Re: [GtkGLExt] GtkGLExt 1.2.0 is out ...



On Sun, 2006-02-05 at 08:35 -0700, Timothy M. Shead wrote:
> Kouhei Sutou wrote:
> 
> >>I have an application using GtkGLExt where the GdkGLContext created in
> >>the call to gtk_widget_get_gl_context doesn't get destroyed. Unsure of
> >>the proper usage pattern, I went back to the GtkGLExt examples to see
> >>how they do it. I think they might have the same problem.
> >>
> >>Looking at the code in gtkglwidget.c, it looks like this GdkGLContext
> >>should get destroyed in response to a call to gtk_gl_widget_unrealize.
> >>However, when I run the "button" example, that function never seems to
> >>get called. valgrind notes the GdkGLContext as "possibly lost".
> >>
> >>I'm observing this in 1.0.6. (I would have checked CVS before sending
> >>this, but SourceForge anonymous CVS appears to be flaking out again at
> >>the moment.)
> 
>  > It's a good news, but I have a question: Is the problem
>  > reported by Braden fixed?
> 
> For reasons of risk mitigation, I did not attempt to address this issue 
> in 1.2.0.  My rationale was that if this is in fact a bug, it must be a 
> pretty minor one to have escaped attention for so long.  Now that there 
> is a stable release that addresses immediate concerns (GTK+ >= 2.8), 
> I'll take a look at this.

gtk_gl_widget_unrealize seems never to get called; and I don't have a
sufficiently thorough understanding of Gtk to know why. It's not hard to
observe, though--just put a print statement in it, run button, and close
the app. (Surely it *should* be called for that example... ?)

-- 
Braden McDaniel                           e-mail: <braden endoframe com>
<http://endoframe.com>                    Jabber: <braden jabber org>




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]