Re: real cross-platform bugs

[Havoc Pennington <hp redhat com>]

Hans Breuer <hans breuer org> writes:
> 
> All bugs described here are verified to exist not only with the 
> win32 port of gtk+, but also with the same cvs version running 
> under Linux (Suse 6.3), window manager Sawfish.
> 

Yes, "Move to current position" has issues with most window managers -
they are a bit buggy.

> Thinking a little more about it, I think I haven't understood or 
> think it doesn't make much sense at all to move a window to it's
> 'current position'. Isn't it always already there ?
> 
> But you are right that the gravity seems to influence the direction ...
> 

The purpose of the "move to current position" test is to see if window
managers properly handle gravity and that gtk_window_get_position()
works properly. If they handle it correctly, then moving a window to
the result of gtk_window_get_position() will not move the window.
So if things are working, the window will not move. If they're broken,
then it will.

On Windows, you want to emulate the correct window manager behavior. 
Read the freedesktop.org window manager spec "Implementation notes,"
and also the ICCCM, for details on this.

Havoc