Re: memory efficient compisiting

On Wed, 2004-03-03 at 03:47, Jon Smirl wrote:
> --- Michel Dnzer <[email protected]> wrote:
> > On Tue, 2004-03-02 at 19:01, Jon Smirl wrote: 
> > > 
> > > Refetching a window's contents over a network will also take a lot of 
> > > time. 
> > 
> > What would have to fetch them over the network?
> 
> If we've thrown them out of the cache and they are from a remote user. Just a
> standard fullscreen repaint command for a remote window. Nothing new in X.

Surely the X server will keep a local copy?


> For example how long does it takes to blend ten 5MB translucent textures [...]

Ten screen-sized, translucent windows on top of each other is asking for
trouble...

> Continuous composition does look doable with a normal desktop mix where 
> all of the windows aren't translucent. 

Indeed, under 'normal' circumstances I'd expect the amount of data
required for the compositing to be of the same order of magnitude as the
visible screen data.

> Also, nothing bad happens if it takes two retraces to do a complex 
> composition pass, I doubt if you'll even notice.

Actually, I think you will notice if it's usually done every retrace but
not every now and then (I think this is called judder?). Even only doing
it every second retrace might be better than that.

Anyway, let's see how it's working now, we can worry about clever
optimizations when we know it's not good enough.


> > Also, IIRC I've seen you argue before that cards with GBs of video 
> > RAM are imminent. :)
> 
> They are, but not everyone is going to have one. If every card had 1GB RAM we
> wouldn't have a problem. The argument about 1GB cards was for a different reason
> (where are we going to get 1GB of kernel address space to map them into?).

Sure, kinda my point.


-- 
Earthling Michel Dänzer      |     Debian (powerpc), X and DRI developer
Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer