Doesn't this cause lockups with the logout dialog if g-p-m is missing?
I've just spent some time chasing a bug (which I'm about to file) where my logout dialog hangs X11 (with a screen-shot its about to fade) for about 50 seconds whilst it tries to find g-p-m on its old DBUS name. (Gutsy). The new g-p-m has a different DBUS-path, so that needs changing.
I'm going to file a bug against gnome-session regarding the change of path in g-p-m, and hopefully just add my other "bug" as a comment here:
gnome-session should not hang trying to communicate with g-p-m !!
PS - from the app (gnome-session-save)'s point of view, the hangup is seen whilst accessing /tmp/.ICE-unix/{session_manager_pid}. A strace of gnome-session identifies the real hangup.
Doesn't this cause lockups with the logout dialog if g-p-m is missing?
I've just spent some time chasing a bug (which I'm about to file) where my logout dialog hangs X11 (with a screen-shot its about to fade) for about 50 seconds whilst it tries to find g-p-m on its old DBUS name. (Gutsy). The new g-p-m has a different DBUS-path, so that needs changing.
I'm going to file a bug against gnome-session regarding the change of path in g-p-m, and hopefully just add my other "bug" as a comment here:
gnome-session should not hang trying to communicate with g-p-m !!
PS - from the app (gnome- session- save)'s point of view, the hangup is seen whilst accessing /tmp/.ICE- unix/{session_ manager_ pid}. A strace of gnome-session identifies the real hangup.