not really a problem, just a question.

[nighty]

they do say that questions make the room for problems though

well its a general question about the future of windowing enviroments in linux
why isnt there an opengl based windowing enviroment on linux?
on what level do you need to start building it?
do i just need to write yet another X window manager using a library (glut for example) or do i need to rewrite X as a whole?
i just dont understand why such projects dont exist because of the tremendeous hidden potential that lies in gpu equiped computers
not mentioning the possibilities of gui innovation.
if there are some cons to such enviroment i dont know about ill be more than happy to know about them.

i ask about this subject because from what i see both microsoft and apple are moving toward a semi 3d enhanced enviroments and if linux desktop enviroments and window systems in general will remain the same until 2006 a gap will form.
most home users wouldnt want to use 'normal' desktop enviroments if theyll have a cpu above 5ghz and more than 1 gig of ram for a personal computer.


i heard about the 3dwm but its development barely showes any signs of life. and its purpose isnt to create a desktop enviroment for day to day use but to lay a foundation for future development.


just interested in what you have to say

cheers

[ManicMailman]

define "such projects don't exist"

because it seems to me....
http://www.fresco.org
http://picogui.org/

[pjp]

Moved from Desktop Environments.
_________________
Quis separabit? Quo animo?

[Verteron]

I'm increasingly uncertain about the future of XFree and X in general, and what should be done in the long term to improve it...

On the one hand I see a network enabled protocol that I use every week to access my box from various interesting places. I see GTK2 and QT3 with beautiful antialiased fonts, SVG vector icons and increasingly decent skins. It's almost enough to make you think "there's nothing wrong". We lack alpha transluent windows, but this is coming in XFree5. I look in to the future and see things on the horizon like E17 and cairographics which look pretty damned cool.

Then again, I look at Apple's Quartz Extreme... and drool. Sure, they have control over their platform that OSS developers can never have, and as such it's easy for them to make the changes necesary to implement these nice new features like Expose and Alpha Blending.

XFree's development cycle is so much slower, and of course it can't be guarenteed to even run on a system with an OpenGL card, so an OpenGL compositing is a non-starter already. But considering that Longhorn will probably be out before the "XFree problem" is solved, one wonders how the Linux Desktop intends to keep up in terms of eye-candy and things like Expose which are genuinely useful.

There's no doubt whatsoever that X's protoco could do with some work. But of course it can only be changed by X.org who are even more conservative than the XFree developers.

What solutions are there? Wait for XFree or Xouvert to sort themselves out and find solutions? I just don't know, reading their mailing lists sometimes makes me wonder whether they appreciate what a hugely important role they have in the whole OSS movement... Keith Packard understood some of these issues, but he, of course, has now gone.

Fresco has been under development for many, many years and they still only have a barely functional code base to show for it. Other projects just don't seem to gain enough momentum because as far as end users are concerned, if it doesn't yet have accelerated 2d/3d/XUV it's not worth trying. Chicken and egg, since for graphics cards manufacturers to produce drivers for a new system, you need momentum, but you can't get momentum until new graphics card drivers are available. Getting the GPU manufacturers to make drivers for XFree was an uphill battle, goodness knows how hard it will be to get them to support another alternative system.

Call me pessimistic, but something needs to be done. I guess we'll just have to wait and see.

[nighty]

in answer to ManicMailman

you are right ofcourse fresco does use opengl, but its a relatively small project that doest have alot of buzz going for it.
picogui is a whole different story, i didnt really see any mention to 3d hardware in its agenda, and since picogui currently targets itself as a gui for embedded systems some time will pass until you will actually see it turn in that direction.

ofcourse theres also E17 which has buzz but so little people develop it its scary and since so little people develop it, the final product might not please everyone.
since E17 currently looks alot like E 16.5 in terms of usability,
and most users that left E for other wms didnt do so not because of graphics or speed but its feel or intuitiveness. so i tend to doubt that its going to be the next mainstay of linux DEs.

so such projects exist, yes, i made a mistake, but it doesnt really answer my question to its full extent