[cairo] XDBE

[Jim Gettys]

Yeah, but we have to get vertical retrace synchronization all set;
sometimes hardware makes that easy, but often not.
			- Jim


On Tue, 2005-11-22 at 17:27 -0800, Keith Packard wrote:
> On Tue, 2005-11-22 at 13:03 -0800, Bill Spitzak wrote:
> 
> > Another possible better fix is that resizes should copy *into* the 
> > backing buffer the current contents of the screen. Then if it 
> > accidentally gets swapped back before the program finishes drawing it, 
> > the user will not see any change in the undrawn areas.
> 
> Composite does this.
> 
> > I agree. It would be nice if *all* backing copies went to the same area, 
> > no matter how they are created. Ignore any minor glitches (ie I think 
> > the composite extension could show partially-updated backbuffer, just 
> > try to avoid recompositing except when a program sends the 
> > back-buffer-swap command).
> 
> Most apps that 'double buffer' do it themselves with a separate pixmap;
> it's more portable, and gives reliable results. This makes solving the
> DOUBLE-BUFFER case a lot less interesting.
> 
> Of course, backing store should be implemented with Composite in any
> case; it's far more efficient and follows the protocol specification
> correctly (unlike the existing backing store code which saves far too
> much).  If we do that, we'll eliminate miles of X server code, and the
> separate buffering for backing store as well.
> 
> -keith
>