Otto Wyss wrote:
> Michel D�nzer wrote:
> > 
> > Pretty much everybody seems to agree on this goal. The opinions differ
> > on how to achieve it though, see the flamewars^Wdiscussions about this
> > on the DRI, fbdev, kernel, ... lists.
> > 
> Anyway this issue has to be solved and I do as much as I can.
> 
> IMO DirectFB has a strong position since it's already working but that's
> not enough. It's time to join efforts so everybody is working for a
> common goal. And this goal has now to be discussed and written down so
> anybody interested doing some work knows what should be done.

BTW, there are some papers by James Gettys and Keith Packard
from OLS which I think are worth reading in this context:

"The (Re)Architecture of the X Window System"
(briefly mentions DirectFB ;-)
http://www.finux.org/Reprints/Reprint-Gettys-OLS2004.pdf

"Getting X off the hardware"
http://www.finux.org/Reprints/Reprint-Packard-OLS2004.pdf

(Abstracts:
http://lwn.net/Articles/94202/
http://www.linuxsymposium.org/2004/view_abstract.php?content_key=44
http://www.linuxsymposium.org/2004/view_abstract.php?content_key=47
)

The way I understand it, their plan is to pull lowlevel hardware
drivers out of X and either base X on exisiting OpenGL/Mesa/DRI
drivers (2D graphics via 3D API because it is faster), or create
a external 2D driver library, and only put some minimal
infrastructure into the kernel, but keep acceleration in
userspace.


Johannes


Reply via email to