[Techtalk] New OpenOffice bails out at 'Welcome to OpenOffice.org" screen

[Akkana Peck]

Conor Daly writes:
> On Tue, Jul 01, 2008 at 02:57:24PM +0200 or thereabouts, Wim De Smet wrote:
> > On Mon, Jun 30, 2008 at 10:50 PM, Anne Wainwright
> > <anotheranne at fables.co.za> wrote:
> > >
> > > Subsequently no sign of the blue screen or anything else, strace is
> > > showing the program is in a loop, doing what I have no idea, because the
> > > same lines come up again and again.
> > 
> > It doesn't tell me much either. It's hard to make out much from system
> > calls unless you see it continuously failing to open some specific
> > file or something. Guess we're not so lucky. :-)
> 
> Typically a program such as OOo will run an idle-loop while awaiting input
> from the user.  If OOo thinks it has a window up but you're not seeing it,
> it will sit in the idle loop forever waiting for an input event.  Not saying
> this is the idle-loop though...

Probably not the same as what you're seeing, but ...
I've seen an older version of OOo (2.2.x on Ubuntu feisty)
get into a loop trying to do X operations. It's probably not the
exact same problem -- in my case sometimes I'd get a window, but it
was super-slow and OOo was looping and repeatedly printing errors like

(process:7600): GLib-GObject-CRITICAL **: gtype.c:2240: initialization assertion failed, use IA__g_type_init() prior to this function

(process:7600): Gdk-CRITICAL **: gdk_screen_get_font_options: assertion `GDK_IS_SCREEN (screen)' failed

There are several Ubuntu bugs reported on that old issue, with
fixes that solved the problem for some people (it had something to
do with interaction among several desktop packages like libgtk+,
openoffice.org-gnome, and openoffice.org-kde). Some of them:
https://bugs.launchpad.net/ubuntu/+source/openoffice.org/+bug/127944
https://bugs.launchpad.net/ubuntu/+source/openoffice.org/+bug/128211
https://bugs.launchpad.net/ubuntu/+source/openoffice.org/+bug/73444

There might be something in those bugs that gives you a clue or
a debugging idea ... unfortunately it doesn't look like anybody
ever figured out what the root problem was, they just found
combinations of packages that made the problem stop.
Still, maybe there's something in there that might give you
an idea for how to debug the problem you're seeing.

	...Akkana