[gtkmm] Re: Bug #134786

[Martin Schulze <martin-ml hippogriff de>]

Am 2004.02.19 22:41 schrieb(en) Bryan Forbes:
> Martin
>     I know I probably shouldn't be doing this, but I am because of  
> the
> severity of this
> bug.  I emailed the gtkmm and libsigc list two days ago about it and
> filed a bug
> yesterday.  If you could get back to me about this bug, it would be
> great.  This is
> definitely critical and has been confirmed on x86 and ppc.  Thanks in
> advance!

You are right, this is a serious bug.
Unfortunately, I have big problems to get gtk+-2.3.x installed on my  
notebook in parallel to gtk+-2.2 so I can't debug gtkmm-2.4. My  
troubles are increased by the fact that gtk+ and gtkmm take hours to  
compile on this slow machine.

Please be patient or help us tracking down the problem. You could e.g.
set a breakpoint at adaptor_trait.h:155 and find out how the endless
loop is triggered. Then we could write a simple, libsigc++2-only test
case.

Similarly I need help testing the most recent patch here:
http://bugzilla.gnome.org/show_bug.cgi?id=125061
This probably also affects bug #134786.

I'm cc-ing the mailing list and attaching this correspondance to the  
bug report.

> -Bryan
>
> P.S.  I've been browsing through gstmm and fixing stuff for 0.7.4 and
> I have some
> questions regarding Gst::Object for you.  Should I email the list or
> just correspond with you?

Please, always use the mailing list. You will have noticed that my  
activities on gtkmm were idle for a long time now. This is partly  
because I have less time these days but also because I found the  
gstreamer interface quite messy by the time I helped out starting the  
wrapper. I'm afraid it will still be difficult to write a clean gstmm  
API and a lot of things must be hand-coded rather than generated by  
gmmproc.

Regards,

 Martin