Re: Problem with reparent (Gtk+-2.12)

["Andrew E. Makeev" <andrew solvo ru>]

В Пнд, 04/02/2008 в 18:04 +0300, Andrew E. Makeev пишет:
> В Пнд, 04/02/2008 в 15:48 +0100, Murray Cumming пишет:
> > On Mon, 2008-02-04 at 17:33 +0300, Andrew E. Makeev wrote:
> > > Good day,
> > > 
> > > 
> > > When we are trying to reparent container with toolbar inside from one
> > > window to another, and then reparent back again, application throws GTK-
> > > Critical about some widget not ANCHORED or VISIBLE, and next we get
> > > "widget is not realized" inside map/unmap signal handler.
> > > 
> 
> > Can't this test case be made much smaller?
> > 
> > Can you try it in C too?
> > 
> 
> Yes, it is all possible. I just attached code to show the problem.
> First, I would to know if someone reported it already.
> 
> Btw, on Fedora Core 8 OS (Gtk+-2.12.5, Gtkmm-2.12.3) I had only
> accelerators warnings. So, the problem shows only on Win32 OS. Otherwise
> I could determine exact problem and suggest patch myself.
> 
> > > 
> > > Where in bugzilla I could track it and, probably, attach simple example
> > > to show the problem?
> > 
> > Under gtkmm or GTK+ in bugzilla.gnome.org, depending on which test case
> > shows it.
> > 
> 

There is simpliest example I could make in C.

Run application, create new window. Hit "dock/undock" button and see
following:

(reparent.exe:4924): Gtk-CRITICAL **: gtk_widget_realize: assertion 
`GTK_WIDGET_ANCHORED (widget) || GTK_IS_INVISIBLE (widget)' failed

Gtk+-2.12.6, on Win32, built with MinGW environment.

Regards,
-andrew

P.S.
Please, let me know if I should to submit a bug.