Earlier this month, we reported that Debian had announced a new release schedule; a freeze during December, a release some time in the first half of the following year. After outcries from the Debian community, the December freeze aspect of the plan was reversed. Since most of the ire about this situation seemed to be directed towards Ubuntu, Mark Shuttleworth decided to step in and offer to put several Canonical employees to work on Debian instead of Ubuntu.
A lot of the ire within the Debian community about the release cycle decision seems to stem from a “Ubuntu vs. Debian” mentality that only works to the detriment of both projects. In an email to the debian-project mailing list, Shuttleworth detailed why coordinating releases between not only Debian and Ubuntu, but also upstream and other distributions, can only work to better the Linux ecosystem.
“We’re already seeing a growing trend towards cadence in free software, which I think is a wonderful move,” Shuttleworth writes in the email, “Here, we are talking about elevating that to something that the world has never seen in proprietary software (and never will) – an entire industry collaborating. Collaboration is the primary tool we have in our battle with proprietary software, we should take the opportunities that present themselves to make that collaboration easier and more effective.”
In order to meet the upcoming December freeze, Shuttleworth is willing to commit several Ubuntu developers to the Debian project. “If the Debian community is willing to consider a December freeze, then Ubuntu (and Canonical) will commit resources to help Debian meet that goal,” he writes, “It means we’ll get less done in Ubuntu, but the benefits of having a schedule which could attract many other distributions would outweigh that.” He believes other distributions will join the December freeze. He added that if the December freeze is not possible, he’s also willing to disrupt Ubuntu’s release cycle to make it all work better – but he won’t commit to both developers and a change in cycle.
He closes his long email by adding: “This is a good faith offer of help and support in order to reach a tough but noble and achievable goal. It won’t be easy, the first time or the next, but it will kickstart a process that will bring dividends to Debian, and to the whole broader ecosystem. Ask upstreams what they think, and whether they would want to participate, and you’ll hear a very positive response.”
I think this is indeed a golden opportunity for the Linux world to show that by being open, they can collaborate openly to make not only the lives of users easier, but also of distributors and upstream developers. A release cadence is a great idea that fits the Linux world well, and will improve the experience for all involved.
This was the situation last I heard anything:
http://lwn.net/Articles/344712/
Considering everything on Debian’s plate at this point, a December freeze for this cycle was always, shall we say, exceedingly optimistic. The people in charge of the different pieces that have to come together to make it happen are being consulted as to the feasibility of the plan for this release cycle.
Stating that “After outcries from the Debian community, this idea was reversed.” is not really an accurate description, and makes it appear that OSNews is trying to stir up a controversy. Though I’m certain that is just a misunderstanding, since OSNews is known to be above such “tabloidish” behavior.
It would be very good to see this offer accepted.
Edited 2009-08-09 19:32 UTC
Yes, the decision on a December freeze date was reversed:
http://www.debian.org/News/2009/20090730
As for the controversy bit – what’s controversial about it? As the announcement above states, the December freeze date was abandoned after community feedback… Exactly as the article here states.
From your link:
The decision was not reversed. Only the actual date of this freeze is in question. Not unexpected, since a December freeze for this release would have been pretty aggressive, to say the least. A “reversal” would have meant going back to the old policies, where the freeze date would remain nebulous and unknown for some time to come, accompanied by a chorus of “It’s read when it’s ready when it’s ready”. My perception is that the plan to synchronize releases is still alive and intact. But this December was just impractical for Debian. In that context, Canonical’s offer to loan paid developers to Debian makes a great deal of sense. (If Debian were rejecting the plan outright, it would not make so much sense.)
I think it would be foolish of Debian to refuse the offer. And I don’t expect that they will refuse it.
Edited 2009-08-09 20:03 UTC
Ah yes, in that sense there is a bit of a reference issue in my English there. Let me fix it. Thanks
Sorry to be so nit-picky. But when it comes to the Debian-Ubuntu powder keg, one really can’t be too careful. And it’s pretty important to me to see those distros and communities living and working in harmony. It’s my fallback goal after I finally gave up on World Peace…
Edited 2009-08-09 20:12 UTC
I don’t see Debian biting on any of this. It always seemed to me that Debian was about stability and doing their own thing. Not success, not cutting edge, not involvement with the prop software world.
I think, or maybe I’ve been out of the Debian loop for too many years, that this is going to fall on deaf ears. I never thought Debian devs cared much about success and market share or competition. But instead a quality stable product created in ways that Debian collectively sees fit.
Who knows? I’d love to find out. Comment away!
“It’s ready when it’s ready” is more often an excuse for poor release management practices. By poor release management practices, I mean things like:
1. Vague definition about feature goals.
2. Unrealistically ambitious feature goals.
3. Inability to say, or difficulty in saying, “No” by key managers.
4. Lack of anyone with enough authority to make the big decisions.
5. Unwillingness, or not having anyone with the authority to drop features which turn out to be best left to the next release.
This is not to say that Debian does not stand for quality. But from what I’ve seen, there is plenty of room for improvement regarding discipline in Debian’s release management plans. Injecting proper discipline, etc. into it’s policies could improve both the predictability *and* the quality of Debian releases.
4. Development freezes which become development “slushes”.
As far as I can see, and as far as most of Debian’s developers are concerned (read the replies to Marks mail, phew!) it is absolutely none of Ubuntu’s and Canonical’s business how Debian’s developers run their project. Indeed, Debian’s developers can argue that Ubuntu hasn’t given a shit about Debian over the years with the removal of many developers and resources. What gives now?
I might personally agree that Debian’s long winded development and release schedule isn’t terribly conducive at times, but that’s their prerogative and you don’t go about changing it by attempting a coup of a project by the back door where the wider development community learns of what is happening through their own press releases.
This is really crap. KDE released 4.0 when it wasnt ready and people are still complaining. Debian, releases when its ready and you are saying its because they cant manage their schedule?
People: stop bashing these projects. If kde was released when it wasnt ready from YOUR point of view, it doesnt matter. KDE has a community and they all know better than you about when to release KDE.
Also, stop bashing about Debian. It has a community and they know how things work better than anyone else.
If you think debian sucks because its not released when you want, it doesnt matter. It doesnt mean Debian was/is doing wrong.
People jumping from middle of nowhere and telling how big upstream FOSS projects suck is stupid. All of such decisions in these projects are made by their community, not a single person; You really think all those people involved were wrong and you know best?
Communities often don’t know what’s best for them. Elections are a prime example for this.
If you are inside a project, you tend to lose the outside perspective. But the outside viewpoint is the one your users have. With an inside view you tend to become too concerned with all the details and lose the overall focus.
Don’t misunderstand me. Communities are great. But the KDE 4.0 disaster is the prime example for a bad decision. Everybody, the individual and the community make bad decisions from time to time.
When talking about a community of millions of people where most of them are not even involved in the decisions, thats right. On small communities like this, where most contributors are users theirselves, its not right.
Its a clear example of people yelling at others when they dont know many things.
They released 4.0 because they had their reasons. The most important thing for projects like KDE is to keep developers. KDE 3 to 4 transition was a huge risk for KDE because the task was so huge and it should’nt have took long because that would cause developers lose their interests. I think not releasing 4.0 at that time might have caused KDE’s dead or heavy loss of developers.
4.0 had its downsides but was necessary at its time. It was a tough decision for kde to release it of course. They even postponed 4.0 release for a few months. They warned users but immature distro’s released it as their default DE.
(not that im a kde guru.I’ve been following kde from close for a few years and thats what i got from what happenned at 4.0 release)
Edited 2009-08-10 09:34 UTC
Part of a FOSS community is feedback from the users. There will be people who complain simply for the sake of making noise but there will also be valid points expressed for the sake if improving the software. Demanding everyone outside the clique shut-up and stop complaining is what Apple and Microsoft do. When FOSS projects stop listening to users and telling them to bugger off then it looses a big part of what makes FOSS great.
With your specific examples:
Debian Lenny has been great but a quick roll over to Squeeze Stable would mean not going without OpenVAS updates in Lenny backports. Provided the rolling distro upgrade process remains clean, it’s not much of a problem for my servers either. I really do like the release when ready approach for Debian but regularity would help with conservative thinking management trained to accept Microsoft’s monthly Tuesday schedule. I think it’s imperative for those who feel more strongly about it to be heard though. When one is on the inside looking out, you need someone outside to tell you when your getting a little crazy.
KDE4, well.. I’m still waiting for that to mature. I put a KDE3.5 liveCD in a notebook and it ran create. I put a KDE4 liveCD in the same notebook and it nearly locked up trying to deal with the increased resource requirements. Not really the step forward I was hoping to see when KDE4 was originally announced. We’ll see how it’s doing with Mandriva’s next release. In this case the “release early, release often” was not kind to KDE’s new direction. The points I’ve heard about KDE4 have been well founded. Increased resource waste with only cosmetic displays to show for it. A file manager that is not yet on par with the 3.5 preferred file manager. Change for the sake of change is supposed to be less of a motivation for such cornerstone FOSS projects so I hope KDE4 matures quickly and shows the benefits behind the changes. For me, it’s not there yet so KDE3.5 remains in place on my machines. Again, if your on the inside of the project ignoring all the comments from third party observers; how do you steer your project or judge how effective the changes are for the end users?
Open means inclusive; all may comment. Open does not mean “only if your one of the select few intimately involved in the project”.
Im not talking about how good 4.3 is or how many people are satisfied. You guys just have to complain no matter what the topic is?
If the complaint is justified then it’s worth hearing and considering. if a project simply says “sorry, your not part of the community.. you have no opinion of value” then it looses that feedback. You mentioned KDE and that people complained when it was released early; many of those complains where justified against what was supposed to be an improvement over the previous version. What if the complaint actually leads to identifying a bug that the developers have missed?
It’s not about complaining for the sake of making noise. Hopefully, those types are very obvious and can be filtered out. It’s about not ignoring all complaints so that the valid one’s can be considered. FOSS is inclusive not an exclusive elitist club; joining in with what one has to offer is it’s life blood, even if that is simply a user experience report.
You are totally misunderstanding what im saying. Noone says projects should not listen to their users or that their users are not a part of their community. KDE fixed a lot of the problems and issues their users had and is still continuing that. But on situations like release schedules and such, developers have their reasons and points that users might not be aware of them. So yeah, i still think Debian developers know best about Debian, as well as KDE developers who know best about KDE.
Your first post read more like anyone complaining had no place unless part of the developer circle. I’m open to that being my own miss-reading though.
In terms of release dates, I’d say the developers can still derive value from listening to what outsiders have to say. Assuming those outsiders are the target audience of course. The final decision does come down to the project manager though as they should have all the insider information along with the outsider opinions. I wouldn’t suggest that release dates be purely based on who complains the loudest from outside.
Of course criticism is not bad and Debian should listen to its target audience. Im just saying people should not criticize everything, no-matter-what.
You got a 6 score, yet Debian is still one of the best distros there is. Stable gets on every server I lay my hands on, and testing is always very good on a working desktop. Everything else – release schedules, other distros’ release schedules, funny people’s “feature” requests, etc. – just doesn’t matter. It’s working, it’s good. EOL.
Why wouldn’t the Debian project not want people who (either because of incompetence or cold-blooded sabotage) ship broken KDE packages since years?
KDE 3 under Kubuntu looked like this: http://www.flickr.com/photos/19616885@N00/2535182834/in/set-7215760…
Just yesterday a blogger who appears in Planet KDE wrote this: http://soliverez.com.ar/drupal/node/94 (Quote: “It looks to me like some system, Rosetta or whatever, doing a carpet-bombing translation without taking into account the parameters.”)
Even a Kubuntu team member blogged about this a while ago: http://apachelog.blogspot.com/2009/04/facts-about-rosetta-and-kubun…
Years of bug reports and it’s still broken.
So yeah, Debian. If you want a horribly broken 6.0 release, go and accept that offer!
I probably shouldn’t be feeding the troll. But… the KDE packages are not germane to this discussion. It would not make sense for Canonical to put people on Debian’s KDE packaging, but to select areas where the efforts would avoid duplication of said effort.
If Ubuntu (or Debian, for that matter) used KDE for its default desktop, like… uhhh…. like… like… uhhhh… you know… like… (Google Google Google) …like PCLinuxOS and Slackware, then, assuming Kurt’s accusation held any veracity, it might be a point worth consideration when planning the details of their joint effort.
Edited 2009-08-09 21:22 UTC
How am I a troll when I point out that Canonical breaks packages since years and “helping” Debian would probably cause more work for the Debian people, because they have to fix breakage caused by Canonical.
I’m not making that breakage up. See for yourself if you don’t believe me.
Which DE is the default is absolutely irrelevant. Packages should never be broken by the distributor.
Edited 2009-08-09 22:48 UTC
See where? I don’t see you siting any examples or links to look at, and I’ve not had any breakage with Ubuntu in years although I do sometimes question their decisions to package certain software a certain way. If it’s just the KDE stuff you’re talking about, I don’t think you’ve any reason to worry if Debian accepts this proposal, as I doubt Canonical will commit much if any resources to KDE anyway.
Read my first post or boot Kubuntu in VirtualBox using another language than US English.
The X server crash on an update, problems with Enterprise Volume Management when people don’t even use it, the Rosetta system that rides roughshod over how projects do translations as he’s said……… If you haven’t encountered at least one of those, well, let’s just say that you’re beyond help.
There’s certainly been enough to question Ubuntu’s overall competence, let’s put it that way.
I’ve never encountered these problems but maybe that’s because I know what I’m doing and don’t endlessly whine and went my irrational dislike for Ubuntu on public forums.
If by “I know what I’m doing” you mean just never ever try to run Kubuntu in another language than US English and pretending that the packages are not broken for ~95% of the world’s population (the ones that don’t have English as mother tongue), then yes… you know what you are doing.
95% of the worlds population is not running localized Kubuntu.
It’s not really the point, is it sunshine? The fact is that Ubuntu’s braindead Rosetta system has broken translations by riding roughshod over upstream, and it’s done it for an awful lot more than KDE.
A great deal of the world doesn’t speak and use US English, and worse, there is there any reason for their translations to be broken versus upstream, so to categorise it in that manner is the sort of denial you’d expect really. What does Ubuntu mean again? I forget.
Maybe not, but I’m not the one who brought up that inaccurate, made-up number to make a point.
For good reason. It’s terribly broken.
I’m the one behind the above linked Flickr set, btw, and I agree fully with the comments about Kubuntu’s bad shape.
I would be in favor of simply dropping Kubuntu. Leave KDE to the distros which have decided to focus upon it as their default. Which is admittedly not many. And the ones that do are not all that popular. But it makes sense to choose a DE and focus on it, without getting distracted by trying to support others.
Well, apparently this Rosetta translation problem is affecting all of Ubuntu so that must mean “regular” Ubuntu sucks balls too.
I don’t use a localized version so I don’t much care though.
Well, maybe more people fix the breakage in GNOME manually.
Yeah, that’s probably the best solution.
Many distros are DE-agnostic or mabe even default to GNOME ans still have great KDE releases.
openSUSE doesn’t have a default and its KDE is among the best. Fedora defaults to GNOME, but its KDE is also good.
KDE doesn’t need special treatment. Just shipping vanilla packages would be enough, but instead Canonical breaks them. If Kubuntu was just Ubuntu + vanilla KDE, it wouldn’t be anything special, but it wouldn’t also be that bad.
All you had to do with those things was do updates to your distribution. The X server crash was certainly a biggie and the EVS problems are well, well documented. Just type the relevant terms into Google if you don’t believe that you had those problems ;-).
No. You whine about your undying love for it and deny that you’ve had any problems at all, which is what most fanboys do on Ubuntu’s forums when rational newbies have a problem.
Segedunum, you seem to live in your own fantasy world. I’ve certainly never noted Soulbender whining or otherwise emoting about his undying love for Ubuntu.
And I have great respect for Ubuntu’s forums. There’s lot’s of good information there. Lot’s of answers. And when there is a genuine problem, that gets discussed, as well. Oh, perhaps there is an anti-Segedunum here or there, preaching some gospel. But they are few and far between. And obvious to the rest of the forum members.
Don’t bother feeding him Steve. You can’t explain to someone who closes their ears… I’ve tried.
I don’t really care about your examples, I don’t use Ubuntu and think that anything other than their LTS releases is pure balls.
I also wasn’t arguing about it either, all I’m saying is if they put people to work on the source instead of the break you won’t have half the crap you get.
The more unmodified Debian packages they use in Ubuntu the better distribution it will be. Win, win for everyone.
Kurt, I assert that you are acting like a troll because you are taking your own claims of breakage to packages which are mostly irrelevant to the team effort being proposed, and using them to further your own selfish agenda.
This tentatively agreed upon alliance between Debian and Ubuntu is too important to both distros, and potentially to the Linux eco-system itself, for me not to point out when someone with a chip on his shoulder, and a personal agenda in mind, shows up to try to disparage one or the other party for his own purposes.
Edited 2009-08-09 23:09 UTC
How is it that good for the Linux eco-system? I’ve said it before, and I’ll say it again: Ubuntu doesn’t contribute as much to the Linux community as a whole in relationship to its marketshare. Point blank, Ubuntu is the largest home desktop distro, largest OEM installed distro, and gaining significant ground in the server/enterprise field. However, its contributions to the development community are dwarfed by its competitors. Exactly how important is this alliance with Debian? Debian did fine forever without Ubuntu, and it will continue forward without it. Ubuntu needs Debian more than Debian needs Ubuntu.
Well, the point you are missing is clearly demonstrated in the above quote. Ubuntu is, indeed, far and away the most popular desktop Linux distro. And it has expanded the boundaries of what the Linux desktop market means.
You like Debian? Fine.
“A rising tide lift all boats.”
That’s what Bob Young kept saying, over and over again. Till we were sick of it, even. But he was right. We are all better off because of the name recognition that Red Hat has achieved in the server market. Our tide has risen with Red Hat’s
Debian was never going to become a wildly successful desktop distro. And if you thought it was, you were dreaming. Debian has lot’s of greatness about it. But it needed something else to really take off in popularity among the “non-elite”. Ubuntu has been that something else. Ubuntu has been a reminder that there is more… much more… to winning the game than just developers.
Which brings us to the concrete point that you raise, which is a good one. Can Canonical contribute enough paid developer time, strategically targeted, to allow Debian to meet an accelerated schedule for this release while still maintaining quality?
Debian may have scads of developers on their developer rolls. But being on the developer rolls is Debian’s substitute for “payroll”. And I’d wager that the *strategically targeted* bit that the *paid developer* factor affords can offset any difference in total number of developers handily. Because cats are awfully difficult to herd. And Debian has a *lot* of cats.
Popular Linux distributions achieving some sad number one status on Distrowatch (all that Ubuntu has achieved) over the past ten years have been two a penny. By any metrics that actually matter we’re not seeing browser usage stats from Ubuntu desktops increase, Canonical can’t rattle off a list of software vendors who are writing software for their desktop and we’re seeing zilch overall that indicates any adoption beyond the current average Ubuntu user. Nobody, in other words.
After all these years and in view of that mountain to climb, what are we arguing over that is now going to tip the balance and is so, so important to the general Linux ‘ecosystem’? A coordinated release schedule between Debian and Ubuntu. That’s it. It’s so stupid it ceases to be funny.
I’d love to see that ‘something else’ that Ubuntu is doing that other distributions aren’t. No one can ever come up with anything when that gets asked though.
The more interesting and amusing question is why, if Canonical are maintaining their own distribution that has diverged greatly from Debian, are they eager to contribute developers to a cousin distribution to coordinate releases? Would those developers not serve Ubuntu better directly?
It makes little sense until you conclude that Canonical might be doing this to share more development work with Debian. Why now? Well, the only conclusion you can draw is that Canonical might be finding it difficult to maintain their own distribution. Notice that it’s an aligning of Debian with Ubuntu schedules. From the replies to that mail most Debian developers are rightly puzzled as to what they get out of it.
I can only see this ‘alliance’ that I haven’t seen the wider Debian community agree to as anything other than a vague, attempted takeover of Debian by the back door for Canonical’s own benefit. I think most Debian developers only learned about this from their own press release.
How about getting non-geeks interested in at least trying Linux? I’m not an Ubuntu fanboy by any stretch of the imagination, but they have arguably done a *ton* for Linux in terms of mindshare.
My “own selfish agenda”??? WTF? I’m not selling a competing Linux distro, I don’t have any investment in a competing distributor or anything else. What would I have to gain by discrediting Canonical under false allegations?
Packages shipped by Canonical are broken and they are not broken because the upstream packages are buggy, but instead the bad shape was applied afterwards.
How is it selfish that I report bugs for an OS that I don’t even use by myself?
Canonical’s broken packages give FOSS a bad reputation. They don’t harm me — I don’t even use Kubuntu outside of VirtualBox.
Having developers fixing up bugs in Debian will undoubtedly help both Debian and Ubuntu. Ubuntu bringing in more stable packages initially will reduce possible breakage later.
You are trolling when your argument boils down to “They broke packages before so they’ll do it to you too!!”
You can look at what actually happening with Debian migration to Python2.6:
http://article.gmane.org/gmane.linux.debian.devel.project/16726
http://article.gmane.org/gmane.linux.debian.devel.project/16739
http://article.gmane.org/gmane.linux.debian.devel.project/16740
http://article.gmane.org/gmane.linux.debian.devel.project/16743
edit:post preview lied to me so no fancy url
Edited 2009-08-10 12:41 UTC
Experience can be a funny thing you know….
Edited 2009-08-10 13:49 UTC
So? Then don’t put those people on the KDE packaging team and let them do something they’re good at. You sound like you’re promoting hostility for the sake of it.
I can see Canonical wanting to help with the Debian freeze in December. That would give them time to hit the April Ubuntu release which should about be scheduled for an LTS one. If Debian could really commit to a two-year release schedule, then Canonical could commit their schedule of LTS releases to follow a few months behind… if Canonical can hit that schedule then they can pay developers to work directly on Debian.
Debian lost it’s place because common packages in Debian have the last TWO popular releases flagged as “unstable” because the crew drew a line X years ago and never called it “stable” so new software versions can’t even be accepted until the NEXT version. Two years lag for programs like for example Firefox is obsolete, not stable leaving the majority of users fending for themselves over on “unstable” to get “less old” versions.
“When it’s done” can grow fondness for a project but in Debian’s case it’s only growing fungus.
I think Canonical should take their core-dev’s (most are DD’s) and give something like 25% of the workday to participate in Debian, without changing the rules of it, because I really doubt this propose will be accepted.
Debian should give away BTS and adopt something like Launchpad. It will remove most of the third-party scripts to check for bugs, statistics, and have a better UI to submit bugs.
Another idea.. how about to have apport in Debian ? It could be useful for both sides.
Is Shuttleworth still banging that daft drum on coordinated releases? Just because a release at some point is good for one project it doesn’t mean that a release at the same time is good for another project. Flooding the Debian project with Ubuntu developers just to make that happen doesn’t strike me as a good idea.
I would have thought he’d have better things to do.
He’s still making the point. And the release folks at Debian have pretty much agreed that it makes sense. (In so far as agreement is ever reached about anything at Debian.) The question now is whether the synchronization can reasonably be accomplished this release cycle. Canonical is offering to try to help Debian out on that account.
For my part as an admin, I would accept a late Ubuntu LTS release if it would help accommodate the sync up.
Edited 2009-08-10 01:12 UTC
Yer, and he’s got no clue what he’s talking about. He seems to believe that he’s going to get Debian, Ubuntu and many upstream projects all releasing at the same time for the benefit of Ubuntu without a thought or a care in the world as to whether a given release schedule is right for a given project. All time based releases seem to have achieved is a version bump and nothing new as far as I can see. If the product is hobbled then I don’t see what having a synchronised release gets you.
Basically, if you want to help upstream help you then get off your lazy, fat arse and contribute what you need to help you achieve the release date you want for your own project. It won’t get done in any other way.
Roughly translated, Mark’s idea is “What I want is what you do”. Truly groundbreaking.
Ha, ha, ha. Yer, the release folks seem to have agreed but most of Debian’s developers learned of this through one of their own press releases. Has it been decided that because Debian’s decision making processes are so antiquated that this is the way things will now be done via the uber new visionary?
Hmmmm. Well that would be a first ;-).
Shuttleworth, for his part, has never seemed interested in more than giving back to Debian, contributing to the upstream project he relies on so much. This statement is in keeping with all of that.
On the other hand, if you read some of the first comments on his post, you can see that at least some Debian developers are far from forgiving Shuttleworth for taking their hard work and building a commercial product off of it, siphoning off their developers with money, and not following the Debian code of conduct they hold so dear. They’re already apparently angry that he knows more about Debian’s plans than they do.
I don’t think they’ll accept his offer. To do so would give Shuttleworth more control (or at least the appearance of it) over Debian decision-making.
And you think that being interested in giving back to Debian is a bad thing. That’s… very odd.
Adric, please. Try to stay reasonable. Firstly, Ubuntu is hardly a commercial product. Canonical does sell some commercial support. But people can and do provide Linux support commercially without making the investment of time, sweat, and money that Canonical does in the actual product. Red Hat builds a commercial product out of other people’s hard work, moves hundreds of millions of dollars a year in revenue, and does contribute something back. Although what it contributes back is *substantially* less than it gets them in revenues. And there is absolutely nothing wrong with that.
So some Debian devs are in a snit. Debian dev snits are generally measured in kilo-posts per thread.
I’ll bet they will. Because it is the reasonable thing to do.
So you are saying that Debian should make its decisions based upon politics rather than what makes sense for the project. Amazing. I give the project leaders somewhat more credit than that.
Perhaps I should have prefaced my comments a bit… I was restating the issues as I saw them in those threads, and from experience. I think Mark Shuttleworth is basically doing all he can to work with Debian, and doing his darndest to convince them of his sincerity. I think it’s a great idea, and I continue to try to wrap my head around why some Debian developers can’t see it.
Meanwhile, I think the Debian developers are being stupid if they can’t see that someone who is willing to do a job they love for free, wouldn’t be more willing to do the job they love for money. And they blame Shuttleworth for having the money to do so (and don’t blame Red Hat, this I have also noticed). Their own internal communications problems are no more important to the discussion than Ubuntu’s alleged problems with reporting bugs upstream.
The commenter I was principally paraphrasing blamed Shuttleworth both for an exodus of Debian developers to the Ubuntu project, AND an influx of Debian developers FROM the Ubuntu project. Shuttleworth can’t win.
And that’s why I don’t think he’s going to. At least, not without strenuous objections from people like that. I’d love for this to work, but whenever I hear about Debian I invariably see comments with that level of vitrol, and I hear about projects stonewalled by those people. I’m sure I’m committing a fallacy of generalizing “some” to “all”, but as I recall the attempt to set up a developer fund to pay to get the last release out, didn’t work so well.
No, I was saying that Debian shouldn’t make decisions based on politics, but that they would anyway. I hope you’re right and I’m not.
Edited 2009-08-10 02:46 UTC
I misinterpreted what you said. I should have known better, based upon your previous posts. My Apologies.
Indeed.
Yep.
So who will prevail in the Debian camp, I wonder? The hot heads? Or the cooler heads? To be sure, this should be good for some first rate knock-down, drag-out entertainment on Debian’s mailing lists. (I wonder how many melodramatic resignations this one is good for?)
In the end, I suspect they will do what makes sense for everyone, even if some folks have to swallow their pride (or resign in an elaborate flourish!) before that happens.
Or maybe they’re jealous because Ubuntu has succeeded where Debian has failed. or they’re just whining morons.
Uh, they seriously whine about that? For real? Yeah, what kind of horrible person pays people to work on something they like?
Considering Debian’s incredibly long and erratic release cycles maybe they need someone who can actually make decisions.
Apparently this is the OSS way to control other entities by controlling what they are doing. Oracle was more direct and honest when its both Sun to control this company. Actually Oracle pays also for Sun’s employees.
The OSS scene has its own business logic…
Thom Holwerda wrote:
Actually, the initial criticism was that such an important decision was announced without any prior discussion on the Debian mailing lists, and that remained the main cause of discontent in the following discussion.
The discussion thread on the debian-project mailing list was started with this mail:
http://lists.debian.org/debian-project/2009/07/msg00148.html
The discussion that followed can be read here:
http://lists.debian.org/debian-project/2009/07/threads.html
… and the discussion thread continues here:
http://lists.debian.org/debian-project/2009/08/threads.html
Mark Shuttleworth chimed in because, for some weird reason, he seems to believe that centralization and synchronization of the “free software industry” would be beneficial to everyone involved (and not just to the biggest distros). Personally, I believe that diversity and decentralization are vital to the free software community because they ensure that business-oriented corporations (like Microsoft or Canonical) cannot fully control the production and distribution of free software.
However, I’d expect Debian developers will find Shuttleworth’s offer to put some Canonical employees to fix RC-bugs in Debian a most welcome gesture. A derivative distro like Ubuntu is actually helping itself most efficiently by helping its mother distro, since most of Ubuntu’s codebase flows from Debian.
Reading most of the “conversation” in these comments is very depressing. It reminds me of 2 denominations in Christianity getting together and “conversing”. By the time they have nuked each other, there is not much left. This is EXACTLY how most of the world views the Free Software community. I think that is when many people are tempted to give up and just run shiny new Macbooks – let the children fight.
[deep breath]
But then you go on, and just try to do a good job, and let the chips fall where they fall. It’s just a shame that a small group of vocal people can cause so much destruction. It all stems from the “my kingdom” mentality. People fight over their turfs. It’s ugly, it’s destructive, but it does make good television!
I know that my next statement won’t make me many friends here but I don’t think that Debian has anything to gain at all from this proposition and that Mark is basically looking for some free labor for Ubuntu to make sure that their LTS releases will do what they’re supposed to do.
For starters, Debian is much bigger in scope than Ubuntu: Debian supports many architectures as equally as it is possible given the technical restraints and boundaries; Ubuntu cares mostly about x86 with PowerPC as an afterthought. I can’t possibly see how Debian can sync its release cycle with Ubuntu’s.
Debian developers maintain pretty much all the 20000+ packages in their repositories whereas only packages in Universe are supported by Canonical/the Ubuntu community. People installing software from multiverse are on their own.
Speaking of Universe, it is mostly a snapshot of a *really* small subset of the packages available on Sid taken every six months where Canonical developers will mostly provide some polish with little or no contribution to their upstream providers. Of course, it is easy to commit to have 2000 software packages working every six months when the upstream has to care about 20000 packages for many architectures and will be lagging far behind to get their release up to the snuff.
Debian developers are mostly volunteers that work on their free time so I can’t see how they can coordinate themselves to work for free and target the needs of a competitor that will most likely give little back (Let’s face it: Mark committing to provide developers is self-serving. Mark himself said that he is not willing to align their release cycle with Debian’s and also part with the developers. It either one or the other…).
Please don’t get me wrong, I don’t like Ubuntu but I don’t necessarily hate it and even recommend people to use it when it is clearly the better choice. I hate the fact that KDE is always measured by what Canonical does with Kubuntu, though: Kubuntu is disgraceful and KDE is often blamed for the things wrong with it. I rarely agree with sbergman27 when it comes to DE choices but I think that he nailed it on a previous post: Canonical should either improve the effort put into it or just stop pretending that KDE is supported at all and drop it altogether. But having said that, I am honestly struggling to understand what Debian stand to gain from such a deal.
The reason that the Ubuntu community is up in arms over this issue is that Ubuntu stands to gain the most from this. Otherwise they could care less about what Debian is doing. I have yet to see the outcry because CentOS packages are outdated…
Debian’s policy of “release when it is ready” is one of the reasons that makes its users love it and that I use it on my servers and my desktop. People that want the latest and greatest can – and should – use testing (which despite the name is as stable or more stable than most distros), backports or even Sid with apt-pinning.
The Debian developers are rightfully entitled to evaluate if their project stands to gain anything from this deal or not.