Re: NSWindow question

[Andrew Ruder]

On Mon, Jul 21, 2003 at 12:30:11PM +0100, Richard Frith-Macdonald wrote:
> 
> I think it's theoretically possible (the window is not specially 
> retained during event processing to prevent it) since, for instance, a 
> keyboard event to close the window could arrive, and that could cause 
> the window to be destroyed.
> 
> Perhaps this patch is good anyway, and what's needed is for the event 
> system to retain a window while it is
> processing an event.  I'm not happy to apply the patch straight away 
> though ... perhaps someone else has some thoughts?

afaict, both of the variables in question are only destroyed, never
just released.  In this case, it would be safe to destroy them
in the dealloc, regardless if its needed or not.
-- 
Andrew Ruder