Re: multiXnest results

[Ander C. de O. <an...@c3...>]

Michael Pardee wrote:
> multiXnest looks great!  I always thought about creating that
> capability by hacking VNC, I never really knew about Xnest.   I
> haven't tried Aivils variant yet, that's next on my list.  Of course
> there is no 3D acceleration and no xv extension, so the old single
> xserver per card approach still has its place, but multiXnest looks
> very promising.  NOTE: the keyboard leds are all working fine for me.
Are you using a ruby or faketty together with multiXnest?
> 
> I set up a single nvidia dual-head card for two desktops, and am
> having one significant problem:  sometimes after the machine sits idle
> for a few hours (sometimes it goes for a long time OK) the gnome
> sessions mysteriously exit and it goes back to the gdm login screen.
> This same machine in a non-multiXnest configuration doesn't have that
> problem.
> 
> So, where is the log for multiXnest?  Not in /var/log/Xorg* because it
> is not an original X server, and ~/.xsession-errors only goes so far
> before filling up with some stupid gnome error so I can't see if there
> were gnome messages right before gnome dies.  Also, I didn't see a
> debug command line option, what do you recommend for debugging
> Xnest/multiXnest?
It actually doesn't have a log. All error messages are sent to stderr, 
but, if you're using gdm, there sould be a /var/log/gdm/:*.log.
> 
> On a much less important note, gnome is acting a little strange with
> Xnest, I think because the nested servers are indeed slower with
> higher cpu usage, and the focus stealing prevention is kicking in way
> too much.  Windows are sometimes getting launched behind others with a
> flashing taskbar entry.  I have to figure out how to turn that off in
> gnome.
> 
> Also, it probably doesn't matter, but I would feel better if there was
> a way to completely disable the keyboard(s) on the root hardware x
> server.  I found limited info, but couldn't start X with several
> approaches.  One would think some kiosks would be without a keyboard
> and there should be a way to use X like that.  We are using
> AllowMouseFailOpen to disable mice on the root hardware x server.
Xorg does not allow this. It won't start without a keyboard. But if your 
problem is ctrl-alt-backspace ande ctrl-alt-f[1-12], then you can use 
the options DontZap and DontVTSwitch.

Ander