Wouldn't it be nice if we were older
Then we wouldn't have to wait so long
And wouldn't it be nice to live together
In the kind of world where we belong - the beach boys
it would be very nice if we our X server could use OpenGL directly for its display and composition. because then we could have hardware accelerated effects that are not only cool looking, but also very useful. well, there is just such a project underway, called XGL
but don't hold your breath. the development of XGL has been largely removed from the community and is being done behind closed doors. when i first heard about this state of things i figured there must be more to it but ... nope. it's pretty much the situation.
so instead of getting as many people working on it as possible, instead of allowing other projects to test their software and play with its capabilities, instead of going through the peer-review process that keeps things healthy ... one company has decided that it would be a better idea to hire a few of the developers and put the development into a source repository that only they have access to.
this is worse than your garden variety fork since unlike your usual fork the forked codebase isn't available for others to poke, prod and work on. who gets hurt? everyone who would benefit from this technology, which are, in no particular order: users, users, desktop projects and their users.
this is resulting in other vendors to go off and do their own things. as you probably know, we're already fairly short on graphics programming talent, so this balkanization is not really what we need. it also smacks distinctly of non-Free practices.
i've waited for some time now for this to come to a happy resolution but i see only the current status quo being kept. it seems nobody is willing to openly talk about it, and i understand why: people don't want to rock boats and have uncomfortable conversations in public. but at some point i have to wonder that if nobody is held publicly accountable, if no feet are put to the fire of public awareness and scrutiny, will anything change? i don't think so, because this is a "business decision". this isn't a decision being made within the open source world, but within the board room. and i'm really not happy about seeing a large opportunity go sailing past because of board room politics.
so ... who is this company, you ask, that would take the development of something as potentially important as this out of the community and put it behind closed doors? novell.
You know it seems the more we talk about it
It only makes it worse to live without it
But lets talk about it
Wouldn't it be nice - the beach boys
47 comments:
Well, so I know which company it is. I guess you don't know _why_ they're going that way? I mean, some fd.o blogger supposed they are doing it so that they can put out a big announcement and praise themselves as The Great X-Server Saviours, but that just can't be it, can it?
They can't just go that way without giving a reason, no? Maybe there's someone who gets an answer from the Novell people when asking publicly enough?
I'm confused. The link you give shows how to get the code from CVS, and thus presumably test it. Is your problem that they haven't given out committing rights?
And is it just me or does that page make it look like Xgl is dead and Xegl is the replacement?
Richard Stallman has long been saying that X development is broken because it doesn't use the GPL or any other license with forced-sharing provisions.
That's also why graphics companies are allowed to get off with not giving the source code to their binary drivers.
I wonder -- what would it take to create a fork of X.org that allows people to submit GPL'ed improvements? Not much more than a webserver and SVN server. Such a change just might revamp the enthusiasm around it.
I can't actually see the problem. CVS checkout, Feature completeness check, installation instructions.
What's missing that would make it open?
Sadly, they seem to have caught the Ximian disease and that seems to have mutated with the Novell disease they already have. You've mentioned that it will hurt users and desktop projects, but it will ultimately hurt themselves - fatally. It's a simple equation:
Bringing open source projects in-house == shortest way to bankruptcy
Quite why they think that's ever going to work, I don't know. You're hiring developers to hack on a piece of software internally that you're going to perhaps release the code for at a later date (otherwise it's pointless), all for something that doesn't make any money at all for you yet, perhaps ever? Only Microsoft have the cash and resources to bring development like that completely in-house, they have the massive revenues to do it and they're not going to give it away at a later date.
That was and is the whole point of open source software and groups like Freedesktop - cost and resource sharing with the community and other interested companies and projects out there. I'm sure the Red Hat people would be interested, and it would be a good opportunity to try and get nVidia and ATI involved as well as various desktop projects rather than everyone having to do their own thing. The situation of a graphics infrastructure and hardware acceleration on Linux-based platforms is as bad as it is right now:
http://www.freedesktop.org/~jonsmirl/...
As a whole, the X.org community barely has enough resources to build a single server. Splitting these resources over many paths only results in piles of half finished projects.
Bringing a not even half-finished, under resourced sub-project in-house? Sounds like a great business decision to me.
I can't actually see the problem. CVS checkout......What's missing that would make it open?
Errr. The actual source repository?
yes, the CVS on freedesktop.org is over six months out of date. the actual work is being done in a repository that is purposefully being kept internal to a small group of hackers within novell.
the "beauty" of it is they can point to fd.o and use it as a ruse to fool those who are familiar with the details of the project.
the next expected code dump will likely come sometime after they release it as a "finished product" in novell linux OS, sans peer review or other involvement from other people and projects.
why they are doing this is left as an exercise in speculation for the reader.
I'm having mixed feelings about this. Part of me is thrilled because we actually have people working on this. I was afraid that the whole project was more or less dropped. But apparently that is not the case. Not only are they working on it, but they even have some major backing!
On the other hand, I'm very, very disappointed at Novell. Doing things like this behind closed doors is NOT kosher, IMO! I can't really understand what their motives are. This certainly doesn't make developement faster.
Is XGL built on current X-server, or is it written from scratch? Point of my question is that could they re-license the new server if they wanted to?
I don't think the license for X is what lets the drivers stay binary. There are closed binary drivers for all sort of other hardware on linux. They don't actually use the Xserver's code. If you want to argue that I think what you mean to argue is that the Xserver should take pains to make itself incompatible with modules that can't be built from source. Seeing as the linux kernel is a GPL project and is fairly friendly to it though I just don't see it happening.
That doesn't mean X wouldn't benefit from a different license though.
Aaron -- how do you know about the development going on internal at Novell? Where is this being discussed?
You are missing out on the issue that there is a good chunk of the development communtity that does not want XGL to happen. These are the "open source or die" types. This group does not want the desktop to become centered around OpenGL when the only source of decent OpenGL drivers is proprietary. Instead this group wants to limit the desktop to only using hardware features with open specs. Since we don't have specs for any chips designed since 1995 this severely limits what can be done with the desktop. Pursuing XGL with closed development is one solution to working around this group.
Jon: and just how does closed development work around people who are "open source or die" exactly?
were they blockading commit access to cvs by chaining themselves to the routers? ;-P
there are lots of projects that hardliners don't agree with (e.g. media projects that use proprietary codecs) and they don't need to retreat behind closed doors.
more importantly, you are missing out on the fact that there are several people in the community who would like to work on these technologies but can't due to the approach being taken at novell.
blaming others who may not agree with the project's approach for the project's retreat into the bowels of novell is really a pretty lame response.
were they blockading commit access to cvs by chaining themselves to the routers? ;-P
That's a decent description of what is going on.
I suppose it may not be appropriate to make reference to the method with which Oxygen is being developed, here?
This isn't enough information to go jump off the deep end like you appear to be. If you have more details, please share them. Otherwise, until there's more one way or the other, I'll give them the benefit of doubt.
Perhaps this is an internal pet project by an employee? Did you have information showing it to be an official company project?
Or, could this be an attempt to shore up a stable release using a smaller core group of developers? After all, there are some times when a smaller group is more effective than a large group.
I can see this going both directions - attempting to create a closed, proprietary version for Novell and SUSE; or, getting to a certain point and then pulling off the cover at some point when they feel they are ready. What I don't see is enough information to determine which path they are going down.
@zeta: it's called "communication". we even have a word for it! ;)
i've spoken with people on both sides of the issue, including higher-ups within novell and the disaffected hackers outside of novell.
this is not a pet project, it's an officially funded project with a few developers working on it full time. it's part of novell's product plan, etc...
as for being more efficient to have a small group .... there never *has* been a large group around this technology. the number of people who have the skills and ambition to work on it has always been pretty small. and what this is doing is splintering that talent even further and ensuring that people like red hat are doing their own thing instead of jumping on the xgl bandwagon as well (and that's not a good thing)
@john (smirl): your cvs account was revoked by a different group of people for pretty good reasons at the time. it was a completely different issue and has absolutely zero to do with the situation here at novell. nobody was blocking their cvs access, probably because they were actually doing productive work.
(sorry, didn't know it was you otherwise i would've responded more directly the first time)
This group does not want the desktop to become centered around OpenGL when the only source of decent OpenGL drivers is proprietary.
http://r300.sourceforge.net/
not quite finished, but it works. also, all radeons up to 9250 run perfectly on linux and most bsds.
Your idea of open and free seems more than a bit radical. Why do you expect that a for-profit company should not do any internal development? They are still apparently allowing read access to CVS. You can vet the source code. Do you know their reasons? Do you know that they won't release it to the community once they've reached their goal for it?
Apparently not. But you're whining that someone won't let you have a part in doing something that you've got no inherent right to have a part in. Whether they're stabbing themselves in the foot is a different question entirely.
If you don't like their approach, start a rival project and try to get a broad base of supporters from across the industry. We'll see how far you get.
For my part, I'm quite happy with Novell and what they've done for Linux in particular and OSS in general. SuSE has long been among the top distros, and Novell has made SuSE 10 a particularly nice release, IMO.
The problem with the OSS community is that it has to be all or none. Many of its members/proponents are too rabidly against any private development or any commercial software. There's really no problem with having software produced by experts and sold, except that you don't want to pay for it. Unless you just don't trust anything you can't modify yourself. That's paranoia, except when its Microsoft. So go write it yourself then.
Really. Grow up.
I can understand the motivations for writing free software and using software that's free. I can even understand wishing that certain proprietary software was free as well. However, Novell has no obligation to anyone (except perhaps to their paying customers and shareholders) and does not need to justify their decisions in this matter.
"if nobody is held publicly accountable, if no feet are put to the fire of public awareness and scrutiny, will anything change?"
Novell has done nothing wrong. Contributing to open software is nice, but Novell's proprietary software projects are nobody's business but Novell's.
If all you're saying is that Novell is a big meanie, then I can agree with that.
By the way, why is everyone a "hacker" and not just a "programmer" or "developer?" More glamorous sounding, I guess.
I'm curious about people's opinion on this: Steven Weber in his great book The Success of Open Source makes a claim that open source projects do better when they start with a central vision or engineering concept that has been developed up to a point where it can be opened up to be worked on. He also talks about the Mozilla source release case study, where the code was found to be unuseable because of the closed-source coding practices. So on the question of whether Novell is making a good business decision, is there a consideration that they need to develop a solid engineering concept *in order* to guarantee their investment, so that it will be good enough for people to want to work on it? And is there a point where in-house development can break a project for opensource purposes? Sorry about the incoherence, I'm writing this at work within sight of my boss.
Novell hired David Reveman in feb. 2005. (source http://www.nat.org/2005/february/#9-February-2005). I'm sure Novell is looking at XGL for their NLD product. It's a good thing that there is development on XGL. I'm sure Novell will make this open source over time.
Hmmm, seems Novell doesn't think that an open source development model works in this case. (Or they want a proprietary X server that they can sell...)
I've blogged a little about this (with eye candy).
You are missing out on the issue that there is a good chunk of the development communtity that does not want XGL to happen. These are the "open source or die" types.
"Open Source" and "cool XGL-based graphics" are not contradictory. In fact they can support each other and be achieved simultaneously. Instead of keeping XGL development in-house, Novell would be much better off sponsoring the Open Graphics project to create accelerated graphics cards with open specifications and fully-supported under Linux.
Here's an interview with Timothy Miller, founder of the Open Graphics project, giving an overview of it purpose.
I'm also very excited about Xara Xtreme, which has possibly the fastest graphics algorithms in existance, and which will soon be released under the GPL.
Dear Jon:
Stop being wrong.
- ajax
Dear ajax,
Prove me wrong and make XGL happen now that you have Redhat to help. Looks to me like there has been zero public progress in the last three months.
jon
>This group does not want the
>desktop to become centered around
>OpenGL when the only source of
>decent OpenGL drivers is
>proprietary.
>http://r300.sourceforge.net/
>not quite finished, but it works.
>also, all radeons up to 9250 run
>perfectly on linux and most bsds
Also supported with open drivers is Intel's i910 family, which is the most popular 3D chip by far (though not the fastest, to be sure). And I believe (though I'm not sure) that S3 and XGI also have open drivers.
So the often repeated idea that there are no open source 3D drivers available actually boils down to no open source 3D drivers for Nvidia, which is not quite the same.
Not wanting XGL because the only good OGL drivers are proprietary seems wrong because if you don't have an accelerated OGL driver than the drawing is going to go through Mesa, aka software, like it does now anyway. Or am I missing something
Jon,
The reason that there hasn't been any public progress on XGL in the past three months (hint: longer than that, really) is because Novell has taken it in-house and shut out the community members who wanted to help.
Hence, the blog post we're all commenting on.
And ajax hasn't been working for Red Hat for three months or anywhere near that long; he just started two weeks ago. (And we're not paying him to work on XGL, anyway; there are more important things that need fixing first.)
(btw, last post was me; no idea why Blogspot lost my credentials, but here, so I'm not just some anonymous loser with a big mouth.)
btw, the anonymous loser with a big mouth in the last two talkbacks was chris lee, who's an otherwise great guy.
now if only he could learn how to log into blogger ;-P
/me runs
Well it was me who posted on fd.o a week or two ago giving out about the same thing, I just didn't want to give out until some people talked to some other people,.. that seems to have happened..
To all the wrongbots who've replied in this thread " you are wrong ".
The last Xgl code drop was from davidr@novell about 6 months ago, there are a group of people who would really like to continue Xegl development (I've been fixing up the r300 driver to it can fully support Xglx and hopefully Xegl). However if I wish to continue to work on Xgl I need to fix a lot of issues that I know davidr has already solved (he told me so by e-mail...).. now it just isn't in the open source spirit if work is being repeated by two people...
Novell I don't believe want to have a prop. X server (like wrongbot 1 or 2 said) they will eventually open source this, but they don't seem to want any community involvment or feedback, they are going to do a huge code drop probably into X.org CVS without anyone ever commenting on it..
To be honest if Aaron says Novell, I'll say Nat.
(And Jon's crap about people blocking people out is just more wrongbot spiel, get over it Jon, you publically gave up working on Xgl, your CVS access to Xgl was revoked, unfortunately it revoked access to freedesktop.org as well, some developers do support other methods like EXA but they are out in the open working on it, they aren't stopping anyone else from working on Xgl....)
I do not agree with "take open source project inhouse = quickest way to bankruptcy"... actually, I see the benefits of taking something like this inhouse... something which has been needed for a LONGGGGG time... and why havent we seen it? because very many opensource projects have no focus, no planning and no management... basically alot of people who cant agree and fork because they think it would be better another way.....and these things usually get to version 0.9 before dying.... By taking this VERY important project inhouse, Novell are speeding up the development and this will benefit them in payback on increased desktop users... (as of course it will for other distributors too)... think about it... everyone is talking about Desktop Linux being the final hurdle.. at least Novell have decided to take something and bring corporate / business planning / management into it.... of course the peer review is VERY important... and as far as I can see... they have NOT closed the source... you can still get it... so I ask... WHAT is the problem?
Wouldn't it make more sense to first improve EXA performance for the r300 before going X(e)gl? I thought that is the whole point behind EXA, to provide decent performance now and the "blue sky" stuff later with Xegl. Right now moving a window around on the desktop makes Xorg use 90% of the CPU spending virtually all its time in fbcopyareammx.
I would rather like to have decent performance in a "normal" setting first before doing all that funky shadow/translucency stuff.
"they have NOT closed the source... you can still get it..."
Please, people, have a closer look at that repository. The last commit that wasn't related to fixing up XGL for the modular tree was half a year ago - none of the changes since then have been made public.
Again: You can't access the code Novell is working on, they keep it secret.
to the last two comments ( I WANT THREADS, lets move this to slashdot :-)
1. EXA on r300, yes I could do it, but I've no interest in EXA, so someone else can do it and I'll be happy. this is my spare time and I'd like to use it in my own way.
2. In one way Novell have not closed the source, we have access to a 6 month old code drop that doesn't build against the latest glitz or Mesa and is missing many of the cleanups added lately. I've actually got a newer code drop by a few weeks that I'm trying to merge into the main tree, however when Novell do finally release all their in-house code, all the time I (and anyone else) spend on this is going to be wasted. What is the point of open source if there is no collaboration, nobody is fighting about what direction Xgl should go, we are mostly quite happy to let davidr lead it in any direction he wants, but some of use would like to hack on it to enable new features and environments (like Xegl and r300) for it to work in.
Who says it hurts the community ?
You can't mix KDE with proprietary software, because of the license. Don't you get it why GNOME will be, and KDE isn't? We need to be able to have people integrate software on and for the most popular free Unix out there: Linux. PROPRIETARY SOFTWARE. All those independent software vendors are *not* going to provide solutions for the Linux/free Unix desktop if they can't integrate. They are *not* going to GPL their software. Dream on...
You GPL zealots haven't a clue how software is done in an enterprise, how seamslessly integrated Microsoft Office is to the document workflow of large corporations and organizations, and how these softwares are built around Windows. Linux/Unix needs that and the only way to do it is to have ISVs writing software for the desktop.
That, my friend, is why Novell and RedHat chose GNOME, because they know their business, because they know they need a business-friendly license.
Why do you badmouth a company because they did a legal thing?
They don't own you a goddamn thing. Grow up!
If the GPL weren't so hostile to companies, we'd have more software. Instead, we get GNU zealots.
BSD folks always understood that. Why is it so hard for Linux developers? PR and media hype gave Linux momentum and now you're having a hard time going cold-turkey on the GPL and facing reality. GPL purity is fine for IBM, because they sell hardware, buddy. It's fine to have proprietary solutions, as long as the core components and protocols are free. Only then you can *really* start talking about selling support as a software developer. It took two multinational companies to drive that message through the heads of Linux fanboys...The more GPL software, the less business...
"If the GPL weren't so hostile to companies, we'd have more software. "
"BSD folks always understood that."
Your holding BSD up as the way to beat a closed source OS? An OS that legally uses BSD inside it...
Are you a flamebot?
Take a look to www.amanith.org , it's the best candidate to build an accelerated desktop... the only pity is the qpl licence.
> You can't mix KDE with
> proprietary software, because of
> the license.
people do it all the time. the licensing does not prevent using KDE technology from proprietary apps, nor running proprietary apps on top of KDE.
it does prevent you from closing the source code up and trying to sell it as your own magic black box. and so does GNOME.
this blog entry has nothing to do with license zealotry (though you obviously have a massive chip on your shoulder regarding that topic). it's about open source development processes.
i have no idea where you get your ideas from, but they aren't informed or intelligent ones.
Novell hasn't chosen for GNOME.
They tried and failed because of
overwhelming protest from the user
community.
If you want to have commercial desktop apps on Linux, the ISVs creating them have several choices.
1. Use gtk/gnome libs.
2. Use the GPL and go Qt/KDE.
3. Pay TrollTech for q Qt developer license (and support) and use Qt/KDE libs.
4. Use whatever other toolkit.
Whatever they do, looking at some of the freedesktop.org standard will be helping to integrate well into both desktops.
I am not a c++ programmer, and i only get to grips with php programming. However after using linux for a year i look close at xgl. To my believe, on one site, it said there wern't enough people to develop these new x technlogies. Whilst any wouldn't be expirienced. Leaving it to say novell or redhat would get it working substantionally done.
Then when it is released people can look at it and drastically modify it. The only thing i don't like is that they haven't give us any news
Novell recently released the Xgl source code. It's actually fairly stable now, and getting better all the time. Reveman has done great work on making this happen.
Let's take account:
- Novell hired the primary hacker for Xgl, David Reveman, thereby allowing him to do what he does best
- Novell seems to have allowed Reveman to manage his project as he saw best
- Novell supported Reveman's re-opening of Xgl
With these facts, when you look back at this old post, are you still certain that Novell was acting with pure evil intent?
@Rev: i don't think they were acting with evil intent. nor did i say that.
there is a saying about never ascribing to malice what one can ascribe to stupidity. i think novell was making some very poor decisions and acting in way that i can best describe as negligent (towards the broader community). everyone makes mistakes, the power of the community is to provide robustness against individual mistakes through peer review.
i laud their support of open source, but throwing money at problems is not enough if it comes at the expense of the community at large.
so now things are good, huzzah!
would they have been had everyone kept quiet?
after all, they had agreed to open up their repository to select individuals who wanted to work on it with them and then, because it was a private agreement, renigged on it. *shrug*
Hey, Aaron:
Funny that we ended up speaking about this face to face at Socal Linux Expo. I think I get your points pretty well. Thanks for your time on that.
I just finished interviewing David Reveman for Novell Open Audio. If you're interested, it should be online sometime late tomorrow.
I just stumbled upon ur blog while trying to find the song "Wouldn't it be nice?" by the beach boiz! cool thing u doin' abt openGL n stuff...well..open source is the way!
wouldn't it be nice if the management sees things...
then we wouldnt have to crack our minds...
and wouldnt it be nice without auditing?
in the kind of world where we belong?.....
You know its gonna make it that much better
When we can say goodnight and stay together
wouldnt it be nice?
oh no...the song just got grounded on my head!!!
black mold exposure,
black mold symptoms of exposure,
wrought iron garden gates,
your next iron garden gates, here,
hair styles for fine thin hair,
search hair styles for fine thin hair,
night vision binoculars,
buy, night vision binoculars,
lipitor reactions,
lipitor reactions,
luxury beach resort in the philippines,
beach resort in the philippines,
homeopathy for baby eczema.,
homeopathy for baby eczema.,
save big with great mineral makeup bargains,
companies marketing mineral makeups,
prodam iphone praha,
Apple prodam iphone praha,
iphone clone cect manual,
manual for iphone clone cect,
fero 52 binoculars night vision,
fero 52 night vision,
best night vision binoculars,
buy, best night vision binoculars,
computer programs to make photo albums,
computer programs, make photo albums,
craftmatic adjustable air bed
craftmatic adjustable air bed, info here
boyd night air bed
boyd night air bed, low price
air bed in wisconsin
best air beds in wisconsin
cloud air inflatable bed
best cloud air inflatable beds
portable sealy air bed
portables, sealy air beds
aluminum rv luggage racks
aluminum made, rv luggage racks
air bed form raised
best air beds form raised
support equipments aircraft
best support equipments for aircrafts
informercials bed air
best, informercials bed air
mattress sized air beds
best mattress sized air bed
antique doorknob identification
antique doorknob, identification tips
troubleshooting dvd player
troubleshooting with the dvd player
flat panel television lcd versus plasma
flat panel television, lcd versus plasma pic the best
causes of economic recession
what are the causes of economic recession
free printable tax forms
free printable IRS tax forms
bed air foam adjustable
bed air adjustable foam
hoof prints antique and unusual equestrian prints
hoof prints, antique equestrian prints
buy adjustable air bed
buy the best adjustable air beds
air bedscanadian stores
air beds, cheap canadian stores
Post a Comment