Submitted by Aaron Siego of KDE: “It would be very nice if 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. 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.”
Nothing more to say really. Disappointing…
To be honest if this is going to give us an OpenGL X server then I’m for it. I think it’s better to have it developed internally and then released(open source of course) than not being developed at all.
Regards
Agreed. Rumors say Novell is also pushing a Microsoft programming language, so no surprise here.
I wouldn’t go based solely on rumors. Novell is doing good things for the community, this wil turn out to be awesome when it’s complete.
This project would die becuase of lack of developers afaik so it’s good thing that somebody picked it up.
They cant build it first and get it working then release it to the community? Just becuase it isnt being developed in the open initially does not mean when it is released it will not be released to teh community.
Maybe Novell realized that to get it out the door they couldnt sit there wrangling with all the political crap in the OSS community and decided that they would start it themselves.
I like the OSS method but sometimes it takes forever to get things off the ground because you have to bow to every person. They can forgo this if they build it and make an intial release and then release it to the community.
Actually they waited years for Xgl being developed (by only 2 active developers)
Xgl is way too large project for 2 developers, meaningless to say that it didn’t even come close to usable/stabile X server
Exactly. Who cares about running alpha quality product. Let them bring it to market in at least a semi-workable state. The OSS community is good, but sometimes you don’t have the time for distractions from the work and just want to get something done. Why does everyone insist that every single line of code should be on a cvs. When it’s ready it’s ready. When it comes out decide if it’s good for you. All this protesting is ill.
I believe you are right as long as the final product is open source. I agree with you until we see the code kept proprietary. It is not bad to close your ears and do some serious work, sometimes. s far as the lack of developers for X is concerned, the recent open solaris move maybe a positive action. The only problem I see with X is lack of documentation and serious tutorials (IBM can help here). When is the last time you saw an X-Windows book on amazon? I think the ammendment to X for the transition from R5 to R6 plus some very high level X books is the only source. We need more Xlib and X internals books from
O Reilly. Especially those books needed to write device drivers. R7 seems to provide more help.
Keep in mind that the transition from Xfree to Xorg is recent. We need time and tutorials, even for vga graphics. EXA and XAA + Xgl are not that documented. And please refrain from aswering go read the source code. I could not have learned C with reading programs only. Cheers.
Open Source development is always about momentum.
If your initial idea is REALLY REALLY good then you can probably get developer interest with little or no effort or work on your part.
With a project like this, though , there isn’t that much community interest in the project to begin with. if there was it would have been done by now, or there would have been 3 or 4 GPL projects which do the same thing.
Look ! There are different ways to start an Open Source Project. Sometimes you want to make architectural decisions before releasing it to the organic, sometimes unpredictable process of Open Source Development.
Give them a break and see what they do. If they botch it up they botch it up, but as it is ,there are no other projects out there trying to fill that need open or closed , We either should just shutup and code our own or let them continue in peace.
It’s not that hard ladies and gentlemen. Quite the bickering
“If your initial idea is REALLY REALLY good then you can probably get developer interest with little or no effort or work on your part. ”
I find that to be somewhat to completely untrue. The only time Developers want to do something is if it is in their direct and best interest. Meaning they are scratching an itch or they are getting paid for the developement.
You could have the best idea on the planet, but if it does not coincide with what the developers are already doing it doesnt go anywhere.
Gnome developement is a good example of that. Where is the functional menu editor in gnome? The one they supply you with (2.12 branch) is so crippled to be worthless.
Xorg is another example, they had to fork to get where they wanted to go.
Consider this Novells fork.
So Novell wants to be the first with a “shiny bling bling Linux desktop TM”. So what, I don’t really blame them if they are putting developers behind it. Once they release it, they have to release the code and then the whole community will benefit. Sure it is a bit shady to quietly develop it without others helping, but if thats what it takes to have hardware accellerated x, I’m ALL FOR IT.
Sure it is a bit shady to quietly develop it without others helping…
How is that inherently shady?
This is the Real World (TM). If Novell put resources behind XGL while everyone else left it to languish, then they should get more in return than everyone else. The return on their investment should be a limited period of exclusivity as long as they open the code after that.
I do not understand this bias against businesses.
Of course they can do it. But that’s not how OSS works. Look at RedHat. They probably are the single company who made most investements into OSS, and yet what they develop is all publicly available and released under the GPL.
Too bad that all the SUSE management quit long time ago… Monkey-heads are now running the company… and it shows.
Let me rephrase that… Novell is touting themselves to have “embraced open source”. While they are doing EXCELLENT things for the Linux / OSS community, the best development model is full open source.
I’m not saying they should OSS all of their products, I’m saying that they should follow the open source development model when they work on open source.
As stated before, I think it’s great they have picked it up at all. It’s just that it would be developed faster if more people could work on their code.
Faster if open? Not necessarily. Others have noted that there is a lack of people with the necessary skills for this work. Maybe the team is best served by being able to focus on getting the work done, and not having the extra chores that go along with the more open process at this point…at least for the first cut. I’m sure it’ll be quicker, and having the first good such product of its type out there *is* what will get others to line up behind it. No doubt it will be fully opened as soon as a good base implementation is ready. Don’t worry so much.
I know at least one developer who said in public he would be interested in helping the XGL effort. It is the same one that brought us EXA: Zack Rusin.
Aaron Seigo spreating fud, is not the first time.
care to show why he’s spreading fud, Yuske?
Im sorry, I don’t answer to Anonymous.
well, then, answer me. aaron seigo has always been very sensible, unlike you (you’re on a -0.13 score now). so please, tell me, why is he a troll? (except from the fact he is employed by trolltech, now ;-))
This is not a big deal.
RedHat has done this many times with projects that have in the end benifited the whole linux community.
They often in the past have had a bunch of programmers go off on their own and develop something to an inital version, then release it to the community and start taking code submissions after that.
Heck, it might actually be a more efficent development model in some cases. Everyone knows you usually get more done with a small number of programmers hidden away than a huge number trying to all work together with lots of project management etc.
RedHat has done this many times with projects that have in the end benifited the whole linux community.
They often in the past have had a bunch of programmers go off on their own and develop something to an inital version, then release it to the community and start taking code submissions after that.
That’s been software that has been bought by them (RHDS for example), and it has never been part of a wider open source project like Xgl. They certainly haven’t been hosted on sites like Freedesktop which gives it a an apparent look of openness.
The article says “behind closed doors”. What exactly does that mean? Looks like there is CVS access to the source code so whats the problem? or is this a licensing issue? or is this just someone trolling + spreading FUD?
Exactly, looks more like some paranoia.
“The article says “behind closed doors”. What exactly does that mean?”
Take a guess…
“Looks like there is CVS access to the source code so whats the problem?”
Only that there isn’t. You can get six month old source from cvs, but not what they are currently working on.
Aparently, the code that is availabale on the public CVS is much older code than what Novell has in-house. They basically forked it and are keeping their fork secret.
Sounds to me like there wasn’t enough community interest to make a viable community based project anyway. I can only assume that if it wasn’t for Novell’s sponsorship it would be a dead project.
Presumably they’ve hired enough talent that they can live without community “input” for now. Especially if that community input is just complaints about speed, hardware support, and bugginess of pre-alpha code rather that actual tecnical contributions.
When you think about it, most open source “STARTS” as closed source.
Usually, lots of open source software starts out as someone working on something all by themselves, and then once it reaches a point where they feel its good enough(aka isn’t embarrisng anymore hehe), then they release it as opensource to others.
I REALLY don’t see the big deal, sounds like someone just does not like Novell for some unknown reason.
Sounds to me like a classic case of “wait and see”. At this point, everythng is speculation.
It really doesn’t make a difference. Before Novell took over, it was really just one guy who did all the work. Well, for 99% anyways.
The only difference is the testing, not everyone can test it now. But I’m sure that they’ll introduce it to the public when they’re almost done. They do the same with OpenSUSE, and other products.
“Only that there isn’t. You can get six month old source from cvs, but not what they are currently working on.”
How do we know that Novell are continuing the project? and that it didn’t just die 6 months ago?
> The article says “behind closed doors”. What exactly does that mean? Looks like there is CVS access to the source code so whats the problem?
Did you ever try that before posting here? Just because they claim on the website that there’s CVS access doesn’t mean that this is true.
This article is nonsense kde developers are angry because
Novell has rid off their pet into Suse.
“Did you ever try that before posting here? Just because they claim on the website that there’s CVS access doesn’t mean that this is true.”
It’s not Novell’s website. Maybe I’m just not paranoid enough to understand…
“Sounds to me like a classic case of “wait and see”. At this point, everythng is speculation.”
I think I agreed unless someone else knows different. The quality of journalism on this site (and other sites like it. e.g. slashdot) is fairly poor. It’s more of a rumour mill or tabloid paper than a serious news site.
There isn’t only glitz as accelerated technology, to build XGL on top of.
Take a look at http://www.amanith.org
Talking to authors they told me that in the next version a compositing layer will be available, and that the project will remain opensource forever.
Interesting,perhaps Novell is willing to include Xgl in the coming 10.1 retail.
you can try Xgl in opensuse already, but I would not know if this is the Xgl developed by novell
There is no proof that Novell is doing this. This blog post is based on Dave Airlie’s original post, it is not new information. I have talked with Dave and he does not have factual data to back up his suspicions.
http://www.livejournal.com/users/airlied/18241.html
It is unknown if Novell has officially sanctioned this or if it is being done as an off-the-books type project. It is also unknown if the project has sufficient resources to reach completion. As far as the facts go this could just be Dave Reveman working in his spare time and not feeling like committing his code.
This is just a conspiracy theory without facts to back it up.
If Novell is doing this, I welcome it because without their backing it would never get written.
Edited 2005-12-20 15:02
point (and reason of aarons critical stance) is that they should develop it in the open – the development would go faster as more ppl would send in bugreports and probably even patches, and the upcoming users (KDE, Gnome for example) could test it and give comments. also, the hardware driver developers (in kernel, or nvidia and ati) could also adapt their drivers and probably help.
so it sucks they keep it in-house, and imho it is a proof of the fact they still don’t really understand “free software”.
that is such BS .. you don’t go live in a house before its built do you? no one wants to work on this project don’t you get it? its been around forever. Just because you are curious and want to see the code being developed has nothing to do with openness. just let novell do their thing. personally i doubt this project will see the light of day for years.
Yes they should have more interaction with the community on this I think – but of course if they keep the develpment to themselves then they can say where the project will go without any potential interference from “outsiders” .
As long as they come up with something which furfills expectations & needs & in the end all the stuff is made available to the community properly then I dont mind.
If I was a X developer dealing with GL then it would propably be different.
Novell are working on XGL – I asked on the OpenSuSE mailinglist a while ago – & got a link to some site .. i.e. they are working on this.
& yes they want to include this in a futire version off SuSE – well there was this interview years ago in which a Novell person stated that round the 10 realease that Linux would get a desktop that would better & fancier than OSX & Vista.
Go Novell/SuSE
There’s an old post from David Reveman, the Glitz and Xgl guy, in the xorg mailing list:
“It’s more likely that I’ll move my new Xgl code into Xorg after the modularization is complete and 7.0 is released than trying to merge it into the kdrive tree.”
[http://lists.freedesktop.org/archives/xorg/2005-October/010569.html]
Wise decision I think.
David also made this post to the cairo mailinglist more recently:
“Most of the functionality I want is in CVS now and the window system independent code and GLX backend code seams pretty stable. Don’t know about the other backends, though. They probably need some work. Hopefully we can release a 1.0 in a few months.”
[http://lists.freedesktop.org/archives/cairo/2005-December/005793.ht…]
/C.M
What about Luminosity. Doesn’t this do the same thing as XGL?
Stolen from Seth Nickels explanation:
“…Xgl is to X11 as Glitz is to Cairo: it provides the same APIs rendered in a much smarter way. Luminocity, on the other hand, is a compositing manager / window manager fusion that composites using OpenGL. Compositing and Window managing are all about what you do with client-rendered windows. Luminocity doesn’t know what’s inside windows, and it doesn’t care.”
[http://www.gnome.org/~seth/blog/relations.html]
/C.M
I just acquired two pieces of informations:
1. XGL is being developed;
2. it is being temporarily developed under closed doors;
To me, the real news is the first one.
I hope they will succeed.
Well, Aaron Siego, can you explain why the new Oxygen icons are developed behind closed doors then? Those are KDE-artists/developers and not even a company is involved.
Xgl was dying because of the lack of developers. Only one man was working on it, and didn’t have any time to develop a native XServer which would work completely.
Having a company building software isn’t a bad thing at all. Please, let business decide what to do. Pushing Open Source to the extreme isn’t good for anyone! (L)GPL is about releasing code, not purely about community-efford during the development process.
The outcome is that we users can benefit from Xgl instead of it not being released at all. And I’m pretty sure that Novell isn’t going to release the code after KDE 4.0 is released, but long before it. So, do hold your breath, because Xgl will be just Ok.
Don’t let Open Source blind you from what users need and what reality is.
Well, Aaron Siego, can you explain why the new Oxygen icons are developed behind closed doors then?
Well:
Those are KDE-artists/developers and not even a company is involved.
That’s why there’s no problem. There’s also a page (recently edited) and a site that tells you what the Oxygen icon people are actually doing along with samples. Hardly the same thing.
Please, let business decide what to do.
No, because it’s a community project (or it’s supposed to be) hosted on Freedesktop that potentially affects a lot of people and projects. Insert business for Microsoft there and you have exactly the situation the open source community is supposed to solve.
I don’t know how to take this. It does seem a little off. Yes it is better than nothing I guess but it doesn’t seem like good OSS ediquete either.
It makes me wonder about their motives. Open is good because it allows you to have oversight and understanding. If they are keeping it closed my real questions isn’t about the lack of being able to test it (that would be nice) but about their motives. Why why why? Makes me suspicious.
You put that next to:
a) their track record as a big nasty evil company
b) their use of Mono which is known to have some very dangerous licensing associated with it
and you get a company that I am really starting to distrust as a linux/gnome user.
PS – anyone who feels like arguing the Mono thing (aka ass holes who scream “FUD!!!” as soon as someone says something they don’t like. are they being paid by MS?) just google around for old posts of mine or Seth Nickels very well written blog on this subject.
What are you talking about?
“a) their track record as a big nasty evil company”
I don’t know what this is about. I’ve not had that image of Novell. Microsoft, yes. SCO…well, little nasty evil company, not big. Oracle, in some ways. IBM, much improved, but not perfect. Novell? They’ve become a great example of what we *would* like to see from a large player in the industry. Not sure what your beef is.
“b) their use of Mono which is known to have some very dangerous licensing associated with it”
You don’t have to use Mono if you don’t want to. Personally, I won’t ever use anything associated with .Net. At the same time, you’re probably one of those who whine against Sun for not fully open-sourcing Java, even though there are OSS JVMs available. I don’t know what the Mono licensing is, not having looked at it, but I’d suspect it is the way it is because of legal necessity – that’s what happens when you reverse-engineer someone’s virtual machine and framework and produce a port of it.
So…yes, I guess my summary of your comments is: FUD. Bring real-world complaints or else stay in your playpen.
“a) their track record as a big nasty evil company”
“b) their use of Mono which is known to have some very dangerous licensing associated with it”
????? Who are you talking to? I can’t find anithing like this in aseigo’s blog…
Those are quotes from the parent post.
This is not true. There is no “evil” in the article or aseigo posts. The point of using “quotes” is to use the same words as someone else.
There. Now you know.
It was from fooo’s post on page three of this discussion:
http://www.osnews.com/permalink.php?news_id=13040&comment_id=76053
Now you know. 😉
Now you know. 😉
Thanks. He could have used the “Reply” then. This way I wouldn’t think he was talking about aseigo’s post.
a bunch of Gnome bashers. They are worried that it may not work with Qt.
first off, i happen to know that this is happening for a fact due to talking with people involved there at novell. this isn’t speculation on my part, nor am i (by far) the only one aware of the situation.
i *am* very happy that novell is putting people on the project. however, that is completely unrelated to closing out *other* people from participating. the former has nothing to do with the latter; and there have been requests made to get involved with XGL, including code contributed, and the response has been negative.
the cvs on freedesktop.org is (i just checked again, actually) around a year old.
why is this bad?
because it completely does an end run around the peer review process; it ensures that nobody else can work with what could be a fairly important technology while it is in development (and that’s purposeful); and it ensures that other vendors are *not* supporting xgl and going off and doing their own thing which means we get further splintering of the community and direction around X. X has enough problems without creating schisms around it.
there is absolutely no good technical reason for what is happening, it’s purely a “business decision”.
i understand that not everyone here likes hearing that, and perhaps especially from me =), but thems the facts and unlike others who may be content to sit quietly wringing their hands, i’m more apt to do and say something about bad situations within the community in which i work.
“there is absolutely no good technical reason for what is happening, it’s purely a “business decision”. ”
And there’s nothing wrong with that. They can choose the development model that’s best for them, as the Oxygen artists did. What matters is that the code, when released, is under a free license.
If others are not happy with the situation they can either convince Novell to work with them or they can start their own code. Crying about it won’t get anything done.
It would be a different issue if they were releasing software under a non-free license; that would be a serious problem. That’s not the case here.
Maybe you should consider that possibility. One development process does not work equally well for every project. Sometimes the work at hand demands a different approach from that taken on another project. Does it not seem likely that some permutations of the OSS development model would be needed to best handle the wide variety of technical, economic, political, etc., conditions and challenges facing specific projects? Every process does *not* need to follow an exact cookie-cutter OSS process in order to be a “good” OSS project.
Assume the group working on it has a strong and clear vision of what they want/need to get done for a first cut. Maybe they’ll be better able to focus on cranking it out if they don’t have to spend time vetting other peoples’ submissions and ideas during this phase. And a quicker first release of something implemented and working will get others to jump on the wagon, since they’ll be cut off earlier in their competing process.
Sometimes the way to get your sw adopted ubiquitously, and closing the team off can at times help create the focus and momentum needed. They can (and likely will) then open the source and the development process once they’ve gotten a good code cut done. You’ll still get your “peer review”. After all the OSS work that Novell/Ximian/SuSE has done, don’t be so naive as to suggest they don’t know the benefits of that.
it ensures that nobody else can work with what could be a fairly important technology while it is in development
People are always told that’s one of the great things about open source : anyone can take the code and start hacking, now suddenly that’s a bad thing ?
it ensures that other vendors are *not* supporting xgl and going off and doing their own thing which means we get further splintering of the community and direction around X
Another piece of open source doctrine : open source is freedom of choice which is good, the good technology succeeds and competes. Now suddenly you’re worried about fragmentation, how about resolving the fragmentation in the layers above the X server first ?
Besides which fragmentation, this thing implements X11, so it should be compatible with other implementations right ?
there is absolutely no good technical reason for what is happening, it’s purely a “business decision”
Let’s see assembling a small dedicated team of paid professionals to work on a piece of software they would like to see released is not a good technical reason. This would be because we all know that the bazar model is superior in all possible circumstances and configurations to any other development model. Um yeah.
And if the fork stays closed well that’s the option the original author offered them (when choosing the license) and that was his choice to make, even the GPL allow you to keep your code to yourself as long you keep the product in house.
Sorry, but this whole thing just smacks of GPL license fanaticism to me (all code must be Free and all that)
“People are always told that’s one of the great things about open source : anyone can take the code and start hacking, now suddenly that’s a bad thing ?”
Jesus, learn to read.
The bad thing is that the code is _not_ publicly available, so nobody from the outside can take, nobody from the outside can test it and give feedback or adjust other software to work well with XGL, nobody from the outside can contribute, so some would like to.
The bad thing is that the code is _not_ publicly available, so nobody from the outside can take, nobody from the outside can test it and give feedback or adjust other software to work well with XGL, nobody from the outside can contribute, so some would like to.
Read what I said : anyone can take the code, maybe there are 100 other devs out there hacking on their own forks in private. The code is there they can take it, this is supposed to be good.
But no, someone might be actually doing some work of value and the first response is not “cool, someone took our code and it helped them” (which would be more in tune with the purported values of the open source community), but “Gimme, gimme gimmeeee !” instead.
Here’s a clue : there are actually licenses that require changes to be sent back to the original author, it’s licenses like these that open source zealots like to denounce companies like Apple and Sun over.
I just whish they would make up their minds already, you can’t have your cake and eat it too.
_there are actually licenses that require changes to be sent back to the original author, it’s licenses like these that open source zealots like to denounce companies like Apple and Sun over._
Licenses really have nothing to do with this situation. The GPL’s requirements only kick in when the software is distributed and Novell’s XGL work hasn’t been distributed at all (which is the source of Aaron’s complaint).
The bad thing is that the code is _not_ publicly available, so nobody from the outside can take, nobody from the outside can test it and give feedback or adjust other software to work well with XGL, nobody from the outside can contribute, so some would like to.
I believe he was talking about the code that’s still in the public repository, not the code from the Novell fork. His point is that Novell was only able to fork the code because it was under an “open source” license in the first place. And no, I didn’t understand him the first two times I read it, either. 😉
I believe he was talking about the code that’s still in the public repository, not the code from the Novell fork. His point is that Novell was only able to fork the code because it was under an “open source” license in the first place. And no, I didn’t understand him the first two times I read it, either. 😉
Thank you that’s exactly what I meant. I blame English not being my mother tongue rather than just sloppy writing *coughs*.
“Sorry, but this whole thing just smacks of GPL license fanaticism to me (all code must be Free and all that)”
Actually, complaining about the development model used by a company or individual is not something that I would relate to the Free Software Movement. Free software is about freedoms of the user of released code, and says nothing about the development model; the development model is an idea of people who defend the “Open Source” definition.
I don’t know if you’re giving enough weight to the political friction surrounding XGL. “In the open”, there was massive resistance to getting anything done with it. From the very first step, convincing people that 2D was doable via OpenGL at all, it was an uphill battle. If Novell decided to sod it all, and go and just write it and see what happens, I don’t think you can blame them.
for all you people who glibly suggest that this is a case of a “kde guy who dislikes $ENTITY”, let me assure you that i have a bit more adult blood in my veins than that. you can ignore realities if you wish, but i prefer not to.
and if you read my blog you may notice that i both praise and decry as i see things happen. both within kde (yep, i’m critical of my own project when i see issues there) and in the technologies in which we rely and use.
dude you are full of it and your post was nothing but trolling.
One of posters before made interesting argument.
care to explain about oxygen. Why is this closed door project? Why is this cool and Novell XGL not so cool?
care to explain about oxygen. Why is this closed door project? Why is this cool and Novell XGL not so cool?
Agreed. I would prefer that Oxygen would be developed on the open as well.
But seriously, AFAIK, nobody is interested in Oxygen outside KDE. And there are a lot of people who are interested in XGL, and could help there. So I don’t think it is the same thing…
BTW, what do you think that the fact that Oxygen is being developed at closed doors justifies novell doing the same? AFAIK, tango was developed at closed doors too, so when they announced it, it already had the guidelines and most of the main icons. So it wasn’t a open process either.
Agreed. I would prefer that Oxygen would be developed on the open as well.
/*If I remember correctly*/ When I was reading Tango mailing lists they said Oxygen and Tango will cooperate on icon naming, etc. Meaning common spec that all should use.
So, yes in that case it is interesting even for non-KDE users.
/*If I remember correctly*/ When I was reading Tango mailing lists they said Oxygen and Tango will cooperate on icon naming, etc. Meaning common spec that all should use.
So, yes in that case it is interesting even for non-KDE users.
They already do. In a public cvs server (freedesktop.org)
See here: http://standards.freedesktop.org/icon-theme-spec/icon-theme-spec-la…
And here: http://cvs.freedesktop.org/*checkout*/icon-theme/default-icon-theme…
The tango website points to the same locations, see http://tango-project.org/Standard_Icon_Naming_Specification
Let me say that I can only praise the tango guys for working on that, together with the kde hackers. They are already colaborating in an open way.
Edited 2005-12-20 22:49
Hey, guys, let aseigo put a bit of pressure on Novell, for God’s sake. Yes, it is nice that they are people working on it. But, yes, it would be better if it was done in the cvs repository, like other people do.
Imagine if everybody would work on X on closed doors, and release it only together with their product. Imagine the hell of merging these different versions.
Now you understand why it is better if they develop it in the open? If they let other people participate?
Let Novell know that people notice when they don’t work openly. It will maybe change their development to the ideal model: open development.
Hey, guys, let aseigo put a bit of pressure on Novell, for God’s sake. Yes, it is nice that they are people working on it. But, yes, it would be better if it was done in the cvs repository, like other people do.
Yeah let’s all put lots of pressure on these companies until they drop out one by one. Hey that will really encourage other companies to open source their code or even use it. Talk about biting the hand that feeds you.
…that the case against Novell is, for now, indiscernable. I suggest that a motion for dismissal with leave to re-present at a later time is in order.
Even if turns out that Novell is clearly not acting in the best interest of OSS, OSS will still survive.
Edited 2005-12-20 18:21
Lets see…reminds me of another app…that eneded up getting released as Open Source after it actually worked and was in use…
Ah yes…Star Office developed by Sun behind closed doors, and was released to become Open Office…
My point is to get a project done in a potentially quick way, this is the best. They will have dedicated teams for this, instead of just a couple weekend developers who may or may not have time to put to the project. I am in NO WAY saying weekend developers for these are a bad thing, just that they most likely are professional developers as well, and have projects to complete which they are getting paid for. That would take precedence over a free project.
My point is to get a project done in a potentially quick way, this is the best. They will have dedicated teams for this, instead of just a couple weekend developers who may or may not have time to put to the project.
Why can’t them employ people *and* use the public repository? A lot of people do that.
OK, it a lot better than nothing. But now Novell has lost the right to complain about similar moves from other players.
What if Red Hat (who employs most of the GTK hackers anyway) decides to develop it in house and release it only when they release their product? Imagine the mess…
The StarOffice release was a great event for the community, but it was the ugliest piece of Unix software I’d ever seen until the OpenOffice folks got a hold of it and spent a few years eliminating the faux Start menu etc.
StarOffice is a good example of how closed development can be badly misaligned with the values and aesthetics of the community.
Let me sum up things as I see them:
1) Novell gains lots of Kudos points for working on it, and releasing it as open source (if they do).
2) Novell loses lots of Kudos points for not doing it in an open fashion.
3) The net is positive for Novell. But could be a lot better. They are setting a bad example.
A very bad example actually. I believe Siego has some point. how can other projects sync to the changes of XGL to have final library/program/enhancements along with XGL?
I mean when XGL is alpha they can start modifying libraries to use this architecture. I love windowmaker and it would be good if it benefited by the project. If the devel is not in public how can they adapt wmaker to xgl? This closedness can do only harm to the community. Maybe it is time to revive the project. we cannot make deals with the devil (companies).
“Why can’t them employ people *and* use the public repository? A lot of people do that.”
True, they could do that. But has been mentioned this project did not have people wanting to work on it, and it was being resisted due to politics. So in order to get it done it looks like Novell just took the bull by the horns. Providing they release the code there is no harm done, and a stagnant project now has life IMHO.
“The StarOffice release was a great event for the community, but it was the ugliest piece of Unix software I’d ever seen until the OpenOffice folks got a hold of it and spent a few years eliminating the faux Start menu etc.
StarOffice is a good example of how closed development can be badly misaligned with the values and aesthetics of the community.”
It is not the aesthetics I am discussing, but rather the functionality. I hated that Start menu myself with a passion. My point is the main functionality was already written, making it a shorter time span then writing the code from scratch. Another project that comes to mind is Mozilla, which was originally Netscape , until Netscape released it. The difference here is the project started in the wild and Novell has put it in a cage. Who knows, maybe they will open it up for public contribution once they have a handle on it. Time will tell on that one.
Again, this is all just IMHO.
Novell is following Apple MacOS X (OpenGL based) and MS Windows Vista (DirectX based) operating systems.
Maybe Novell is trying to make a closed OpenGL-based GUI on top of linux.
The irony of this is very funny. Isn’t Aaron Siego one of the leaders of the pack that constantly claim that the fact that Qt is developed behind closed doors is good for KDE? He’s blasting Novell for doing exactly what Trolltech does. Does anyone actually pay serious attention to this guy’s rants anymore? Why does he have any credibility? I’m not talking about his development work, which I’m sure is stellar. I’m talking about his half-informed and poorly thought-out blog rants. In that respect the guy is clearly a crank and a troll, the kind who fools himself into believing that he’s doing the world a favor by poking sticks in everybody’s eyes. Obviously the KDE guys aren’t very happy with Novell for other reasons these days, but there’s just no excuse for this kind of half-witted scare mongering.
Look, I’m not big fan of closed development methodologies. I say, if your software is going to be free and open source software anyway, why not let the sunshine in? But the reality is that it doesn’t always work that way. Sometimes we get “code drops” from on high, especially when a project is brand new. Companies including Trolltech and Novell have provided justifications for the practice in the past, which you are free to agree with or dismiss. The practice is certainly not ideal, in my opinion! But it’s also certainly not “non-free” in any way, shape, or form, despite Siego’s insinuations.
Free and open source code is free and open source code, whether you like how it was developed or not. If the development methodology of a project becomes a continuing problem once the code is out there, the community always has the option of forking (just as it did for different reasons with X.org). The free and open source label doesn’t mean that the software is developed in a world of candy canes and unicorns designed just to your liking. It just means that the license is free and open.
“there is absolutely no good technical reason for what is happening, it’s purely a “business decision”. ”
“And there’s nothing wrong with that. They can choose the development model that’s best for them, as the Oxygen artists did. What matters is that the code, when released, is under a free license.”
And yet, an icon set is entirely unimportant compared to something affecting X at such a basic level.
Icons also don’t really need oversight (though some people might want to comment on the WIP). A massive chunk of code on the other hand needs oversight. When they get done and drop a code bomb on X.org, there’s no telling if there will be something about it that makes it unusable. X.org might have changed some internals, they might not like the way something is done (extreme example: if Mono is used).
If there is something about how it is done that prevents it from being integrated into X, it will be much harder to fix once everything is done. This is the sort of thing that needs feedback. It needs more eyes. People miss stuff.
Of course it’s great Novell is working on it. I too thought it had died
Now though we are left hoping their final product does everything right, otherwise it will be some time after the release before it’s usable for anyone but Suse/Novell. Great for them of course 
Aaron Siego is right on this one. While Novel is not breaking any laws by doing the development behind closed doors, XGL is to essential to Linux for them to do it that way. The Xserver is a fundamental component to Linux and by not allowing other companies and developers to contribute to the code base as it is being actively developed, they are untimately going to cause the code base to be forked several times. Trolltech has already stated that they will develope or aid in the development of a GL based server. Had Novel publically announced that they were going to start full fleged development on this project and officially invited other participants, this project could have been a major unifiying project for the open source community. The fact is, not very many developers would have shown up anyway, had they publically announced their intent, becuase this is a very difficult project, but a few very dedicated and talented companies and developers could be contributing to XGL’s development right now. Zach Russian and Aaron have talked repeatedly about XGL for KDE 4 on a number of occations in addition to EXA. If Novel wasn’t trying to control the implementation, We could have Trolltech and a number of the KDE developers, Redhat (rememeber, this is where the XGL code originated anyway) and Novel all hacking away at an XGL server that everyone is going to need in a couple of years.
While I know everyone is just happy to see someone working on this project, they’ve really set the whole project back, because as soon as they publish the code (if they publish the code) we’re going to see multiple forks so that the server can be customized (How is that really helping). They have every right to do the development behind closed doors, but that basically defeats the concept of open. Yes, some people think open means just releasing the code at some point, but those individuals are not fully embracing the true spirit of what open source development means (open to everyone during the process, not just bolt on parts (rewritten code) after I’ve played God with the code)
Here are the most recent screenshots of xgl that I could find. Keep a rag handy to clean up the drool if you’re going to take a look at them.
http://www.cs.umu.se/~c99drn/xgl/
http://www.cs.umu.se/~c99drn/xgl/xgl-shot1.png
http://www.cs.umu.se/~c99drn/xgl/xgl-shot2.png
http://www.cs.umu.se/~c99drn/xgl/xgl-shot3.png
Aaron Segio has, once again, showed the lack of maturity in certain parts of the KDE community. And this time it’s someone who represents Trolltech as a developer. In order for RedHat and Novell and others to compete with Microsoft, they need to be able to deal with people that have a certain level of maturity and the KDE community, once again, demostrates they are lacking there.
Aaron Segio has, once again, showed the lack of maturity in certain parts of the KDE community.
And you showed you can’t handle criticism, even reasonable one. Can’t you discuss stuff without calling names?
Here is a tip for you: focus on the issues. There is nothing objective on your post.
Let me start by saying that, if Novell releases the code later, I think (and aseigo too, as he said) that having people working on XGL is good. That said, it could be better:
Isn’t Aaron Siego one of the leaders of the pack that constantly claim that the fact that Qt is developed behind closed doors is good for KDE? He’s blasting Novell for doing exactly what Trolltech does.
Qt did not exist before trolltech started to develop it. XGL was already open source. Trolltech was generous for giving it’s code for open source developers. Novell is taking public code and taking it to closed doors development, so, until they release the code, they are going the other way, the wrong way.
Qt is not developed in a public CVS in open fashion, because otherwise, Trolltech could not sell licenses. XGL has none of those restrictions. So the question is:
Why is Novell developing it this way? The public CVS was already there…
Can you spot the difference?
BTW, if you want to talk about credibility, what about those ximian guys who considered themselves the heros of free software now giving this bad example. Tsk, Tsk, Tsk. Bad, my friend, bad.
looks great, hope they release some working code, sooon.
And more hope that the fellow debian/ubuntu supervisors accept it and bring it into the distribution and dont wait 5 years until its branded stable or whatever
that xgl won’t end up on the ForbiddenItems http://fedoraproject.org/wiki/ForbiddenItems list of the Fedora project… The GPL zealots should come to their sences. And from what i see form the comments here this is the oppinion of the general public.
XGL is not GPL licensed, it is MIT licensed. You can drop all of the GPL comments; they don’t apply. MIT licensed means Novell can ship and keep the source secret forever if they want.
The license is MIT instead GPL so that X and variations can be shipped with BSD if anyone bothers to port the code.
i am not talking about that particular license. i am talking about that attitude in some projects, that if something is not pure open source or developed “the open source way” it should not exist.
I am talking about the binary drivers issues, the mono issues etc…
Well because the software can’t be distributed freely like EX: ubuntu discs, without agreeing to the EULA of the closed software (EX: see adobe reader license) , the opensource licenses does give you that permission.
The vendor themself must actually give you permission to include it.
And let’s not mention the patented software.
*shrug*(EX: just google microsoft patent lawsuit)
Go read the MIT license. MIT is more free than the GPL. It basically says you can do anything you want to the code — print it on toliet paper or keep it secret. There is no viral requirement to share your source.
If you can legally get your hands on MIT licensed code you can redistribute it all you want without asking permission of the original author. The key part is that Novell has to agree to release the code in the first place, you can’t force it out of them like the GPL lets you do.
Ubuntu aready distrbutes a lot of MIT licensed code since the whole current X server family is MIT licensed.
why would an opensource project end up on that list?
oh right, you are simply an idiot trying to start a flamewar.
that would be ace, go Novell.
but it would be nice if they released the latest version as an install option in 10.1 and dropped the code in CVS at the same time.
I hope this pushes David and Novell to put newer code on the CVS. Because it will help development?
No. Because I am crazy about eye candy and I want to compile and use the newest version on my desktop!
So- gripe on KDE leader! Bring me a CVS on which I can break my box! I need something to tide me till KDE 4 is released!
Perhaps I’m uninformed respect to this, but aren’t major QT releases done the same way more or less?
I mean, there are articles about technology previews of QT 4 ( http://dot.kde.org/1089303565/ ) that seem to indicate no previous source code was released until the work was usable enough.
Trolltech releases daily snapshots:
ftp://ftp.trolltech.com/qt/snapshots/
I decided to remove this post….
Edited 2005-12-20 22:36
Perhaps I’m uninformed respect to this, but aren’t major QT releases done the same way more or less?
If you want to compare a toolkit vendor, which sells commercial licenses, and therefore has to develop everything in-house, with a linux distributor that takes an already public project an bring it in house, fine…
But imagine if every linux distributor started to do the same: everybody would work on their own X or GTK or whatever fork, and sync it only during releases. That would be a nightmare.
AFAIK, Trolltech develops Qt with their own resources, and they plan to continue doing so, and therefore a public cvs makes little sense.
To sum up: on one side you have a company that takes their own code and makes it open source. On the other side, you have a linux distributor, that takes other people’s code (even if the license permits) and develop it at closed doors. (There is manifested interest in the community of joint developement, there is at least one high profile developer who said he wanted to work on it.)
So if you can’t see the difference, I can’t help you.
If you want to compare a toolkit vendor, which sells commercial licenses, and therefore has to develop everything in-house, with a linux distributor that takes an already public project an bring it in house, fine…
But imagine if every linux distributor started to do the same: everybody would work on their own X or GTK or whatever fork, and sync it only during releases. That would be a nightmare.
XGL never had a stable release, and the development was mostly stagnant since there were few contributors. By the way, Novell competes with Redhat for enterprise licenses, but that doesn’t mean they must do all development in-house. Seems you think with toolkits it can’t be the same. Also, Novell hasn’t stated any plans to always keep the development in-house.
Why are you people making such pessimistic predictions? Last I heard, Novell open-sourced things like YaST and the Exchange connector when they had no need to. I’m almost sure that when they make the publicity stunt they want (“next-generation graphics in novell linux” or whatever) they’ll sync the sources in public cvs and that’s it (no “only sync during releases”). And keep paying a full-time developer for it.
So if you can’t see the difference, I can’t help you.
I see the (small) difference: a longer wait to get the first public version. Don’t trouble yourself “helping me”. I just don’t get why some enterprises keep the development in house, will always do so and it’s OK, and when Novell takes a project on the verge of being abandoned and decides to release the changes later lots of people pop around criticizing the incredible evilness of the wretched company.
Only thing I’ve got clear here is this: people are lusting for an XGL release and can’t wait.
I see the (small) difference: a longer wait to get the first public version.
I see another: people who want to help can’t.
Only thing I’ve got clear here is this: people are lusting for an XGL release and can’t wait.
I agree.
I see another: people who want to help can’t.
That’s the funny thing. Now that this piece of news appears, looks like everyone wanted to help. Compare with the situation in August:
http://lists.freedesktop.org/pipermail/xorg/2005-August/009168.html
“The basic problem is that only two people (myself and DaveR) are
working on Xgl. Xgl is way too large for two people to finish. I have
been trying without success for the last year to attract more people
to work on Xgl.”
How about this one then:
http://lists.freedesktop.org/archives/xorg/2005-August/009655.html
I am sure that this hacker would be helpful…
Edited 2005-12-20 23:18
How about this one then:
http://lists.freedesktop.org/archives/xorg/2005-August/009655.html
I am sure that this hacker would be helpful…
Zach isn’t working for Trolltech anymore. He has gone back to pursue an advanced physics degree. I doubt if we will see any software out of him for a long time.
I am sure that this hacker would be helpful…
Oh, I don’t doubt it. Funny thing is he’s being working for Trolltech on QT’s Arthur rendering system until around october-november, and now he’s on physics research. Here, his blog:
http://www.kdedevelopers.org/blog/14
Novell: 1 full-time developer. Rest of the world: good intentions and words of praise.
Novell is taking up the development of XGL behing is door? Stop screaming like it was selling it already!
First Novell is surely doing this because it want to work without being disrupted by people asking for dumb fonctionality or to response to decision not to implement some feature.
Novell does this to develop faster and better. Dont forget that they have the ximian guy with them which have strong knowledge of Gnome. And don’t rave about KDE/QT, it _WILL_ work too. Novell guys are not dumb. They have been around much more longer than a LOT of other software company. Pleasing a lot of people will get the some favoritism from the community.
At the same time, Novell is doing this as it was said to compete with the Vista graphic system. It is for money, big money. Don’t expect Novell to mess with that.
And finaly, why Novell won’t give it back? Open source is supposed to be their moto now. Doing so would simply mean they would be so much hatred that they would surely die from an horrible death, the community sending them a big FOAD.
First Novell is surely doing this because it want to work without being disrupted by people asking for dumb fonctionality or to response to decision not to implement some feature.
Well let’s all throw the towel in and can the concept of open source software. Microsoft has it right then.
Novell does this to develop faster and better.
Bringing open source projects in-house to hack on is not economically viable.
At the same time, Novell is doing this as it was said to compete with the Vista graphic system.
And they’re going to compete with Vista by bringing a very raw project in-house to hack on with two or three developers? Right………
It is for money, big money.
Oh my God, when will this ever end? The desktop, and any 3D effects on it, accounts for zilch, zero, nada, sweet FA of Novell’s revenue. They are merely dorking around with code trying to create their their own corporate desktop in their own little world as they have always done.
The reason why it’s important is that it needs to be developed to give Linux desktops a future in their current state and a fighting chance for the future. That can only be achieved in a community, which is what Freedesktop was set up for. Also, X is incredibly important to anyone who runs a graphical environment.
The notion that Novell is going to create the equivalent of the Vista graphics system doing this is going to have the Microsoft fanboys rolling in the aisles.
> Isn’t Aaron Siego one of the leaders of the pack
> that constantly claim that the fact that Qt is
> developed behind closed doors is good for KDE?
care to point to a place where i’ve said this, ever?
> He’s blasting Novell for doing exactly what
> Trolltech does.
as others have pointed out, it’s a slightly different set of dynamics here.
but even if it were, this isn’t how Trolltech develops either. they offer snapshots that are updated regularly. you can contribute patches, and we (KDE) often do.
> Does anyone actually pay serious attention to this
> guy’s rants anymore? Why does he have any
> credibility?
perhaps because once in a while i’m right?
> I’m talking about his half-informed and poorly
> thought-out blog rants. In that respect the guy is
> clearly a crank and a troll,
if you actually read my blog, as opposed to just entries that get picked out for publication on a website every so many months, you might notice that i’m usually pretty upbeat and easy with the support and praise.
the “shocker” is that osnews.com (among other sites, it’s not unique to here) like to publish stories about controversy more than they do about good things.
> the kind who fools himself into believing that he’s
> doing the world a favor by poking sticks in
> everybody’s eyes.
it may surprise you, but i hate writing this kind of stuff. fortunately i don’t feel like i need to very often. but once in a while something happens and rather than stand by silently while trains derail i step up and say something.
yes, i know that that is an unpopular thing to do. i learned that back in elementary school, but i also learned that sometimes unfun, unpopular things are necessary or at the very least beneficial.
and again, i pat people on the back waaaaay more often than i poke with sticks. fortunately there’s a lot more going on that’s positive than not =)
i would urge that you consider the message rather than the messenger, because even if i am a warty, ugly, horrible ogre of a human it doesn’t change what’s happening …..
I think that if this story reaches slashdot, Novell will open the development to respond to the community requests. XGL is a too big story to be developed behind bars.
As long as Novell doesn’t kill XGL like they did to WordPerfect…
Hmmm, here’s a clip from Novell’s portion of the OSDL Desktop Architects meeting this month (link: http://kegel.com/osdl/da05.html#novell ) :
Challenges: OpenOffice is the key! We have 6000 internal users. It’s viewed as a cheap clone of MS Office by users. UI is klunky and bad. Looks worse than MS Office. MS Office users have trouble using it. It would be nice if it were faster, good UI, and had some compelling original functionality. We’re really dependent on Sun’s efforts here. It’d be nice if there was more multilateral development on ooo.
Q. What’s stopping Novell from putting more engineers on OOo?
A. We have a few already. It’s money, really. If we could have a revenue strea (sp.) from it that’d help.
So OOo2 is identified as one of their biggest challenges for the desktop due to it’s design, and they’re relying on Sun to improve it because they don’t have the resources to justify heavy development on their end.
Fair enough.
But seems to me that investing in improving OOo2 would do far more to help them advance linux as a business desktop than creating an OpenGL based desktop environment. I have to assume they’ve got more than one or two guys working it if it’s really going to fly without external testing and development, even in pre-Alpha stage.
Novell has claimed on more than one occassion that they are going to blow away Microsoft on the desktop in terms of wow factor, a wholly misguided strategy in my opinion. Hype sells home users, it doesn’t sell business users (once they manage to cut through it, that is).
I’d like to think they’re keeping it closed for now to help speed development before re-opening it as a public project, but I’m cynically inclined to believe that they’re keeping it under wraps because they want to showcase it for NLD and be the first out of the gate with it. Then again, NLD 10 is supposed to hit in the first half of next year, I can’t see them having it ready by then. An ultra-cynic might think they intend to close it and keep it as a proprietary advantage, but as was pointed out, that would poison them in the OSS community and rightly so.
Still, when you listen to Messman speak about Windows you get the impression they’re focussed on replacing Windows at all costs rather than simply providing an alternative to Windows. It might seem like a fine line, but with one you’re setting the bar unattainably high and with the other you can focus on specific opportunities that exploit your strength(s). I hope it’s just hyperbole (“Vista will drive users to linux!”) and that they haven’t actually drank that koolaid, otherwise they’re doomed.
If they’re serious about opensuse being an inccubator project for corporate technologies, maybe it will find it’s way into a development version of Suse for testing and development. Maybe.
Who knows. It’s all speculation for now. Could all be much ado about nothing.
So OOo2 is identified as one of their biggest challenges for the desktop due to it’s design, and they’re relying on Sun to improve it because they don’t have the resources to justify heavy development on their end.
Fair enough.
But seems to me that investing in improving OOo2 would do far more to help them advance linux as a business desktop than creating an OpenGL based desktop environment. I have to assume they’ve got more than one or two guys working it if it’s really going to fly without external testing and development, even in pre-Alpha stage.
Good points, but the problem is that the OO.org codebase is so massive, so convoluted, so packed with cruft from possibly as far back as the late 80s that it’s next to impossible to get anyone to work on it – outside of Sun.
Look at it this way. Novell hires and engineer or assigns one of their engineers to start working on OO. They take a couple months (or longer) to get up to the speed on the codebase and then start submitting patches. But hold on. This guy works for Novell and these Sun guys maybe aren’t so thrilled some Novell guy bogarting his way into the scene. Politics is everywhere and it’s recognized, but rarely discussed except by developers.
XGL is one of those sexy things that Novell can control development on. Just like Trolltech controls Qt and RedHat is de-facto controller of Gnome.
People need to stop acting like there’s some kumbaya, utopia in open source development.
That was a damn hilarious read actually, and shows how clueless they still are:
Email call center workers wanted good music software, too. So we worked on that; see Banshee.
Yer, it’s called Amarok and it didn’t need Novell to spend money and time on creating a new open source project.
And this I find particularly funny:
Q. What’s stopping Novell from putting more engineers on OOo? A. We have a few already. It’s money, really. If we could have a revenue strea from it that’d help.
ROTFL. Well, that just tells me that Open Office is not viable as a project to do what Novell wants it to do. Either they get more companies outside of themselves and Sun to help improve it or it’s tough luck I’m afraid.
WMV/WMA/MMS support needed! Real’s been great, maybe they can help. If Linux can’t play these, people will think it’s a cheap piece of crap.
Errr, well no. That support is ultimately down to Microsoft. The way you get around it is by really pushing support for open media formats like ogg. They should concentrate on getting solid MP3 support first, because that is something that is feasible to do.
I shake my heads at these people who are supposedly fashioning desktop Linux’s future.
Edited 2005-12-21 10:24
“Yer, it’s called Amarok and it didn’t need Novell to spend money and time on creating a new open source project.”
I use Banshee and I’m very glad they decided to do it. Amarok is no Banshee.
(that may be a good thing in your opinion, of course. I’m just saying that, to me, Amarok doesn’t come close).
Slightly OT, but I’m getting rather sick of the modding system (or lack of it) around here. It’s time for a Slashdot style modding system where people have to earn the right to be modders and where trolls who mod down without replying actually giving a sensible, adult reply are strongly discouraged.
Edited 2005-12-21 09:58
The fact that Jon Smirl was abadonning Xgl was widely discussed in the OSS community. KDE could have stepped up to the plate at any point and helped the project along with either funding or developers (as could any other organization). KDE didn’t. Novell did. If they don’t like they way Novell is running the project, KDE is more than welcome to fork the last available source and do the development their own way.
Adam
??? Who talked about KDE?
But anyways, the problem is not that Novel took up Xgl, thats all fine and dandy and Novel is to be lauded for that.
The problem is that Novel took the development private, meaning that people need to
– either duplicate work (fork),
– or wait
and that is not an ideal situation.
The author of the article is a KDE developer.
As for the non-ideal situation… Tough. Any hypothetical person who wants to fork the code could have jumped in and helped Jon Smirl when he was working on Xgl and prevented this situation. They never did.
Adam
> The author of the article is a KDE developer.
The author in question seems to be also:
– human
– living in canada
– male
Here are some alternate versions of your headling:
“It’s humanities own fault.”
“It’s the canadians own fault.”
“It’s the mens own fault.”
I must say the candian one has a certain ring to it…although the kde version is far more catching in these surroundings 😉
> As for the non-ideal situation… Tough. Any hypothetical person who wants to fork the code could have jumped in and helped Jon Smirl when he was working on Xgl and prevented this situation. They never did.
Well, Novel could simply work in the open. They chose to not do it, thus these reactions…Tough…
As you see its getting to a lot of coulda’woulda’shoulda’…fact is we have this non-ideal situation. What do we do about it?
I think we will wait, and while we’re at it, we will complain. Then at some point Novel might have made something nice and the whole thing will be resolved and we will be a big happy family. Or Novel will have screwed up, and we will complain some more and sulk about the lost oportunities.
And that will be it.
(Novel probably could have had made the whole thing easier but…ah…lets not start over here 😉
I must say the candian one has a certain ring to it…although the kde version is far more catching in these surroundings 😉
It’s also more appropriate, given that he’s introduced as “Aaron Siego of KDE”.
Adam
Whta license is this XGL code under, and when it was previously released more openly what does the license allow?
I think the license may be under the XFree86 license but I can’t find where it is explicitly stated though. What does this license allow?
Linux is really a XFree86/GNU/Linux system…
Wow, there’s some rampant cluelessness here.
The Xgl development used to be in the open, directly (as far as I could tell) on the CVS tree you see on freedesktop.org. There weren’t many contributors, sure, but its utility up until very recently has been quite limited (and there’s been rapidly rising interest on the part of current X/DRI developers in the last few months). However, in the last 5 months there haven’t been any commits to it by its principal developer, David Reveman. He’s currently working for Novell, and I’ve heard (though not directly from him) that he’s been instructed to keep changes private, basically until Novell can make a big flashy release.
Some of these changes that Novell’s holding, like xvideo support, are things that current Xgl users would really like to have. So open-source developers looking at writing and committing it themselves, duplicating effort (and there’s few enough X developers already, without us wasting our time rewriting each others’ code). But there’s been a lot of “Well, I won’t bother working on it for now, since any day now there’s going to be a huge code dump from David that will change everything.”
The issue here is that Novell is taking developers away from the open-source community for internal closed-source development, resulting in duplicated effort from a limited pool of talent or diverting new development efforts, all for a tiny bit of theoretical marketing gain.
— Eric Anholt