Re: bonoboui changes break nautilus

[murrayc t-online de (Murray Cumming)]

Could somebody please go ahead with this revert? I'm not enough of a cvs
expert and I don't want to compound this by getting that wrong too.

I'll investigate what Nautilus (and apparently libgnomeprintui) are
doing and provide patches for them.

Sorry, nobody knew that this would happen, and we can't rebuild all of
GNOME for every little patch.

On Mon, 2002-04-22 at 21:52, Owen Taylor wrote:
> 
> With a build of current CVS, nautilus doesn't come *remotely* close to
> working (all menus and toolbars missing), with messages "Trying to
> invoke CORBA method on unbound UIO Container".  Looks like the
> _construct() functions changes are causing these changes.
> 
> Reverting all changes since Friday, namely:
> 
> ===
> 2002-04-22  Murray Cumming  <murrayc usa net>
> 
> 	* bonobo/bonobo-window.c: (bonbo_window_construct) Recommitted,
> 	with bonobo_window_construct() implementation which was ommitted
> 	somehow before.
> 
> 2002-04-22  jacob berkman  <jacob ximian com>
> 
> 	* bonobo/bonobo-window.c (bonobo_window_construct): revert
> 	previous commit as it broke the build and wasn't approved by the
> 	release team
> 
> 2002-04-22  Murray Cumming  <murrayc usa net>
> 
> 	* glade/glade-bonobo.h: reverted part of previous commit - I
> 	had committed part of another pending patch that depends on a
> 	not-yet-applied libglade patch.
> 
> 2002-04-22  Murray Cumming  <murrayc usa net>
> 
> 	* bonobo/bonobo-selector.[h|c] BonoboSelector now has an
> 	"interfaces_required" property, and the child BonoboSelectorWidget
> 	is instantiated in the *_instance_init() function, so that
> 	bonobo_selector_new() can now be just a convenience wrapper for
> 	g_object_new().
> 	
> 2002-04-22  Murray Cumming  <murrayc usa net>
> 
> 	* bonobo/bonobo-window.[h|c]: bonobo_window_new() is now just
> 	a convenience wrapper for g_object_new(). This is possible because
> 	BonoboWindow now has a "win_name" property and the UIContainer is
> 	created in the *_instance_init function.
> ===
> 
> Fixes the problem. I would advise making this revert in CVS and then
> making sure that we have consensus and an approved patch before committing
> anything again.
> 
> Regards,
>                                         Owen
-- 
Murray Cumming
murrayc usa net
www.murrayc.com