On Tue, 18 Mar 2025 23:51:21 +0300 Andrei Borzenkov <arvidjaar@gmail.com> wrote:
18.03.2025 19:35, Lew Wolfgang wrote:
Asked and answered.But here's a simple example. Consider a situation where you have terabytes of large raw data files on a remote server that doesn't have a console. Jane logs in with ssh -X and runs code to visualize some of the data and produces lots of jpg files. Thousands of them. She visualizes some of them with xv. She does this simultaneously with multiple servers with the displays appearing on multiple local virtual desktops.
I already answered this in another mail. X11 applications (or any application that can fall back to X11) work as before when you are running Wayland desktop. For pure Wayland on both sides you need to use some helper - exactly as you need for X11.
Your use case has nothing to do with alleged "X11 network transparency". Your programs talk to the *local* X11 server; they are not aware that this connection is forwarded over network to somewhere else.
There is no X server on the remote machine! Or at least there does not need to be. The remote machine just runs the X client, and it is connected to the X display server on the local machine. Wayland doesn't have that capability. Whether it matters or not I don't know.
So, in your case there is no practical difference between X11 and Wayland. It is possible to forward connection to the *local* X server over network using "ssh -X"; it is possible to forward connection to the *local* Wayland compositor over network using Waypipe on top of SSH.
If you want to compare, at least compare apples with apples.