Xgl and Xinerama
alexdeucher at gmail.com
Thu Mar 9 13:14:09 PST 2006
On 3/9/06, Matthias Clasen <mclasen at redhat.com> wrote:
> On Thu, 2006-03-09 at 14:32 +0100, Michel Dänzer wrote:
> > On Thu, 2006-03-09 at 07:54 -0500, Matthias Clasen wrote:
> > > On Thu, 2006-03-09 at 12:19 +0100, Xavier Bestel wrote:
> > > > On Thu, 2006-03-09 at 11:44, David Reveman wrote:
> > > > > I don't think there's that many apps that are xinerama aware. Except for
> > > > > the window manager, most apps will work just fine without the xinerama
> > > > > info available. Some xinerama compatibility thing could be created if
> > > > > necessary.
> > > >
> > > > All GTK+ apps are Xinerama-aware at some level: menus and other pop-ups
> > > > won't cross a xinerama limit for example. Maybe Xgl/Compiz should take
> > > > advantage of that and declare its rotating cube as 6 xinerama screens,
> > > > to have windows crossing 2 faces behaving correctly ? (I don't know if
> > > > it's even possible).
> > > >
> > >
> > > The only information needed on the client side for this support is
> > > just the monitor geometries, ie a list of rectangles. It should not
> > > be hard to make that information available somewhere, e.g. a root window
> > > property.
> > I guess I'm not sure why you'd re-invent Xinerama in this way but not
> > use it in the first place.
> I'm not reinventing anything here, just pointing out that this is all
> thats really needed on the client-side. Why Xinerama was not done this
> way in the first place, you have to ask HP...
That's also my the various drivers have to re-implement xinerama to
support "xinerama" on mergedfb/twinview setups.
More information about the xorg