>
> > > My question is....when X Windows starts doing this, how the hell do we
> > > debug it?
> >
> > look at the :0.log file in /var/gdm
>
> THANK YOU! It was complaining about not being able to connect to some
> socket. I found the socket in /tmp; it looked like an old copy. I
> removed all the old X-related stuff from /tmp (e.g. .X11-Unix,
> .ICE-Unix, .font-unix, and a few others), and now X Windows will start
> correctly.
>
> One note I wanted to add, though....X Windows was dying/starting in a
> loop even though I've removed it from /etc/inittab -- just typing "gdm"
> all by itself on a root console was enough to start the loop.
>
Three problems about that:
-Frail font solution: kill the font server in run level 5 and X enters
in a loop of death. It would have been far more sensible to put a
normal (file based) font path in XF86Config and use xset during
initilaisation for replacing this font path with one pointing to the
font server. However if font server is not running then X does not
change the font path and doesn't die.
-Frail [GKX]DM solution. Normally GDM is preferred over KDM and this
one over XDM and RedHat 6.2 is smart enough to go to the "lower" DM
if the higher one is absent. But there is variable who allows the
user to override the selection of DM. However if that preferred DM
is missing the script does not try to loook for another DM. It is
not a good idea to assume the sysadmin is infallible.
-No detection of X loops of death. There is another well known
distribution who detects them and automatically goes to runlevel 3 so
it would seem logical RedHat looking at how they did. Don't they
have someone for looking at what competition is doing?
--
Jean Francois Martinez
Project Independence: Linux for the Masses
http://www.independence.seul.org
--
To unsubscribe:
mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
