Public bug report changed: https://launchpad.net/malone/bugs/28493
Comment: A little bit more information for you, in the hope it might be useful. I've tried to obtain a stack trace at the point where the trouble occurs (instead of after the damage has already been done). My main problem is the graphical window system completely locks up at the point of failure, so I can't access the gnome-terminal shell to be able to type in "thread apply all bt". I tried Ctrl-Alt-F1 and sending various signals to the gnome-panel and gdb processes, but no luck. But I did manage to get a couple of clues... [EMAIL PROTECTED]:~$ gdb gnome-panel ... [snipped boring stuff] (gdb) run Starting program: /usr/bin/gnome-panel [Thread debugging using libthread_db enabled] [New Thread 46912585372432 (LWP 4451)] (gnome-panel:4451): Gtk-CRITICAL **: gtk_icon_theme_lookup_icon: assertion `GTK_IS_ICON_THEME (icon_theme)' failed (gnome-panel:4451): Gtk-CRITICAL **: gtk_icon_theme_lookup_icon: assertion `GTK_IS_ICON_THEME (icon_theme)' failed [New Thread 1074006368 (LWP 4474)] (gnome-panel:4451): Gtk-CRITICAL **: gtk_drag_source_set_icon_name: assertion `icon_name != NULL' failed Program received signal SIGINT, Interrupt. [Switching to Thread 46912585372432 (LWP 4451)] 0x00002aaaaafa7cf9 in gmenu_tree_item_get_type () from /usr/lib/libgnome-menu.so.2 (gdb) Quit Note the 3 "critical" assertions. The first two happened immediately after running gnome-panel. I often get these types of messages when I run stuff from the command-line; is this normal? But the third assertion is much more interesting, it appeared only when I actually triggered the bug. It appears there's a problem with the parameters passed to gtk_drag_source_set_icon_name(). But there seem to be several places in the gnome-panel source where that could get called from. I'll keep playing with it, see if I can persuade my system to give me a better back-trace. Lambros -- desktop-bugs mailing list desktop-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/desktop-bugs
