Linked by Thom Holwerda on Mon 2nd Jan 2006 16:43 UTC, submitted by Anonymous
X11, Window Managers David Reveman has made the latest XGL source code available to download. This comes a few weeks after development of the project was criticised for being done 'behind closed doors'. There have been huge changes to XGL, the most significant being restructuring of the code, allowing XGL's GLX support to function on other drivers than the proprietary Nvidia one. Xcompmgr can currently be run under XGL with full acceleration provided that the proprietary ATI or Nvidia drivers are used. An OpenGL based compositing manager, 'Compiz' is currently in the works and a release is expected in February. David intends to get the code into freedesktop CVS as soon as possible, after which the code should eventually merge with Xorg.
Order by: Score:
Good News
by abraxas on Mon 2nd Jan 2006 17:01 UTC
abraxas
Member since:
2005-07-07

With the release of Xorg 7.0 and now this it is an exciting time for the Linux desktop. 2006 looks like it is going to be an interesting year for Linux.

Reply Score: 2

v RE: Good News
by Tom K on Mon 2nd Jan 2006 21:16 UTC in reply to "Good News"
RE[2]: Good News
by zerblat on Tue 3rd Jan 2006 00:25 UTC in reply to "RE: Good News"
zerblat Member since:
2005-07-06

I absolutely agree. It's easy to forget the amazing progress that has been made every year since Linux first was born.

Every year, LWN publishes a timeline of the most noteworthy developments during the past year. The one for 2005 is only avaliable for subscribers, but previous timelines are available here: http://lwn.net/Articles/Timeline2004/ It's pretty interesting to go back a couple of years and see the things that were big news back then.

Every year has been a great year for Linux and Free Software in general. I'm sure 2006 will be even better.

Reply Score: 2

RE[2]: Good News
by abraxas on Tue 3rd Jan 2006 01:36 UTC in reply to "RE: Good News"
abraxas Member since:
2005-07-07

Don't you have anything better to do than be negative all the time? Every year since I've been using Linux has seen tremendous changes on the desktop. With XGL and Xorg 7.0 we should see improvements on the desktop that rival what is planned for Vista. At least I have seen more concrete demonstrations with XGL/Xorg than with Vista betas.

Reply Score: 2

v RE[3]: Good News
by Tom K on Tue 3rd Jan 2006 04:29 UTC in reply to "RE[2]: Good News"
RE[4]: Good News
by somebody on Tue 3rd Jan 2006 04:45 UTC in reply to "RE[3]: Good News"
somebody Member since:
2005-07-07

The Linux cake is obviously loaded with hallucinogens, as I don't see how else you people could make the claims you do. ;)

Obviously, but still better than having need for regular lobotomy checkup as you obviously have. Comments you make about Linux could be a nice psychological example of mental anxiety disorders and pussy deprivations.

Living in this bitter world must be cruel for you, isn't it?

Reply Score: 1

v RE[5]: Good News
by Tom K on Tue 3rd Jan 2006 08:36 UTC in reply to "RE[4]: Good News"
RE[6]: Good News
by poofyhairguy on Tue 3rd Jan 2006 09:15 UTC in reply to "RE[5]: Good News"
poofyhairguy Member since:
2005-07-14


It really is simple: I think Linux is garbage, and I'm going to say so.


It does not bug us that you have an opinion. It bugs us that you have a roll to fill (aka the anti-Linux guy on the OSNews site) and you don't do it well.

Say "the Linux Desktop sucks." Or in this case something like "What news! David Reveman almost has done what OSX did years ago."

Not "Linux is garbage." Linux is used as the basis for the popular TiVo systems, many kinds of devices like cell phones and it is the operating system running 7 of the top 10 supercomputers in the world. How can it do such things if its all garbage? Obviously your claim is false.

We don't want to force you to like Linux, or stop saying bad things about it. We just want you to be a better troll than you are. There is a need for balance is every situation, but with your obviously erroneous statements you do not hold up your end of the bargain!

Reply Score: 3

v RE[7]: Good News
by Tom K on Wed 4th Jan 2006 08:48 UTC in reply to "RE[6]: Good News"
RE[6]: Good News
by superstoned on Tue 3rd Jan 2006 09:18 UTC in reply to "RE[5]: Good News"
superstoned Member since:
2005-07-07

well, it's easy - you'll be voted down until you start making sense, or stop trolling...

linux sure has some disadvantages, but hey, i have to work on an XP box right now - and there are several reasons i hate it more than my Linux box at home. so linux is at least doing better than XP, right?

but yes, that's nothing new... i guess you're a mac user, as Mac OS X might be the only OS that can beat linux+KDE...

Reply Score: 1

v RE[7]: Good News
by Tom K on Wed 4th Jan 2006 08:50 UTC in reply to "RE[6]: Good News"
RE[6]: Good News
by deepspace on Tue 3rd Jan 2006 09:54 UTC in reply to "RE[5]: Good News"
deepspace Member since:
2006-01-03

If Linux is garbage, why are you posting a website in your "user page" runnig linux and php?

Reply Score: 1

v RE[7]: Good News
by Tom K on Wed 4th Jan 2006 08:51 UTC in reply to "RE[6]: Good News"
RE[6]: Good News
by somebody on Tue 3rd Jan 2006 14:29 UTC in reply to "RE[5]: Good News"
somebody Member since:
2005-07-07

I like how you Linux fools pretend to be psychologists, and concern yourselves with the deep reasons for my trollage.

No need for PhD in your case, anyone can notice you're not really IQ-based.

It really is simple: I think Linux is garbage, and I'm going to say so.

Yes, do that. As long as you finally start posting comments. I've always enjoyed good critical view against anything, no matter how much I preffer that thing.

Actualy, the best ideas come from people bitchin'. Reffering to Chaos theory.

While I do that, the Linux apologists just give excuses for why things are how they are, all while proclaiming its superiority over anything else in existence.

We did this theme already, haven't we? Browse trough my comments and say I'm OS-agnostic with preference in Linux. It is true. But no way can you say I'm apologist. My comments are just as Linux critical as OSX or Windows. It is called realistic view.

Perhaps all of *you* should take a step back and perform some psychoanalysis?

Wake up Dorothy! You've never been to OZ.

Only some cases like you are moronic enough to deserve that. Some of us are just to ordinary to be fun to disected by psychoanalisis.

p.s. I don't mind if you troll. Just do it good and based on truth, instead of your fiction-based mental grief with Linux.

Reply Score: 1

v RE[7]: Good News
by Tom K on Wed 4th Jan 2006 08:52 UTC in reply to "RE[6]: Good News"
Linux Zealot, aren't ya?
by Snifflez on Tue 3rd Jan 2006 05:45 UTC in reply to "RE[3]: Good News"
Snifflez Member since:
2005-11-15

"The Linux cake is obviously loaded with hallucinogens, as I don't see how else you people could make the claims you do."

Ah, but you seem to be as obsessed with Linux as those you profess to despise so much, LIP. You really need to stop this whole "maggot calling pus white" thing.

Reply Score: 1

v RE: Linux Zealot, aren't ya?
by Tom K on Tue 3rd Jan 2006 08:55 UTC in reply to "Linux Zealot, aren't ya?"
RE[4]: Good News
by abraxas on Wed 4th Jan 2006 18:28 UTC in reply to "RE[3]: Good News"
abraxas Member since:
2005-07-07

It's funny how you never make counter claims. You just poo poo anything Linux. How approriate your username is.

Reply Score: 1

Your Linux obsession is not healthy...
by Snifflez on Tue 3rd Jan 2006 06:03 UTC in reply to "RE: Good News"
Snifflez Member since:
2005-11-15

You know, you remind me of a little kid who just got installed Fedora Core for the first time, he thinks he's a real Linux guru now, so in a voilent fit of pants-soiling joy he prances around various online forums screaming "linux rox! linux rox!!!11!1", liquid feces streaming down his legs, his face a mask of self-righteous glee and disgusting self-importance. Well, you're just like that kid, only you opt to inform that "linux sux". Other than the sign in front of your attitude (+/-), there is very little difference between you and a Fedore Core virgin -- for both of you, Linux is a religion.

Reply Score: 4

amazing news!
by JrezIN on Mon 2nd Jan 2006 17:05 UTC
JrezIN
Member since:
2005-06-29

Amazing news! Who doesn't hope for a hardware-accelerated composition and drawing in Linuxes/Solarises/BSD ? It may be closing to costumer-release quality soon for Linux... maybe not too far to the other ones who use X.

...another great thing about making OGL use more common in desktop, are the pressure for more and better drivers from every vendor who supports Linux and all these platforms!

Reply Score: 2

Correct decision
by mals on Mon 2nd Jan 2006 17:44 UTC
mals
Member since:
2006-01-01

Xgl is far too important to be developed behind closed doors. Thankfully Novell realised this, although it did take some pushing. Remember, one of the reasons Xorg came into existance was because the development of XFree86 was not open and transparent enough!

Hopefully this will kickstart X development for 2006. KDE 4 is going to have a lot of graphical enhancements and Xgl can ensure there is no performance hit.

Reply Score: 4

RE: Correct decision
by molnarcs on Mon 2nd Jan 2006 17:51 UTC in reply to "Correct decision"
molnarcs Member since:
2005-09-10

I absolutely agree. When news of criticism reached osnews, people completely missed the points raised by aseigo - now I hope everyone soon will see the benefits of open development.

Reply Score: 4

RE: Correct decision
by kensai on Mon 2nd Jan 2006 20:45 UTC in reply to "Correct decision"
kensai Member since:
2005-12-27

Agreed. 2006 is looking good for Linux. KDE and Xorg are tacking steps in the right direction and now XGL. I'm loving 2006 as we speak.

Reply Score: 2

Thanks Aeron!
by Guppetto on Mon 2nd Jan 2006 17:53 UTC
Guppetto
Member since:
2005-07-06

For everyone dumping on a certain KDE developer for bringing the closed door development of XGL to light, those harsh words should be converted into an apology. Had he not publically brought Novels little project to public attention, does anyone believe that this would have happened.

Now the good news.

I think that the news story of the XGL and EGL work that the KDE developers have started doing, also played a part in this project suddenly becomming open.Novel knows that given the choice of using thier version of XGL or the version developed by the KDE developers and Troltech for KDE, that they would be fighting a loosing battle. The good news is that now, XGL and more importantly EGL will be here a lot sooner rather than latter, if Novel, KDE, Troltech and the Red Hat developers take a few moments to sit down in a room and combine their efforts.

Also note:

It's clear that Novel has also been talking to both ATI and NVidia, because as the article states, if you're using their proprietary drivers, you can already run a fully accelerated desktop with the upcoming cvs commit; that means a lot of work has already been done on this project.

Reply Score: 5

Important to note
by siride on Mon 2nd Jan 2006 18:02 UTC
siride
Member since:
2006-01-02

I think it's important for everybody to read this post by one of the Xorg devs: http://lists.freedesktop.org/archives/xorg/2005-December/011803.htm...

Basically, he says that there probably wasn't anything sinister going on. Now that the code is being merged, I think it's proof that a lot of people were getting upset over nothing.

Reply Score: 5

RE: Important to note
by segedunum on Mon 2nd Jan 2006 19:03 UTC in reply to "Important to note"
segedunum Member since:
2005-07-06

Basically, he says that there probably wasn't anything sinister going on. Now that the code is being merged, I think it's proof that a lot of people were getting upset over nothing.

Rubbish. Without people getting upset, as you put it, there would have been no public development and Dave Reveman would not have been so anxious to get his previously closed code (which might not have been released to anyone until February and beyond) into the public CVS.

1 - 0 to Aaron Seigo there I think.

He also says:

If people work in different directions and don't cooperate on a given feature then it will take longer for that feature to happen and the user community will suffer as a result while the efforts converge.

Which proves many of the points made when people got upset.

And also:

If any given party stops contributing open code, well, tough for them, it's not like their commits can be undone.

Which is why Dave Reveman is back-pedaling trying to get his code in, which might end up being useless if it doesn't get committed soon.

No, it proves that there was indeed something sinister going on, people getting upset has actually got results and things might actually really happen now as a community.

Reply Score: 0

RE[2]: Important to note
by null_pointer_us on Mon 2nd Jan 2006 20:04 UTC in reply to "RE: Important to note"
null_pointer_us Member since:
2005-08-19

Rubbish. Without people getting upset, as you put it, there would have been no public development and Dave Reveman would not have been so anxious to get his previously closed code (which might not have been released to anyone until February and beyond) into the public CVS.

When I was little, I remember reading a little story about a rooster. You see, every morning he would jump onto the fencepole, take in this gigantic breath of air, puff out his chest, and crow. Then the sun would rise, filling the farm with light and warmth. Pretty soon he began to think that his crowing made the sun rise. So he told all his friends, but no one else believed him. To prove himself, one day he decided not to crow; naturally, he was crushed when the sun rose anyway.

It's easy to say that because the code is opened now, your criticism has cause it, but the bottom line is that I don't see any information on why the code was released. It could be that this date was set before the criticism began or that the code was going to be released anyway when it was ready for merging into X. The latter is what seems to be the case from the link provided in the post you were responding to. Not only that, but it specifically asked people to avoid the silly political debates:

There haven't been any checkins to the xgl DDX in the old modular X server tree for a few months. This does not necessarily mean that people have stopped working on it. It doesn't even necessarily mean that the developers have "taken it closed". Heads-down development time is simply a reality, even for open projects, and if the developers feel their work is incomplete or not ready for public consumption then that's their right. There are many examples of highly successful open source projects that simply were not available to the public before their initial release - Mozilla, just as an example.

...

I will continue to work on enabling X on GL in the open code, and I suspect I'm not alone there. Political debates about who's playing nicely with whom don't get development done any faster, so let's not have them, k?

That cuts both ways. Don't start those debates, and don't give cause for them to exist.


Why are you using his message to further such debates?

Which is why Dave Reveman is back-pedaling trying to get his code in, which might end up being useless if it doesn't get committed soon.

What is the basis for saying that he is back-pedaling?

Reply Score: 5

RE[3]: Important to note
by Guppetto on Mon 2nd Jan 2006 20:41 UTC in reply to "RE[2]: Important to note"
Guppetto Member since:
2005-07-06

You make some good points, but I think all of the crowing that occured deffinitely influcenced their decision to suddenly give back some of the code that they weren't obligated to return. That or the fact the the KDE developers were about to embark on their own version of XGL and EGL, which would not have been the best development for Novel considering how many people use and love KDE.

The open source communitty has been known to bitch and moan from time to time, but we don't screem bloody murder unless their's a reason. Combine Novels suspect apparently closed door development along with all the SUSE and Ximian crap that has happened over the past few months, and I believe it was reasonable to question thier motives, even if thier motives were not corrupt. They are at the end of the day, a big business, and the primary goal of large companies is to keep the money train rolling. A pretty picture has made a number of people welthy throughout history and XGL and EGL will provide the prettiest picture your going to see on Linux.

All that aside, checking the code in still doesn't insure that the KDE developers, Troltech, Redhat, or Novel will work together to complete this project. I would love to hear what the KDE developers stance is on working together on this projeect along with Redhat's position. Remember, this project started out in the hands of a number of Redhat developers if I remember correctly, so their are a lot of people (but only a small group of knowledgeable developers) that have the ability to make this project a reality if they could all get together.

Reply Score: 1

RE[3]: Important to note
by segedunum on Mon 2nd Jan 2006 21:52 UTC in reply to "RE[2]: Important to note"
segedunum Member since:
2005-07-06

When I was little, I remember reading a little story about a rooster..................................

Your long-winded attempts at saying that this would have all happened anyway are pretty laughable.

I have something a bit shorter that actually means something. If it walks like a duck, flies like a duck, swims like a duck and quacks like a duck it's a safe bet that it's a duck ;-).

but the bottom line is that I don't see any information on why the code was released.

What code was released? If you're talking about Dave Reveman's then it hasn't been merged yet and the original plan (if there was one) was to release it around the time of XDevConf. There was certainly going to be no code dump or merge until this started being debated.

Why are you using his message to further such debates?

Because I'm not - you are. The pointless stuff you've quoted does not contradict what's actually happened. What's happened has happened, and it's pretty clear why, as I've stated.

As he says, the subject deserves no further debate.

What is the basis for saying that he is back-pedaling?

This message:

http://lists.freedesktop.org/archives/xorg/2006-January/011922.html

At this point in time after development has just been happening on XGL, and it even gives a description of what's going on - something which has been completely lacking at any stage until now.

The positives are that this looks as if it is all going to be merged, hopefully, but I don't think anyone should be insulting anyone else by painting over or denying why. Let's just leave it at that and let them get on with it.

Reply Score: 0

v RE[4]: Important to note
by segedunum on Tue 3rd Jan 2006 13:14 UTC in reply to "RE[3]: Important to note"
RE[4]: Important to note
by null_pointer_us on Tue 3rd Jan 2006 17:31 UTC in reply to "RE[3]: Important to note"
null_pointer_us Member since:
2005-08-19

Your long-winded attempts at saying that this would have all happened anyway are pretty laughable.

I didn't say that it would have happened anyway. I asked you to stop claiming otherwise until you have some evidence to support your claims.

I have something a bit shorter that actually means something.

Of course my story meant something. I even went to the trouble of stating the meaning so that you couldn't accidentally miss it: It's easy to say that because the code is opened now, your criticism has cause it, but the bottom line is that I don't see any information on why the code was released. So just how did you manage to miss my point?

If it walks like a duck, flies like a duck, swims like a duck and quacks like a duck it's a safe bet that it's a duck ;-).

That's a flawed analogy. Not only are you claiming to have seen and heard a duck that no one can present any hard evidence for, but you are demanding recognition for having shot it to save the community.

What code was released? If you're talking about Dave Reveman's then it hasn't been merged yet and the original plan (if there was one) was to release it around the time of XDevConf.

My bad, I meant to say that the project was being opened.

There was certainly going to be no code dump or merge until this started being debated.

I see that you can restate your hypothesis. How about backing it up? C'mon, show us that you know what Dave was thinking when he began the discussion on the mailing list. Where's the crystal ball?

Because I'm not - you are.

I asked you to stop demanding that everyone accept your viewpoint in lieu of evidence that it is correct. IOW, put up or shut up. How is that continuing the debate?

The pointless stuff you've quoted does not contradict what's actually happened.

"What's actually happened" is that the project is being opened. I never contradicted that. Moreover, the stuff that I quoted explained just how little hard information we had on what was going on with XGL. I can understand you disagreeing with my posts, but how do you continually fail to understand them?

What's happened has happened, and it's pretty clear why, as I've stated.

You've stated this repeatedly.

> What is the basis for saying that he is back-pedaling?

This message:

http://lists.freedesktop.org/archives/xorg/2006-January/011922.html

At this point in time after development has just been happening on XGL, and it even gives a description of what's going on - something which has been completely lacking at any stage until now.


We have established that the project has been opened. What we are talking about is why.

The positives are that this looks as if it is all going to be merged, hopefully, but I don't think anyone should be insulting anyone else by painting over or denying why. Let's just leave it at that and let them get on with it.

I asked you back up your claim. You really shouldn't take that as an insult.

Reply Score: 2

Good news
by mOrPhie on Mon 2nd Jan 2006 18:34 UTC
mOrPhie
Member since:
2006-01-02

I don't understand the reply on this good news posted by (I asume) an x.org developer: http://lists.freedesktop.org/archives/xorg/2006-January/011925.html

Novell may not have done it in the public (until now), but they have done a lot. Now, the developer is asking for some help to get this into X.org. This would be good for everybody. X.org benefits from it, so the users will too. So, why the angry reply? This mentality doesn't bring the users a better xserver. The only thing I see from that reply, is frustration. I think we should all be reasonable about this. It might not be the best way of developing open source software, but it got us result. Great results. ;)

Reply Score: 2

v RE: Good news
by segedunum on Mon 2nd Jan 2006 18:51 UTC in reply to "Good news"
RE[2]: Good news
by superstoned on Tue 3rd Jan 2006 14:51 UTC in reply to "RE: Good news"
superstoned Member since:
2005-07-07

its stupid they are voting you down, as you're just right...

Reply Score: 1

X development...
by Hamman on Mon 2nd Jan 2006 18:52 UTC
Hamman
Member since:
2006-01-02

From having been almost dead, the development of X reaches heights no one dared to predict a few years ago. First the release of X11R7 and now XGL.
When I first heard about the plans for Xorg 7.0 I thought it was a mayor undertaking and wasn't sure that the devs could finnish it. XGL always seemed like a pipedream to me, to be honest, and with one of the devs leaving I thought the project to be dead.
To my great joy I was definately wrong in the first case, and my predictions about XGL seems equally untrue. I'm not gonna finnish my post with a cliche like "this will be the year of the linux desktop", but with the recent turn of events Linux's future as a desktop OS appears a lot brighter to me.

Reply Score: 3

RE: X development...
by Snifflez on Mon 2nd Jan 2006 19:33 UTC in reply to "X development..."
Snifflez Member since:
2005-11-15

"I'm not gonna finnish my post with a cliche like 'this will be the year of the linux desktop' [...]"

Thank you. Whoever conned the phrase ought to be shot in the head, since it assumes that Linux is not present on the desktop, which is utter nonsense.

Reply Score: 3

RE[2]: X development...
by kaiwai on Tue 3rd Jan 2006 01:00 UTC in reply to "RE: X development..."
kaiwai Member since:
2005-07-06

Thank you. Whoever conned the phrase ought to be shot in the head, since it assumes that Linux is not present on the desktop, which is utter nonsense.

I personally prefer the old KDE phrase of "is the desktop ready for UNIX?"

The whole 'year of linux' was started by, IIRC, a magazine writer a long time ago - and its based on the stupid assumption that all end users have the same requirements.

Reply Score: 3

Drivers and video cards
by bsharitt on Mon 2nd Jan 2006 19:02 UTC
bsharitt
Member since:
2005-07-07

I just hope they don't lock XGL into high end video cards and propriety drivers. If I'm not mistaken, even OS X(the x86 version) can use Intel's integrated 3D graphics chipsets for its eye candy.

Reply Score: 1

RE: Drivers and video cards
by segedunum on Mon 2nd Jan 2006 19:09 UTC in reply to "Drivers and video cards "
segedunum Member since:
2005-07-06

I just hope they don't lock XGL into high end video cards and propriety drivers. If I'm not mistaken, even OS X(the x86 version) can use Intel's integrated 3D graphics chipsets for its eye candy.

Unless people were to go mad, like with Jon Smirl's proposals, then yes you might, but there's no reason for things to be like that. Hopefully it will push the boundaries of what's possible with today's (and yesterday's) graphics cards and get hardware acceleration into a sensible and rational state first. Cart and horse in the right order.

Edited 2006-01-02 19:12

Reply Score: 0

RE: Drivers and video cards
by kaiwai on Tue 3rd Jan 2006 00:56 UTC in reply to "Drivers and video cards "
kaiwai Member since:
2005-07-06

I just hope they don't lock XGL into high end video cards and propriety drivers. If I'm not mistaken, even OS X(the x86 version) can use Intel's integrated 3D graphics chipsets for its eye candy.

The ideal senario, in a perfect world, would be for an opensource, openstandards graphics card architecture to be created, like what SPARC is, and is completely focused on just supporting OpenGL 2.0 - if enought companies got behind it, like Apple, IBM, SUN etc. since all these companies sell systems, not graphics cards, they would actually be better off in that the graphics cards they would use, would only need to be made at cost with a small profit for R&D, which by enlarge could be spread accross all companies who wished to be involved.

But hey, like I said, its a dream unfortunately :-(

Reply Score: 3

Good News
by DonS on Mon 2nd Jan 2006 19:02 UTC
DonS
Member since:
2006-01-02

I think these people need a third party to step in and smooth some feathers a bit so they all can get back to coding. The last thing we need in OSS is more forks in code and endless fights over turf.

Reply Score: 1

RE: Good News
by s_groening on Mon 2nd Jan 2006 19:06 UTC in reply to "Good News"
s_groening Member since:
2005-12-13

Correct!

We do not need all the different forkings of code, we need more design before coding to make sure the 'right' decisions are made so that coders can code udistractedly.

Reply Score: 1

X Development
by viator on Mon 2nd Jan 2006 19:42 UTC
viator
Member since:
2005-10-11

Excellent news indeed puts a smile on my face. Open is always the best way to go. ;)

Reply Score: 1

Open Source at it's best :)
by ThawkTH on Mon 2nd Jan 2006 20:04 UTC
ThawkTH
Member since:
2005-07-06

By the community, for the community!

Close development, and you might as well just close the source.

Nobody knows what's best for the users than, well, the users...

Maybe this will help push Nvidia/ATI to help out with some BETTER drivers. Release specs for the OLD cards that aren't even sold anymore, perhaps?

*Dreams of a hardware accelerated KDE 4*

Reply Score: 2

Great news but..
by NxStY on Mon 2nd Jan 2006 20:08 UTC
NxStY
Member since:
2005-11-12

This is of course great news but most cards still lacks good accelerated drivers. Nvidia cards for example needs nvidia's proprietary drivers that distributions can't ship. If XGL is one day going to be the standard X-server (which I believe is the plan) then vendors needs to release their card specifications so good drivers can be written!

Reply Score: 3

One concern is ..
by de_wizze on Mon 2nd Jan 2006 20:17 UTC
de_wizze
Member since:
2005-10-31

.. that I hope that this doesn't mean useless feature requests that distract from the true focus of its developments. Don't get me wrong. I applaud the motion but I just worry for the focus and direction from now on.

Reply Score: 1

closed source WAS good
by indech on Mon 2nd Jan 2006 21:10 UTC
indech
Member since:
2005-12-06

Having a big company that has set ideas like Novell, it can be a good thing for them to get their ideas into the code and once they develop it to a state where they believe they have their ideas developed and are ready for criticism then they can release it.

The problem with open source at times is there are too many voices and the goal is lost in all of the unecessary chatter. Thus I do not blame them for keeping it closed till the time was right. Opening them in the end was a good idea, however.

Reply Score: 2

sweetness
by SEJeff on Mon 2nd Jan 2006 21:31 UTC
SEJeff
Member since:
2005-11-05

For X development to pick up, the largest milestone has already been released. x.org 7.0 modular has been officially released. One of the second biggest things is a hardware accelerated X server and a stable compmgr.

Now that xgl is starting to pick back up, I wonder what kind of work (if any) has went into luminosity?

http://live.gnome.org/Luminocity

Hardware accelerated KDE would be cool, but so would hardware accelerated Gnome.

Edited 2006-01-02 21:32

Reply Score: 1

RE: sweetness
by superstoned on Tue 3rd Jan 2006 09:47 UTC in reply to "sweetness"
superstoned Member since:
2005-07-07

luminocity is dead... it was meant to be an example of what is possible. cairo might once bring what luminocity did show... tough i bet KDE 4 will be the first fully GL/hardware accellerated desktop environment for linux. well, E17 is working on it, too, afaik! gnome tends to be behind on this kind of things, so i guess luminocity was something cool but not to stay.

Reply Score: 1

The path forward
by jonsmirl on Mon 2nd Jan 2006 21:54 UTC
jonsmirl
Member since:
2005-07-06

The are several paths forward that can be taken. Given that we have very few resources pursuing all of them ends up with none of them being finished. Here are some of the basic choices:

1) Least common denominator. This is the current X server. The GPU is ignored and used to heat the room. Same xlib API on all hardware from low end to highest end. Stay this way forever and ignore GPU advances.

2) Start adding acceleration features to the xserver. That is the EXA scheme. Pick the low hanging fruit and implement it one feature at a time. So far ATI/Nvidia have not participated in this scheme. Over the long run it could slowly reimplement the features OpenGL into the xlib API. Software emulation is used when the hardware doesn't have the feature. You may end up with many different xlib versions in the different server releases. (This is the path followed by DirectX - 1, 2, 4, ....).

3) Go full OpenGL. This path capitalizes on the existing OGL drivers from Mesa, Nvidia, ATI. It jumps straight to the final accelerated graphics API. Old hardware is handled with software via the Mesa OGL implementation. SW Mesa needs a lot of performance work to get decent speeds out of old hardware. xlib is a compatibility layer via XGL. Single OGL API accross all hardware. Apps check acceleration levels and adjust UI accordingly. (Mac and Longhorn on this path).

4) Server inside of server. This is the current XGL model. Run the full X server and the full XGL server at the same time. XGL runs inside the XServer as a full screen app. This lets XGL get OpenGL support from existing drivers. Avoids the need to build standalone drivers. Downside - it's a resource hog.

5) Fork - low end machines stay on current X server. Highend goes OpenGL. Highend keeps xlib compatibility layer. Over time, high performance apps will be written to use OpenGL API which does not exist on low end machines.

What is the future graphics API? Is it xlib, xlib plus extensions, OpenGL or something else?

We really need to make a choice. There are clearly not enough resources to chase everything.

Edited 2006-01-02 21:56

Reply Score: 5

RE: The path forward
by IceCubed on Mon 2nd Jan 2006 22:27 UTC in reply to "The path forward"
IceCubed Member since:
2005-07-01

5) Fork - low end machines stay on current X server. Highend goes OpenGL. Highend keeps xlib compatibility layer. Over time, high performance apps will be written to use OpenGL API which does not exist on low end machines.

Maybe not a fork, but a backend.

Make 2 backends - 1st being the highend XGL backend, 2nd - the `traditional` one.

The problem with this aproach is that the first one will get the drivers and the second one will won't.
So eventualy the `traditional` backend will become deprecated.

Reply Score: 1

RE[2]: The path forward
by jonsmirl on Mon 2nd Jan 2006 22:43 UTC in reply to "RE: The path forward"
jonsmirl Member since:
2005-07-06

Forking also involves forking the graphics API. What do you do about things like pixel shaders which are only available on high end hardware? They are incredibly useful in building high end UIs. You can create visual appearances with them that simply are not possible to do without them. Do pixel shaders get added to the extended xlib API or do they only exist in the OGL API? Do you software emulate (performance sucks) them or provide a capablility bit indicating their presence?

Choosing the path forward is not a simple decision.

Reply Score: 1

RE[2]: The path forward
by siride on Mon 2nd Jan 2006 22:59 UTC in reply to "RE: The path forward"
siride Member since:
2006-01-02

> So eventualy the `traditional` backend will become deprecated.

As will the hardware...

There are basically two big options: keep doing the same thing or change course. The former is just not going to cut it anymore for Linux to be competitive, or even usable as software and desktops become more demanding. There are certainly difficulties in changing the architecture, but those difficulties are really just challenges to create a workable system and not excuses to keep the status quo.

Edited 2006-01-02 23:01

Reply Score: 3

RE: Great news but..
by re_re on Mon 2nd Jan 2006 23:46 UTC
re_re
Member since:
2005-07-06

I very highly doubt that nVidia or ATI will be opening up their specs for the world to see any time soon, it would be excellent, but not likely.

What I do see as a definite possability and probably much easier to swollow (for nVidia and ati) is nVidia and ATI changing their liscensing on their proprietary drivers to allow them to ship bundled with a free os.

Reply Score: 2

r_a_trip
Member since:
2005-07-06

3) Go full OpenGL. This path capitalizes on the existing OGL drivers from Mesa, Nvidia, ATI.

This only works if decent 3D accelerated FOSS drivers exist for Nvidia and ATI cards.

A large portion of the community will not accept an encumbered closed source foundation in the form of proprietary graphics drivers just to enjoy some eye-candy.

While eye-candy and more appealing interfaces are nice, they must not come at the cost of being unable to build fully Free OSes.

If the price of OGL graphic underpinnings is going closed source for the drivers, we might as well go Apple or Micosoft, because the core principal of FOSS is lost on a dependant, mixed system.

Reply Score: 2

jonsmirl Member since:
2005-07-06

If the price of OGL graphic underpinnings is going closed source for the drivers, we might as well go Apple or Micosoft, because the core principal of FOSS is lost on a dependant, mixed system.

I don't think we have the right to be making that decision for everyone. Each person should be free to choose what they are comfortable with. For example I have no problem with running binary NVidia drivers while you may have a problem with it. In that case, you shouldn't buy an Nvidia card and instead buy an ATI 9XXXX series which has a free driver available.

If you stick with "open source or die" and ATI/Nvidia never opens their hardware then we might as well forget about Linux on the desktop. Within the next three or four years having an accelerated GUI is going to become a requirement and not an option on desktops. What are you going to do if no hardware on the market in that timeframe has open drivers available?

That is my fear, by the time the community figures out that accelerating the GUI is a requirement, not an option, we will have five years of catch up programming to do and no one will undertake such a large task. At that point Linux on the desktop is a lost cause.

Reply Score: 1

zerblat Member since:
2005-07-06

There are still a few more options:

Creating free drivers for newer hardware through reverse engineering: http://r300.sf.net

If you're not a gamer, there are cards with open source drivers which might be sufficient: http://dri.freedesktop.org/wiki/Status

Creating open hardware: http://wiki.duskglow.com/index.php/tiki-index.php?page=AboutOpenGra...

Reply Score: 3

jonsmirl Member since:
2005-07-06

Obviously we will keep pursuing free drivers where ever possible. I lean on ATI/Nvidia to release specs every chance I get. But it's their hardware and they don't have to release specs if they don't want to.

However, I don't agree with the logic that says, since there are no free high level drivers available then X should not expose high level features. That's option #1, continue with the status quo. In the long run that strategy can become a complete disaster if all future video hardware goes with closed source drivers. At some future point there may be no more open hardware available.

Reply Score: 1

Lobotomik Member since:
2006-01-03

This myth about the unexistence of open drivers keeps popping up again and again.

Truly, nVidia support with open drivers sucks badly, but ATI has pretty solid open support for older cards (up to Radeon 9200 or so) and solidifying support for things up to X800 (and maybe beyond) in the r300 open driver.

Also, there is a big, big world outside nVidia and ATI. Let us not forget about marginal players like S3 or XGI, with varying openness, but the VERY VAST majority of 3D chipsets sold are made by Intel; integrated in most desktop and laptop computers, they are perfectly supported by open source drivers.

Intel 3D chips are not by any means the fastest, but most people seem to never notice, and they are much more than enough for the sweetest desktop eye candy.

So, as things are these times, you really need to depend on closed-source drivers only for anything nVidia and very new Radeons; if best gaming performance is not what you're after, there is plenty to choose from (at much lower prices).

And then, of course, if best gaming performance is indeed what you are after, nowadays Linux is sadly not what you need, because there aren't many games for it, to start with.

Edited 2006-01-03 08:35

Reply Score: 2

poofyhairguy Member since:
2005-07-14


Also, there is a big, big world outside nVidia and ATI. Let us not forget about marginal players like S3 or XGI, with varying openness, but the VERY VAST majority of 3D chipsets sold are made by Intel; integrated in most desktop and laptop computers, they are perfectly supported by open source drivers.

Intel 3D chips are not by any means the fastest, but most people seem to never notice, and they are much more than enough for the sweetest desktop eye candy.


GREAT point. I think everyone misses that. Intel is on the side of the open source crowd and they are the biggest player in the desktop graphics market.

I mean....Intel's 915 video hardware is the only Directx 9 compatible (aka has the pixel shaders) hardware with open specs right? Why not just build the entire future Xserver on that? Then ATI and Nvidia will have to play along or lose the market and those that demand an all OSS system can have it!

http://downloadfinder.intel.com/scripts-df-external/Filter_Results....

Reply Score: 3

Don't discount those who value FOSS
by r_a_trip on Tue 3rd Jan 2006 00:50 UTC
r_a_trip
Member since:
2005-07-06

If you stick with "open source or die" and ATI/Nvidia never opens their hardware then we might as well forget about Linux on the desktop.

A proprietary driver should always be optional, but never a required default. If FOSS has to go partly closed source to "win" on the desktop, it has already lost.

If FOSS again becomes encumbered with any dependancy on closed source it doesn't matter if it has larger uptake. Then the core reason why FOSS exists in the first place has been slaughtered.

First get a fully Free implementation out the door and then enable the choice of tainting the system for those who are comfortable with it. Otherwise I see XGL and EGL chasing the proprietary drivers of the Graphics Leader du Jour for all eternity (be it ATI, Nvidia or an unknown third party) and a fork of X.org to keep it unencumbered from CSS driver dependancies for those who do care.

Even though you don't mind giving up Freedom in core infrastructure, it doesn't mean that others don't. Dependancy on closed drivers for display is the same as the Java trap. Your software is open, but foundation is not, still shackling people to money interests of lockin companies.

Reply Score: 2

jonsmirl Member since:
2005-07-06

If you stick with "open source or die" and ATI/Nvidia never opens their hardware then we might as well forget about Linux on the desktop.

A proprietary driver should always be optional, but never a required default. If FOSS has to go partly closed source to "win" on the desktop, it has already lost.


Optional will depend on which hardware you buy. XGL has free drivers available for a lot of hardware: http://dri.freedesktop.org/wiki/Status

But a proprietary driver is not optional if you buy Nvidia hardware. If you own their hardware then the proprietary driver is required. But you weren't forced into this situation, you got there by choosing to buy NVidia. If you had chosen ATI you could choose between open Mesa and fglrx.

I should also point out that there are no free drivers available for ATI's latest chips either.

Edited 2006-01-03 01:09

Reply Score: 1

kaiwai Member since:
2005-07-06

A proprietary driver should always be optional, but never a required default. If FOSS has to go partly closed source to "win" on the desktop, it has already lost.

Babe, there are alternatives. I used Matrox cards with UNIX's for MANY years, and guess what, they still sell the Matrox G550, which is completely opensource driver, all the features supported, is available for both AGP and PCIe.

Now sure, it isn't bleeding edge, but I am sure for what most people need to use it for, it'll do the job quite nicely - and you can use it with the nice warm fuzzy opensource feeling that you long for.

Reply Score: 1

Maciek
Member since:
2005-11-15

Even though DavidR is now focusing on a new compositing manager (compiz), the previous one which he has been working on until now has been made available:

http://cvs.freedesktop.org/xorg/app/glxcompmgr/

Zack Rusin from KDE already seems to have this working nicely. From his blog entry yesterday:

On the xorg front Eric got Xgl working with xorg cvs, which is extremely cool. I added an older version of glxcompmgr to xorgs/app cvs. Dave has been doing some amazing EGL work. Try it all out, update xorg cvs, compile Xgl, checkout glxcompmgr and give it a testspin. I'm hoping to do most of the Plasma desktop code this week and do some Xgl magic soon after so keep watching cvs.

Reply Score: 4

somebody Member since:
2005-07-07

Zack Rusin from KDE already seems to have this working nicely.

Seems that those are two different things they were working on.

From XOrg mailing list
http://lists.freedesktop.org/archives/xorg/2006-January/011926.html

[from e-mail]"As far as I can tell it seams that they've pretty much only done work on the egl specific code which I haven't worked on so that's going to be really easy to merge."

Reply Score: 1

Headrush
Member since:
2006-01-03

I think a lot of people assume the average user cares about having an Open Source system, but most don't.

I think a large majority of users move to Linux for the flexibility, the cost, and they just want a system that works.

I believe if Windows was what we have in Linux now, but closed source, most of us would be still using it and would not have looked for alternatives.

Reply Score: 1

somebody Member since:
2005-07-07

I use linux 99% of my time for 8 years now.

I think a large majority of users move to Linux for the flexibility, the cost, and they just want a system that works.

No. Cost and flexibility, yes. System that just works??? We could debate on that. If you bought hardware carefully with Linux in mind, yes. Often works better than you could dream on Windows or OSX. Typical user does not do that and that leads to a lot of problems.

XP just works "in their own view" and right conditions. HW devels support this OS.

OSX does the same. But you can't buy wrong hardware.

Linux??? It is mostly set up as second in mind. And people do expect more from it. Too much if you ask me
Two examples,
1. everybody bugs developers for NTFS write support (ntfs is covered with patents), nobody bugs MS for even read ability of ext3 (and no one expects that from Windows, while Linux often "sucks" for not providing that). Hell Windows even screws your MBR when reinstalling. But people treat that as linux problem
2. people often buy some piece of HW (that doesn't conform to any real world spec and HWvendor does not want to give specs) (even Win drivers were a hack in a first place) and say "linux sucks because it doesn't support my ..." not even caring they haven't bought something supported by linux, but they bought something that was just cheap and available.

"Just works" is only a case when you started correctly when buying HW.

I believe if Windows was what we have in Linux now, but closed source, most of us would be still using it and would not have looked for alternatives.

Again, you believe wrong. Most just like to hack inside. For example, I like to change, tweak and screw with services. Can't do that in closed source.

Reply Score: 3

Any instructions available ?
by ghaefb on Tue 3rd Jan 2006 08:47 UTC
ghaefb
Member since:
2006-01-01

Are there any instructions on how to make this baby work ? ;)

Reply Score: 1

Thanks
by zam001 on Tue 3rd Jan 2006 09:10 UTC
zam001
Member since:
2005-08-12

Thanks David!!! and Thanks everyone who gave their voice here at osnews to make this happen....

Reply Score: 1

another perspective
by 25bravo on Wed 4th Jan 2006 17:37 UTC
25bravo
Member since:
2006-01-04

Windows is good. Everybody knows it. And windows is doomed, and everybody knows it. Linux will likely rule the world some day because it is starting to put more food on the table than windows, in a lot of areas of technology.

What will decide the fate of OSdom: Freedom, fear, and food. That's what Windows still has going for it, Fear. In the food department, Windows is still a bit ahead, but linux, with it's freedom, has the potential to feed ten times more mouths. As they say, teach a man to fish...

Anyway, XGL is just another step towards alleviating the fear factor for me. Long live freedom, aye?

Reply Score: 1