it's more likely work would have been on an X server
that didn't really resemble X11 and wasn't 1:1 compatible.

On Mon, Jun 9, 2025, 09:51 Vasily <vasil_tik at yahoo.com> wrote:

>
> >
> > that's not a wayland thing. wayland is just a protocol plus associated
> > tooling and libraries to speak that protocol. panning is a compositor
> thing.
> >
> > if you wanted a compositor to support this then add it to that
> compositor. as
> > there is no single compositor in the wayland universe, you'd have to
> make a
> > choice. you could also build your own compositor.
> >
>
> And that is the main issue with wayland.
> "Very intelligent" people created a protocol plus associated tooling and
> libraries and expected that community will build a new great compositor.
> But if X11 was developed for ages and polished for decades by people who
> loved it, no one want to spend his time and design a new compositor for
> every
> existing GPU
> Because in this case it will be just another X11.
>
> If you take a look on Linux there are few servers in it already - like
> systemd or dbus, pulse/pipewire, named ...
> For my understanding there is nothing wrong with a server/client approach
> and right now there is not any evidence that there is a wayland compositor
> that works like X11
> Second problem that desktop must adopt such compositor.
>
> My personal opinion is that there are some issues need to be solved before
> wayland compositor can substitute X11.
> So I think X11 will work for us another 5 -10 years.
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.x.org/archives/xorg/attachments/20250609/05d89835/attachment.htm>