Re: Gnome-core

[Raphael Hertzog <rhertzog hrnet fr>]

Le Tue, Oct 12, 1999 at 08:57:09AM -0700, George écrivait:
> That's because we can't find it.  If you look at the code, it doesn't make
> any sense that it would do that.  Also I can't reproduce it here on my alpha
> so I assume you have to be quicker then the computer somewhere.

I can assure you that the bug does exist. I can reproduce it with 1.0.53.
Juste move your mice rapidly over the panel while reducing. Move it again
and again to try to catch the panel before it's hidden. :-)

You can try to change basep_widget_do_hiding to slow the hiding and better
reproduce the problem. :-)

The funny thing is that I can reproduce it only for a panel on the
bottom ...

I looked at the code (but I'm not at all a GTK hacker) and in
corner_widget_pop_up you can see this :

        if(GTK_WIDGET_REALIZED(corner)) {
                PanelWidget *panel = PANEL_WIDGET(BASEP_WIDGET(corner)->panel);
                corner->state = CORNER_MOVING;
		....
	}
	corner->state = CORNER_SHOWN;

I don't know what a "realized widget" is but if the condition is false then
corner_state may well be set to CORNER_SHOWN while the panel is in fact
moving (corner_widget_pop_up may be called more than one time, no ?) ...

I guess that something like that may happen with _pop_down and so on.
But that's only supposition... the problem is probably totally different.
:-)

Cheers,
-- 
Raphaël Hertzog >> 0C4CABF1 >> http://tux.u-strasbg.fr/~raphael/
<pub> CD Debian : http://tux.u-strasbg.fr/~raphael/debian/#cd </pub>