Red Hat Bugzilla – Bug 77768
Latest rawhide (9 november) kills gnome session
Last modified: 2008-05-01 11:38:04 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr-FR; rv:1.0.1) Gecko/20021003
Description of problem:
Upgrading to 9 nov. rawhide caused gnome-session to hang on startup on the two
boxes I've tested (one PIII the other athlon). When using a splash screen it
hangs before any icon is drawn.
Reverting to RH 8.0 Orbit* (+ gconf, bonobo, oaf...) sort of fixes it, gnome
session starts, but lot of gnome parts crash afterwards with
+ cannot create instance of abstract (non-instantiatable) type `GtkWidget' ;
Version-Release number of selected component (if applicable):
Steps to Reproduce:
Upgrade to 9 nov rawhide ORBit*/gconf*/oaf*/bonobo* rpms.
Actual Results: Useless login (empty blue screen sicne I had disabled splash
Expected Results: Normal login
This bug has rendered two stations about useless
(Ok, I know that using rawhide can cut, this is one of the most severe breakages
I've seen so far)
Still here after all this time, dropped all RH 8.0 throwbacks and went full Raw
One of the test boxes has an nfs /home, and login in from a RH 8.0 station works
so it's not a user config problem.
Right now I'm reduced to login in failsafe mode, and launching metacity &
gnome-panel by hand in the xterm like in the good old gnome 0.6 days (plus
usually any control-center applet to load user config)
I am the only one with this ? Since I can reproduce it on two independently
installed stations, this would really surprise me.
I can confirm this bug on two machines here.
I have since gone back to stock RH 8.0 bonobo.
Downgrade to bonobo-activation 1.0 (the newer bonobo-activation was backed out,
maybe it's still stuck in rawhide, I don't know).
It may also work to upgrade to a newer control-center.
The crash about 'GtkWidget' is fixed by upgrading libgnomeui, fwiw.
Anyhow, this is working in our current tree, not sure about the ftp site sync.
It works, thanks.
Unfortunately none of my rawhide sync scripts understands reverting to an older
version, so I was stuck with the old new buggy one.
Maybe upping the epoch in this case would have helped ?