Event redirection adventures

[Matthias Hopf]

On Jan 12, 06 17:45:55 -0800, Keith Packard wrote:
> On Wed, 2006-01-11 at 18:37 +0100, Matthias Hopf wrote:
> > Hm. As far as I understand the issue, these two problems basically fall
> > into one (coord trafo) - for hit detection one could just check the
> > normalized transformed coordinates for 0<x,y<1 - and of course check for
> > shaped windows, etc.
> 
> It's not quite that easy; you can easily imagine translucent windows
> wanting to use alpha values for hit-detection, and pure 3D applications
> wanting to use standard 3D pick for hit detection. Assuming simple 2½-D

I see. It's the "etc." that's hitting me here :)

> compositing environments is rather limiting. So, hit detection cannot be
> done within the X server in general. Also, hit detection is done a lot
> more than transformation, so if there is a cheaper data structure which
> is specialized for that job, that's a good thing to consider.

ACK.

> Yeah, Deron has convinced me that attempting to dump the entire geometry
> of 3D objects into the X server is just a bad idea. Pick is a whole lot
> faster when accelerated by the 3D hardware anyway.

In that case we would be again at scene graphs. Maybe something to keep
in mind when we meet at Xorg DevConf in February. I've thought a lot
about this, and right now I'm pretty sure as well that it is a bad idea
to have the Xserver to have knowledge about the whole scene.

> Given that we have to provide a client-side mechanism for this (for the
> 3D environment case), I suggest that it's better to focus on that
> implementation for now and look into the performance benefits of adding
> some server-side computation if it looks like that will be widely
> useful.

My words :)

Matthias

-- 
Matthias Hopf <mhopf at suse.de>       __        __   __
Maxfeldstr. 5 / 90409 Nuernberg    (_   | |  (_   |__         mat at mshopf.de
Phone +49-911-74053-715            __)  |_|  __)  |__  labs   www.mshopf.de