Question about the future of Xorg

[Vladimir Dergachev]

On Sun, 15 Jun 2025, Alan Coopersmith wrote:

> On 6/14/25 12:45, Vladimir Dergachev wrote:
>> Imagine consumers using organic led wallpaper to cover walls and expecting 
>> to put windows with pictures, videos and homework there. Suppose you have 
>> low- resolution cheap OLED wallpaper, 96dpi. 5 meters (16ft) will translate 
>> to more than 18000 pixels!
>
> The X11 coordinate space tops out at 32767, so it's not going to scale to 
> wall
> size.
>
>>> well i couldn't in my std way - i'd run out of x client fd's - the xserver
>>> limits x client count to 128... :) i'd have to enable single process mode 
>>> in my
>>> terminal to keep it a single client.
>> 
>> Hmmm - scalability issue. We need to fix that in X :)
>
> If you look in the Xserver(1) man page there's an option already:
>
>       -maxclients
>               64|128|256|512 Set the maximum number  of  clients  allowed 
> to
>               connect to the X server.  Acceptable values are 64, 128, 256 
> or
>               512.
>
> What it doesn't tell you is that there's a tradeoff involved - the X11 
> resource
> id space is divided up into evenly sized chunks per client, so if you go from
> 128 clients to 256 clients, each of those clients gets half the available 
> number
> of resource ids to use.

Interesting !

I think these limitations in X11 are understable given it dates back to 
1980s. Back then 32-bit machines were big.

Nowadays, we can easily solve such limitations with 64-bit fields. Double 
the size, but our computers are orders of magnitude faster than in 1980s.

Looks like Wayland has 32-bit surface dimensions - much better.

Since I am the one arguing for scalability, I think I should setup a test 
system and check how modern Wayland compositors behave under load. We'll 
see how to go forward from there.

Do you know if anyone did equivalent of x11perf test with Wayland?


>>> it once. all those people windows is alienating and try linux certainly 
>>> won't
>>> be doing it either.
>> 
>> No need for everyone to do checks.
>> 
>> This is a perfect business case for a distribution like Debian or Ubuntu or 
>> Redhat, which offers pre-checked packages, secured with hash sums.
>
> I doubt any distro checks the source code of all the packages they include 
> that
> closely.  I wouldn't be surprised if some of the commits I've pushed to X11
> git repos have never been looked at by another human being.

Another example is the SSH issue Debian had a while back. But these kinds 
of problems cannot be solved with a popup asking user "Trust SSH or not?"
And the past history shows these issues are discovered relatively quickly.

For an app like xmag, all one needs to do is check that there is no code 
trying to send data outside. If there are bugs then the worst thing that 
happens is that it crashes.

If people care about security (like I do), they can go a take a look 
themselves. One can also have a crowdsourcing like project where people 
look at some Debian packages and then we know that specific versions are 
checked.

best

Vladimir Dergachev


>
>        -Alan Coopersmith-                 alan.coopersmith at oracle.com
>         Oracle Solaris Engineering - https://blogs.oracle.com/solaris
>