Re: NSApplication bug

[Ludovic Marcotte]

> if we don't get any event - then nothing it's happening.
> if nothing is happening, there is no reason to update the windows :-)

The fact is .. they won't be updated if they are not marked to be updated! 
But, if they *need* to be updated (see my Blink application), they won't 
be updated until you generate an event.

For example, if you've a thread running that marks a NSButton as 
'disabled' when executing a critical section, the button won't be 
'disabled' until you move your mouse in the window containing this 
NSButton!

> and the less we update the windows, the better and faster.

I've tried with 0.25s.. on my poor celeron and I've seen no speed 
decrease since we don't update windows if they don't need to.

> the blinking cursor is setting up its own timer to call its own selector
> outside the event framework.  I think it makes sense to assume that if you
> fire timers to call you outside the event loop, you call [NSApp
> updateWindows]; directly after marking anything as needing a redraw.
> 
> This would be a very efficient way of implementing it.

It's up to you :-)

I think we should verify how Blink works under MacOS-X and OPENSTEP to see 
if we need to generate an event to get the window updated.

Anyone could make this simple test?

Ludo

-- 
Live as if you were to die tomorrow.
Learn as if you were to live forever.
 - Gandhi