Linked by Adam S on Sat 25th Jun 2005 19:56 UTC, submitted by Diego Calleja
X11, Window Managers A new acceleration architecture is being ported from kdrive by some QT developers, which will make composition managers like xcompmgr really fast and able to do some of the "display tricks" MacOS X has been doing for awhile.
Permalink for comment
To read all comments associated with this story, please click here.
What a mess
by Howie S on Mon 27th Jun 2005 18:34 UTC

Ok, so I'm somewhat confused. I've been following the osnews posts on xgl and mesa-solo, and have even emailed with one of their developers. Here's what I do understand:

* OpenGL is an open source library for rendering graphics (2D and 3D)

* XGL is an X-server which uses OpenGL for rendering, as opposed to the currently used system which is called XAA.

Here's what I'm a bit shaky about:

* In the end, you'll use either KAA or xgl, but not both, since they are in a sense two projects attempting to improve upon the same rendering capabilities of the X server.

* KAA's main "advantage", currently, is it's support of older (non-OpenGL) videocards, but it's debatable if driver support will actually be a hurdle for adopting xgl.

I basically turned my back on my linux partition back in 2000 primarily because of it's poor, clunky GUI experience, which only seemed to improve when I switched to WindowMaker or XFCE as my window manager. In the meantime, there are several efforts to improve upon X, the bane of many linux users's existances:

* there's the creation of, which increased my hope that all the feet-dragging that went on with Xfree86 would come to an end, and things would fiiiiiinaly move forward

* there's Rasterman's work within his Enlightenment system

* there's xgl (which reportedly will appear in NDL10)and mesa-solo (which, as I understand, would actually REPLACE the x-server entirely --- about ___'_ time)

* now there's KAA

* oh, there's also DirectFB

Here's an idea: Why don't all the interested parties work together towards some kind of agreement as to how the graphical subsystems will progress, and then create a modular architecture to allow for these and other forms of experimentation to take place? I thought that's where comes in?

Just an idea.