“Ubuntu caused a lot of friction with and for Debian. In discussions with its founder, Mark Shuttleworth, and other Ubuntu developers during (and before) Debconf6, I was able to spell out the main criticisms from the Debian perspectives of the way Canonical/Ubuntu is handling things (without a claim to completeness). These criticisms mainly stem from discussions with fellow developers over the past 18 months, and I largely support all of them. I am publicising them here to help make the status quo more transparent.”
Debian and Ubuntu
87 Comments
2006-07-09 1:02 am
2006-07-07 11:57 am
-
2006-07-07 12:32 pm
-
2006-07-07 8:25 pm
2006-07-07 12:27 pm
2006-07-07 5:18 pm
very good to see this topic finally being adressed and pushed into the light. ill admit im bias, debian pure. i dont mind ubuntu for what it is doing, i have seen examples of everything he points out. i think some of the jealousy would go away if ubuntu admitted and made it more known that it was based off of debian.
Edited 2006-07-06 20:17
Ummm, if you go to http://www.ubuntu.com and click About Ubuntu the first sentence mentions Debian. What would you like them to do, plaster “BASED ON DEBIAN!!!!!” on every one of their webpages, take out an article in every major newspaper and magazine announcing their connection to Debian, and have Canonical put up a huge sign at their headquarters that says “Just to let you know, we’re based on Debian!” Christ people….
http://www.ubuntu.com/ubuntu
Oh, and also on the right of their webpage, there is a link that says “Debian and Ubuntu” that describes the relationship in much more detail.
http://www.ubuntu.com/ubuntu/relationship
EDIT: Oh, and just to prove my point earlier, on that last webpage is the following statement “Debian is “the rock upon which Ubuntu is built”.”. How much clearer would you like them to be about the relationship.
Down you go.
Edited 2006-07-07 02:02
Ubuntu are using their rights freely under the GPL. Oh, so Debian only like _some_ freedoms provided by the GPL, eh? Take it out with the GNU, not Ubunutu, or relicense to the creative commons.
Don’t look now, but Canonical is doing the same thing (being a baby). Guadalinex is a Spanish distribution based on Ubuntu, but at the Ubuntu Developer Summit in Paris, Ubuntu complained that Guadalinex is not using Launchpad like they want. Of course, they don’t have to (Guadalinex is freely exercising their rights under the GPL). It certainly looks like Ubuntu is trying to create vendor lock-in via Launchpad, which is NOT GPL’d software, and I can’t say I can blame Guadalinex.
http://trends.newsforge.com/article.pl?sid=06/06/28/192244&from=rss
I think that before you use the lock-in argument you should consider the history of Ubuntu.
Your post implies that lock-in = evil which is true if the company trying to do the “locking in” has a reputation of doing so for evil means (MS, Apple, etc). HOWEVER because Canonical has not actually done anything evil yet they DESERVE THE BENEFIT OF THE DOUBT here; they have given enormous amounts to the GNU/Linux community.
I have personally heard Mark talk about Launchpad, Rosetta and the goals of those two projects (Solving cross distrubution cooperation on translation and bug tracking) and do not believe cononical intensions are anything but pure and good.
You might rightly say that Steve Jobs or Bill Gates could just as easily give a speech where they describe their respective “lock-ins” as being “good for the industry / good for the consumer / insert lie here” BUT as i mentioned; compare apples with apples…..
With no evidence that Canonical is an evil company I fail to buy into the lock-in argument (launchpad = lockin) as anything more than a misinterpretation of their goals to improve cross distro co-operation.
Edited 2006-07-06 23:40
Here’s your lock-in:
http://www.joachim-breitner.de/blog/archives/60-Launchpad,-Google-a…
The fact that Canonical has yet to provide any sort of profitable business model (they are no subscriptions for Dapper), their insistance upon Launchpad is an attempt to create profits by means of CONTROL over FREE SOFTWARE. The headquarters of Canonical is now in the Isle of Mann, which is tax haven. Canonical is quickly looking like any other greedy corporation.
However, in defense of Ubuntu, they have a great community which is dedicated to “open” and “free” spirit of FOSS. They are by far the nicest and most active community on the web. My complaints are geared towards the corporation and not the community (which is represented by those who post here). There is fine line between the corporation and the community, and I hope the community will stand up to the corporation.
Maybe Canonical has yet to do anything wrong, but they are certainly headed down that path. There is noting wrong with skepticism, and, in fact, I would hope that more people pay close attention to that which unfolds around them. Complacency is a dangerous luxury.
>> their insistance upon Launchpad is an attempt to create profits by means of CONTROL over FREE SOFTWARE.
Ok, first how is a repository and integration tool “Control” – for those packages under Ubuntu and other projects that ‘buy into’ it maybe… Don’t like it, let’s see a Free (as in beer) Open Source Free (as in freedom) equivalent. THE REASON it’s so good and caught on as well as it has is they can do this… crazy thing… called PAYING THEIR EMPLOYEES with this stuff called… money. How evil, paying people to make a product that allows free stuff to be made better… Oh noes, run, the sky is falling.
Second, if you give something away you no longer have the right to get your panties in a twist when someone else does something that makes money with it…
>> The headquarters of Canonical is now in the Isle of Mann, which is tax haven.
Given the amount of stuff they give away free while retaining a functioning corporation, this is good business sense – GOD FORBID somebody with a head for business get involved in turning OSS/FLOSS/GPL software into something VIABLE.
>> Canonical is quickly looking like any other greedy corporation.
You know, I have to say I am DAMNED tired of hearing the words “Evil” and “Greedy” when it comes to a company trying to make money – especially when said companies are giving things away. Hate to break it to you, but capitolism won… The only people who think otherwise likely being career students having their lives paid for by mommy and daddy or career educators and lecturers – Basically people who’ve never had to do a honest days WORK.
… and free software advocates wonder why they get laughed out of business meetings and insulted as ‘dirty hippies’ – look in the mirror.
“The only people who think otherwise likely being career students having their lives paid for by mommy and daddy or career educators and lecturers – Basically people who’ve never had to do a honest days WORK.”
Wait a sec, isn’t teaching one of society’s most important (and, evidently, underappreciated) jobs?
>> Wait a sec, isn’t teaching one of society’s most important (and, evidently, underappreciated) jobs?
Yeah, but what’s the old saying? Those who can, do… those who can’t, teach.
The comparison between open source developers and teachers works pretty good in my mind – they are often some of the most brilliant people I’ve ever met – and at the same time I wouldn’t trust spot one of them to find their way out of a piss soaked paper bag with a hole in the bottom.
There’s a certain… detachment from reality, that goes with both vocations.
Some of the complaints seem a bit off (e.g. the jeolosy part) and lead to a no-win result (e.g. acknowledging the jeolosy would lead some to believe the Canonical was arrogantly saying they were better).
However, Ubuntu as Mark Shuttleworth readily admits, depends crucially on Debian, and if there are things that can be done to improve the relationship between Ubuntu and Debian it’s fair for Debian to ask. It’s also fair for Ubuntu to make similar requests for Debian. For instance ‘this email’ indicates the flip side of one the the complaints:
http://lists.debian.org/debian-project/2006/06/msg00296.html
It would be a lot easier to comply with Debian’s requests if Debian made it easier to comply, and that benefits not just Ubuntu, it benefits all Debian derivatives.
Respectful talk and negotiation between projects, IMO, is always a good thing.
It’s legal for me to call you all sorts of names but I doubt you’d like me if I did, and you’d probably mention that you dislike it.
Welcome to how functional communities function, by treating people with respect and not by being police states.
From what I have seen, the root of the “problem” is that in all these years there has never been a strong Debian based distibution. There have been many Debian based distros, but they have always been distinctly less influential than Debian, often struggling financially. The Debian dev’s got arrogant. The Debian users got arrogant and set in their ways.
Then along comes a strong Debian based distro that does a lot of things better than Debian, is not struggling financially, and passed up Debian’s user base almost immediately.
I’m more of a RedHat centric guy. But I have tried Debian several times over the years and been quite unimpressed.
I’ve been using Ubuntu Dapper on my desktop and laptop for the last 2-3 weeks, and I just might not go back to my old distro, for my desktop/laptop, at least. It’s been that good. (Ubuntu Server I can’t comment on.)
I’ve read quite a few things from Debian users and devs that sound suspiciously like someone feeling threatened.
The way I look at it, though, it’s like the service business: “Take care of your customers. If you don’t, someone else will.”
Competition, meet Debian. Debian, meet competition.
Edited 2006-07-06 20:45
“Then along comes a strong Debian based distro that does a lot of things better than Debian, is not struggling financially, and passed up Debian’s user base almost immediately.”
Word! Let’s put it like this: before Ubuntu I wasn’t even considering using Debian or Linux for that matter. Now I got it deployed on computers I earn my living from.
So yes, the jealousy is indeed tangible.
That said, patches should be sent upstream in a poper and sane format. Once this is done, I don’t see what more Debian could ask from Ubuntu.
What made you consider Ubuntu then…
An up to date, supported, easily installed apt-get distro with seemingly stable finances and a proper release schedule.
One thing I think that didn’t come up in the discussion but played a big part of the “revolution” was Debian itself…No one has ever tried to claim debian was anything other than a great distro with some features still unmatched by any other top distro…but what hurt Debian for a long time and still to this day is the attitude of “some” of the community members…As a newb is is EXTREMELY difficult to find a central point for help withouth getting wizcracks and smart remarks like RTFM…In my early days with debian i experienced alot of this myself…it was almost impossible to get a quick answer out of anyone in the chatrooms…and even at the time…Debian really didnt have a centralized point of contact for users to finf more info and seek help…outside of the several different sites listed in documentation…there was no central point where the community gathered in the spirit of opensource so to speak…and Debian has/had a long history of ignoring “users” wants and desires moving forward…debian as it should be easy to find stated numerous times, was not in the business of trying to make the debian desktop a happening place…this is where Ubuntu struck oil…they took all the things people loved about the debian distro, and tried to eliminate all the “USER” complaints about the distro…One of the first sites that went up was ubuntuforums…which to this day is only rivaled by the gentoo forums IMO…ask a question, get an answer, and no question is to noobish or complex…they did a stellar job at trying to be all about the user…basically, they gave the people what they wanted and asked for, rather than trying to dictate what they should have…thats why Ubuntu became such a smash hit…I have contact with the communities of all the major distros on the block..and the ubuntu is by far the friendliest and easiest bunch to get along with, and its always easy for n00bs to come in the forums and chatroom and ask question without being belittled…give credit where credit is do….I think quite frankly Ubuntu is at the top of the charts right now because they deserve to be, they have done soem things on and for the linux desktop that needed to be tackled a long time ago…but no one was interested…and now alot of people are just plain sick because they didnt do it first…
Paragraphs. Plase use paragraphs.
I’m not trying to be troll or an ashole here, but I can’t even comment on your opinion since I can’t read it.
Which is a shame since I like to read more on OSnews than simply the articles.
I can’t even argue with u on that, you got me that time lol…was typing heading out the door from work lol…Bad Squishy…Baaaddd Squishy
It just shows how badly OSnews need a proper “Edit” as you could come back and fix it later on.
Hi, I am the author of the blog post that spawned this thread. I acknowledge that there are parts and members of Debian that are not very user-friendly. However, I also think that the situation has gotten a lot better over the past years. If you try to follow the standard guidelines about asking smart questions, I can really not imagine this happening still.
Sure, our debian-user mailing list is very big, and there are plenty of trolls and idiots; thus, the best advice I can give is to try to be as specific as possible. Are you experiencing a laptop issue? Try debian-laptop. Is it to do with X? Try debian-x. Firewalls? debian-firewall. A specific package? Look around on http://lists.alioth.debian.org/ or http://packages.qa.debian.org/ for specific contacts. And so on, you can check out http://lists.debian.org/ for pointers. This is a topic I discuss at length in my book, http://debiansystem.info by the way (chapter 10).
I guess I have to agree that Debian isn’t as much about the user as one might read into our claims. However, to us, that claim is not so much about trying to get as many users to use our distro, as it is about making our existing and potential users happy, and here I think we’ve fared pretty well.
” but what hurt Debian for a long time and still to this day is the attitude of “some” of the community members…As a newb is is EXTREMELY difficult to find a central point for help withouth getting wizcracks and smart remarks like RTFM…In my early days with debian i experienced alot of this myself…it was almost impossible to get a quick answer out of anyone in the chatrooms…and even at the time…Debian really didnt have a centralized point of contact for users to finf more info and seek help…outside of the several different sites listed in documentation…there was no central point where the community gathered in the spirit of opensource so to speak…and Debian has/had a long history of ignoring “users” wants and desires moving forward…debian as it should be easy to find stated numerous times, was not in the business of trying to make the debian desktop a happening place…this is where Ubuntu struck oil. ”
You make very valid ponts here IMHO.
When Canonical entered the market and hired some developers to work on a Debian-related project, a lot of jealousy boiled up among those who didn’t get a job
Pretty much sums up the whole “controversy”.
Ubuntu has brought paid developers, packagers, testers, designers etc to Debian. All advances made for the Ubuntu distribution are made available back to the main debian repositories. All Ubuntu does is take a freeze of the .deb’s, apply their patches and release.
This seems to be just debian people getting upset Ubuntu is gaining market share when debian is just business as usual.
The author of that web page of issues even discounts them within the text. At first, the following is noted “Even though Mark has claimed that the patches should be modular if the underlying package uses a modular package management system (such as dpatch), this does not seem to be the case at first glance…” but then follows up with “You have to look inside the monolithic patch to note that its changes are compartementalised into dpatch files.“. So… where’s the problem?! There are sufficiently seperated patches for the debian maintainers to use. The author then goes on to say “Canonical/Ubuntu either has to start giving back according to their own claims, or stop pretending that it does…” even though it appears Ubuntu’s patches are what they say they are!
The next point is really moot, as is the following one “Debian does not need any one derivative more than any other, and in fact it has been considered harmful if any derivative stands out.” …. ok, so what you’re saying is even though Shuttleworth has invested huge amounts of money into developing software that Debian can freely use, hired debian maintainers and paid them salary, and generated a huge amount of interest (for once) in a debian-derived distribution, he is some how doing a bad thing?! Anyone who uses Ubuntu knows it is based on Debian. If they don’t, it’s more than likely they don’t know what Debian is, nor fedora, Redhat, Slackware etc… all they care about is they have a web browser and email client! Ubuntu has always openly talked about being Debian based. Watch all the videos of Shuttleworth explaining the process Ubuntu is created for proof – there is no pretending or marketing.
On the ubuntu.com front page, Debian is not mentioned at all; if you actually follow the About Ubuntu link, you can find Debian mentioned once in the first paragraph (albeit without a link), and if you did notice the right hand menu expanding, you may even decide to click on Ubuntu and Debian, where the relationship is finally illustrated from the Ubuntu point of view. To many, this is just not enough.
Above is yet another ridiculous point. Why mention something as technical as it’s Debian roots on the homepage. If you really want that sort of information, I’d expect to find it on the ABOUT page, not the main page that customers of all levels of technical ability see first. c.f. my point above about simple users.
With regards the point about the “corporate image” of Ubuntu…. well… Canonical is a company, who wants to make money from support / it’s products. Where’s the problem? The integration of launchpad etc is done at a level that helps end users of its distribution, not hinder them.
It’s not until you get to the final point that you really get to understand this toys-out-of-the-pram attitude:
When Canonical entered the market and hired some developers to work on a Debian-related project, a lot of jealousy boiled up among those who didn’t get a job
Enough said really….
Ubuntu has brought paid developers, packagers, testers, designers etc to Debian. All advances made for the Ubuntu distribution are made available back to the main debian repositories. All Ubuntu does is take a freeze of the .deb’s, apply their patches and release.
I am sorry, but this is not the case. For one, many patches in Ubuntu are not applicable to Debian. Second, many patches are of unacceptable quallity to the Debian developer, and third, we still have to go out and fetch those patches, Ubuntu does not make any effort to reach out to us.
It’s also not true that Mark hired “Debian developers”. He hired people who were previously Debian developers, but who are now on an Ubuntu schedule. Within Ubuntu, the pace is fast, and they are working close to full throttle, but always with the rule in mind “Ubuntu first, other things second” (quote by a core, day one Ubuntu developer). Many of those developers have not passed on their duties within Debian, so they are still the maintainers of the Debian packages, and hence have little choice but to simply follow the path that Ubuntu is going. Debian is a volunteer project that has never been on a time line and would much rather never be (if it weren’t for our users requesting regular releases), but now we’re forced to accept all these changes Ubuntu is indirectly forcing into our sacred OS, without much of a chance to do anything about it.
I acknowledge that a lot of the issues involved are purely emotional, but that’s exactly what all of this is about.
I will not address your other points for there is nothing for me to discuss.
I am sorry, but this is not the case. For one, many patches in Ubuntu are not applicable to Debian. Second, many patches are of unacceptable quallity to the Debian developer, and third, we still have to go out and fetch those patches, Ubuntu does not make any effort to reach out to us.
Basically with this you are saying: “Please Canonical, do our work for us.”
Debian is licensed under the GPL, Canonical/Ubuntu uses Debian, builds from it and beyond, has everything open source, ready for you guys to take it back, so what’s stopping you from doing so? You actually expect to get bite-sized chunks back that apply directly towards Debian? That’s bizarre from any point of view.
Within Ubuntu, the pace is fast, and they are working close to full throttle, but always with the rule in mind “Ubuntu first, other things second” (quote by a core, day one Ubuntu developer).
It’s called ‘being on the payroll’. Ask your boss about it, if you have any.
Debian is a volunteer project that has never been on a time line and would much rather never be (if it weren’t for our users requesting regular releases), but now we’re forced to accept all these changes Ubuntu is indirectly forcing into our sacred OS, without much of a chance to do anything about it.
You are NOT required to push their changes back in. That you do only shows you see some value in them. So again, what’s the point?
I acknowledge that a lot of the issues involved are purely emotional, but that’s exactly what all of this is about.
Reading your posting on your blog it all just sounds like petty jealousy all-in-all. Instead of the whining the people involved could have ported back a bunch of patches already.
Basically with this you are saying: “Please Canonical, do our work for us.”
Debian is licensed under the GPL, Canonical/Ubuntu uses Debian, builds from it and beyond, has everything open source, ready for you guys to take it back, so what’s stopping you from doing so? You actually expect to get bite-sized chunks back that apply directly towards Debian? That’s bizarre from any point of view.
It stops being bizarre when you think about it: Ubuntu needs Debian, and any patch in Debian is a patch they have to worry less about.
However, the problem is not so much with patch availability, but patch quality. Have a look around at http://patches.ubuntu.com and note: automatically generated patches document differences between packages as they were built, but not really between the packages as they are maintained.
Debian is licensed under the GPL, Canonical/Ubuntu uses Debian, builds from it and beyond, has everything open source, ready for you guys to take it back, so what’s stopping you from doing so? You actually expect to get bite-sized chunks back that apply directly towards Debian? That’s bizarre from any point of view.
And yet if this is not done, the patching effort equals almost nothing for Debian. It’s simply not practical to extract changes from an entire package (or several related packages) which have been heavily modified in the context of another distro.
Look at what happened with Apple and KHTML:
http://www.osviews.com/modules.php?op=modload&name=News&file=articl…
Ubuntu vs Debian is actually the second high-profile case that has to deal with this issue: bypassing the spirit (not the letter) of the GPL by having a fast-moving payed team modify code to the point backporting by a slow-moving voluntary team becomes extremely difficult.
I for one hope to see some solution for this issue and I hope more companies do not get ideas from this.
bypassing the spirit (not the letter) of the GPL by having a fast-moving payed team modify code to the point backporting by a slow-moving voluntary team becomes extremely difficult.
That hardly bybasses the “spirit” of the GPL.
If anything, slowing down because someone can’t keep up would be the violation.
>> It’s also not true that Mark hired “Debian developers”. He hired people who were previously Debian developers, but who are now on an Ubuntu schedule. Within Ubuntu, the pace is fast, and they are working close to full throttle, but always with the rule in mind “Ubuntu first, other things second” (quote by a core, day one Ubuntu developer). Many of those developers have not passed on their duties within Debian, so they are still the maintainers of the Debian packages, and hence have little choice but to simply follow the path that Ubuntu is going.
And the problem with this is??? Seriously, this amounts to little more than a hissy fit over “they’re getting paid and we aren’t”. Grow up. People putting a job that pays to do basically the same thing as their volunteer project, and suddenly they are doing everything the way that puts food on the table? THE HORROR!!!
>> but now we’re forced to accept all these changes Ubuntu is indirectly forcing into our sacred OS, without much of a chance to do anything about it.
First off, who’s forcing you – if anything the complaints from the Debian side of the fence ‘it isn’t up to our standards’ or ‘the patches don’t apply to our versions’ sound more like lazyness and just plain whining about it being ‘different’.
SECOND, ‘Sacred’? Ok, people bitch about the use of the word ‘zealot’ – THERE’S WHY IT’S USED… Because in a lot of cases, it MORE than applies.
I mean what’s next, the Debian developers all put on purple robes and start eating cyanide laced yogurt so the aliens will come down in the mothership and take them to heaven?
and to top it off, the whole subject is complaining about how hard it is to reverse-in the patches that come from Ubuntu, then we get someone complaining about the changes being ‘forced’ on them. Whiskey Tango Foxtrot, over?
Ubuntu does not make any effort to reach out to us.
but now we’re forced to accept all these changes Ubuntu is indirectly forcing into our sacred OS, without much of a chance to do anything about it.
When whining, it helps not to contradict yourself.
Is Ubuntu not reaching out, or are they forcing you to change?
How, precisely, do you “force” a bunch of volunteers to do anything they don’t want to do?
Oh, right. You pay some of them to work according to your priorities. What a horrible thing, paying people and then expecting them to do what you pay for them. Even more horrible: people doing what they’re paid to do.
That’s wrong. There are enough users that do not know about Debian, or don’t care about what it is based one.
Why? They never see anything from Debian. Ubuntu devs effectively replaced most of the end-user visible “Debian” strings with their branding. And they send you a CD set for free.
And in the case that you are blind or you do not even KNOW how Ubuntu works… Debian is not just their roots, they actively pull things out of it like a parasite.
On the other hand they have the “universe” community and they recruited too many less experienced devs for it, not knowing or not respecting the Debian developers, or just ignoring them. IMO most are proud enough to not ask the original maintainer, even in case of trouble. The typical case is that at the end many Ubuntu packages have lots of small patches which of them are pointless (not well designed) and sometimes reverted RSN.
That cannot be said about all MOTUs but there are enough of that kind and “they” seem not to care about that. Often it looks like they don’t make any effort to be present to the original developers, ignoring them and what they do and believing that they can do the job better. And having the means to change things, they do change them. Sometimes there are good changes, but often there are just some meta changes, sending all that do the original maitnainer would mean a lot of noise (see above)… and keepign things in sync with
others is work.
But, hey, we don’t want to do the boring work, we are cool, we are Ubuntu, we are The New Generation, the old ones can go to hell!!!!1 (except of supporting us with all we need, of course).
It is about the attitude and the respect for the own work, from the developers and from the end users, and Ubuntu seems to want both but not give it back or forward it to the parents. And since they are not giving money either, what exactly do the Debian devs get for their work?
Why? They never see anything from Debian. Ubuntu devs effectively replaced most of the end-user visible “Debian” strings with their branding. And they send you a CD set for free.
And in the case that you are blind or you do not even KNOW how Ubuntu works… Debian is not just their roots, they actively pull things out of it like a parasite.
I did not realise the GPL required one to plaster your product with the name of the piece of software it is based on.
All according to the GPL the source code of what Ubuntu is made of is out there. So what is the problem? You don’t like them using the source code? Use a different license, when the Debian guys started working on Debian this sort of thing is explicitly allowed by the GPL.
That you might disagree with it or not based on a moral or ethical point of view is a whole different thing. But from a cold hard licensing perspective the Debian guys have no ground to stand on.
excellent article…
I think some miss the point that if ubuntu is going to claim good will toward debian then it is only right to provide good will toward debian. If they aren’t then fine but dont play both sides.
Simply excellent article… I wish/hope we could get the guys together more and see if we couldnt come out of it with a better relationship.
That being said, yes debian could improve some stuff IF we had manpower, income, same time zones and so forth. A lot of tools are there and we are slowly getting better but we could stand to pick up the pace a little.
I would like to see a user project, not about defining how to make things easier but actually doing just that. Possibly work with preseed files, godo walkthrus, video tutorials, and so forth. Things like how to get a desktop system that works and looks as good as any of the other distros. rambling….
There are two main reasons why I don’t use Debian.
1) No Core Distribution. Love the idea that they have over 10,000 applications but I really wish they had a Debian Core consisting of 1-4 CDs. This “core” should also be their main focus for advancement and also very well documented.
2) Fixed Release Schedule. “When it is done” as a schedule usually works for games and for convicted monopolists. Otherwise, please feel free to join FreeBSD, SuSE, RedHat/Fedora, and almost every other major operating system in setting deadlines and commiting to them.
I think the reason why Ubuntu is so popular is that they basically addressed these two issues. Otherwise, Ubuntu is basically the same thing as Debian. I’ve been hoping Debian would do this since ’98. I guess someone else finally listened and the Debian Project is all sour over it. Well, if you want the people back, give them what they want.
1) No Core Distribution. Love the idea that they have over 10,000 applications but I really wish they had a Debian Core consisting of 1-4 CDs. This “core” should also be their main focus for advancement and also very well documented.
The current incantation of Debian is a core. There might be a lack of focus with development, which isn’t exactly surprising since it’s completely community-maintained (without any kind of corporate supervision). Yet, they don’t want to release second-class packages. Perhaps the core is bloated, but at least you got a coherent blob.
2) Fixed Release Schedule. “When it is done” as a schedule usually works for games and for convicted monopolists. Otherwise, please feel free to join FreeBSD, SuSE, RedHat/Fedora, and almost every other major operating system in setting deadlines and commiting to them.
These distributions have delayed their latest releases by weeks… Not to mention that FC5 shipped with an unwanted feature in the kernel and SuSE shipped with a broken package manager. While at it, Ubuntu doesn’t play nice with many ATI cards.
Having a lax release schedule got its advantages. The quality of FreeBSD releases are usually good because they don’t hesitate to push back their deadlines (I believe you mixed them with OpenBSD). Debian releases are really stable. They might not be impressive feature-wise, but they are easy to work with and to extend.
I think the reason why Ubuntu is so popular is that they basically addressed these two issues.
Issues? I would rather say “ideological differences” since it’s definitely not an issue for current Debian users…
It seems to me there is a misunderstanding of the philosophy behind the Debian project. Perhaps that’s why we have the clash between some developers like mentioned in the news?
“””Having a lax release schedule got its advantages.”””
Lax? Four years is not “lax”. It is “complete paralysis”.
Sarge was released in June 2005. Etch is planned to be released in December 2006.
See for yourself :
http://lists.debian.org/debian-devel-announce/2006/05/msg00000.html
Hyperbole… Not once has it taken Debian four years. the longest has been woody to sarge, which was three years, and due to good reasons.
In any case, forget the past. We’ve learnt our lesson from sarge, and even though we’re still learning, things have vastly improved…
“””In any case, forget the past.”””
Well, here is the history according to Wikipedia:
1.1 – buzz, 17 June 1996
1.2 – rex, 12 December 1996 (0.5 years)
1.3 – bo, 2 June 1997 (0.5 years)
2.0 – hamm, 24 July 1998 (1.1 years)
2.1 – slink, 9 March 1999 (0.7 years)
2.2 – potato, 15 August 2000 (1.4 years)
3.0 – woody, 19 July 2002 (1.9 years)
3.1 – sarge, 6 June 2005 (2.9 years)
I know what they are saying, these days.
But there’s a reason that “Forget the past” doesn’t work when applying for a credit card.
I’m still betting on 4 years for the next release.
Just watch and see.
2) Fixed Release Schedule. “When it is done” as a schedule usually works for games and for convicted monopolists.
This stupid, but I’m not surprised by it, not that the others made it a +5. But thing is, for many of us (and I’m not a hardcore debian-er, still) a frequently updated and usable linux is better than “new” releases every x months. In fact this new-release method was one of the reasons for my switching away from redhat to debian back in the days. Since then I’ve found only two distributions that can fit my needs – which stand fairly high above the average clickety joe’s notions of a desktop OS – and they are debian and gentoo. And that on desktops, notebooks and servers.
Having told that, I don’t have anything against Ubuntu. In fact I always have Ubuntu disks ordered in stock to be able to give it away to people who are interested (there were times when even myself installed Ubuntu for a machine I was using).
Why ?
Since Ubuntu is Debian for the masses, in a compact and easily digestible form. And this means good, even in my vocabulary, and from this perspective it doesn’t really matter that how Ubuntu helps Debian.
From the other perspective, it really matters. Since those users who get to know Ubuntu first (and yes, there are very many of them) and later try Debian will think Debian is much more inferior which is stupid. But they will think it is so because they are on a lower linux-knowledge level than the average Debianer, mostly because how Ubuntu makes Debian accesible for the masses. In a way, that _can_ mean it’s superior. And from the perspective of a really average user, it does mean that.
What they (Ubuntu _and_ Debian together) could do to help each other out is to make the above difference clear, to clearly state that they both have benefits for the other, and besides that Ubuntu should give more credit for the “parent” distro but Debian should acknowledge that Ubuntu managed to bring crowds into the game.
Sorry for being this long.
“But thing is, for many of us (and I’m not a hardcore debian-er, still) a frequently updated and usable linux is better than “new” releases every x months.”
Who said anything about Debian having to do a “new” release every x months? I was merely stating almost every serious operating system has real dates for releases and real dates for EOL support.
For example, I know Ubuntu has regular releases every six months and will support these releases for about two years. In addition, Ubuntu has enterprise releases every 18-24 months and it will support them for 5 years.
Another example, I know RedHat releases their enterprise product every 18-24 months. Each of these enterprise release will receive 7 years of support. Also every six months, they release updated versions containing the latest security patches and driver support. On the free side, new versions of Fedora will be released every six months, will be supported a minimum of 18 months, and then will go into legacy support with an unknown support lifetime but given plenty of notice when such support will end.
Yet another example, FreeBSD also has a new release every six months. Each of these releases is supported for two years. Within whole numbers of stable releases (6.0, 6.1, 6.2,…) the have stable APIs so it should be very safe to upgrade. Once every two-three years, they have a major release (4.x series, 5.x series, 6.x serieas). Each of these series generally has a lifetime support of 4 years.
Yet another example, Sun Solaris has a support life of about 10 years. New versions (8,9,10) are released about every three years. Updates (u1, u2…) are released every six months. Like FreeBSD, these updates keep the base API stable, but fix bugs, add new functionality (ex. ZFS) and add support for new drivers.
Now what is Debian’s official policy again regarding new releases and updates?
For example, I know Ubuntu has regular releases every six months and will support these releases for about two years. In addition, Ubuntu has enterprise releases every 18-24 months and it will support them for 5 years.
Yes, that’s the claim.
cd 1 is a core, a core of the most popular packages that fit on one cd
cd1 and cd2 is a core, a core of the most popular packages that fit on two cds
etc, etc, etc…
🙂
Debian releases when the bugs are squashed… would you rather they left the bugs and just set a date?
Ubuntu is popular because they are popular, its their week, month, year, whatever…
Yea, we should join the ranks of microsoft that makes release dates and does so well with them.
> Debian releases when the bugs are squashed… would you
> rather they left the bugs and just set a date?
Actually that’s not the way it works. There are two basic ways of managing a project: time driven or feature driven.
In feature driven project management, you define the features and then spend the time necessary to squash all bugs.
In time driven project management, you define the release date then cut out features that can’t be stabilized in time. Ubuntu, Fedora, GNOME, Gentoo, NetBSD, and several other projects (including the current Linux kernel development to a certain extent) use this approach because it allows the users of their projects a way to plan their deployments ahead of time, it allows them to co-ordinate more easily with other projects, and it drastically reduces feature and scope creep (something feature driven schedules seem to have a hard time controlling) since it’s easier to reject a new addition on stability grounds than on feature grounds (i.e. this feature is useless unless it has these other 12 features).
What’s being asked for is simply to for Debian to shift to a time driven schedule (that will drop features that don’t confirm to Debian’s quality standards). It’s something even the Debian president said was a good thing, although there is some disagreement whether that time should be 1 year or 3 or longer.
Debian is trying to settle for a 1.5 year cycle since sarge’s release. However, that’s a soft goal, meaning that we will not lower our standards just to meet it, like others have done (I am not suggesting Ubuntu is necessarily at fault). Instead, we are defining the features and planning well ahead, just like you said. But we still won’t freeze before the RC bugs are taken care of. So Debian’s in the middle between time- and feature-based management and that hopefully will not change.
Another thought: we have plenty of users who do not want too frequent releases because of the “never touch a running system” secret — which applies even to Debian.
meaning that we will not lower our standards just to meet it, like others have done
I recommend that you read and understand the post above yours. Selecting time based rather than feature based project management is not lowering standards. It’s just a different approach to product management.
feature driven – well imo debian doesnt do “features”! they simply bring sid into testing and then they have to clean it up which would be to squash bugs. I guess you could say they are feature driven.
time driven – what features are they going to cut out? KDE? They dont have any major parts they make themselves. They certainly cannot roll any of the DEs back to stable version so while they can set a general date they cant pinpoint it because of a lot of factors. One factor being that they have a set of packages to work with and their is not cutting out something because it wont be ready.
Sure Debian has features, as do Ubuntu, Fedora, NetBSD, Gentoo and friends. The features are the package upgrades themselves. In a feature driven distribution, if the latest version of the Linux kernel causes has too many problems too close to the release date, you cut that “feature” and go to the tried and true previous version.
In the case of Ubuntu, being time driven meant it had to give up on possible plans to support two very popular features — XGL and some wireless integration — and instead provide a way to play with these features at their own risk.
Debian is partially time driven too (at least at the endo of it’s release cycle). Two releases ago, the the 2.6 Linux kernel was not stable enough to satisfy Debian’s QA, so Debian decided to only support the 2.4 kernel, but since the 2.6 kernel was popular enough, it chose to provide it as a “use at your own risk” feature. In the previous release of Debian, it also cut X.org *even though it was one of the few distros that didn’t ditch XFree* precisely because X.org wasn’t good enough.
The good thing about being time driven is that it’s easy to be hard nosed about it because you’re dealing with an objective criteria of code quality (which can be measured by the number of critical bugs) instead of a softer criteria like how nice to have a feature is (How on earth can you measure it? By whose criteria is it nice to have?).
The problem with the old Debian model wasn’t that it had a long release cycle — sysadmins *prefer* a 2 to five year release cycle since doing “standard operating environment” tests and “critical systems” tests are a real pain — something you don’t want to do too often. The problem was, no-one knew when the next release of Debian was going to happen (or even *if* it was going to happen) so people weren’t able to make plans WRT it.
Anyway, the point is moot now. As was pointed out earlier, Debian is trying to move towards being more time driven (once every 1.5 years, give or take a month or two if things go south unexpectedly) instead of sort of the previous mish mash (be feature driven until people get tired of waiting, then switch to time driven when the complaints are too loud).
debian core fits on one cd, and yes there is one.
the release is irrelevent because things are updated and available DAILY.
debian LIKED being elitist, i didnt and dont see a problem with that.
ubuntus patches cause as many problems as they fix. a big part of that comes from ubuntu devs submitting patches for software that has potentially been updated to a newer version already.
>However, the only way this is going to work from the Debian perspective is as a mutual effort. If Canonical/Ubuntu do not realise this, maybe another derivative has to step in to fill the void?<
A mutual effort in true partnership would benefit both projects and ultimately the user most. However right now I believe a little consolidation would be a good thing and not more derivatives.
From an efficiency point of view wouldn’t it be better to have people concentrate on fewer ‘distros’ and even higher quality instead of duplicating a lot of work? (I know, this was an argument when they started forking in the first place). Let’s hope Ubuntu and Debian can work it out.
Btw- will probably go back to Debian when etch is stable (due to philosophy) as it comes along very nicely, but there are some very visible contributions from the Ubuntu project in testing snapshot right now.
A mutual effort in true partnership would benefit both projects and ultimately the user most. However right now I believe a little consolidation would be a good thing and not more derivatives.
… which is precisely why I am doing all this. I am trying to consolidate (as are a bunch of others), but if it won’t work, it’s pointless to lose time and energy.
but there are some very visible contributions from the Ubuntu project in testing snapshot right now.
Absolutely.
nothing wrong with etch right now! go get it
Installing Debian was a bear at first. It’s all things to all people, potentially, and it’s fine for that. But it evangelizes itself poorly. Knoppix did a good job of addressing this, but it isn’t viable as a distribution when its own maintainer says not to use it as such.
Ubuntu probably comes closest to filling the Free (as in beer) Desktop Linux void as we’ve got. Unfortunately to the Debian team, a company tends to make its decisions more by necessity than consensus, and toes have been stepped on.
As unhappy as some Debian devs might be, I have to say it’s a hell of a lot more gratifying to install Ubuntu in the space of 20 minutes and go than it was to spend nearly an hour manually configuring hardware in Sarge. Left to your own devices, this would have taken another six months or longer, and you know it.
Competition breeds improvements. You want to complain about people leeching off your achievements? There’s FreeSpire over there in the corner pretending to be Ubuntu. Take aim at them.
“say it’s a hell of a lot more gratifying to install Ubuntu in the space of 20 minutes and go than it was to spend nearly an hour manually configuring hardware in Sarge.”
Compare new to old and I bet there will be some difference. Got Etch?
Quite frankly, “Monolithic Packages” – big deal – if Debian’s devs want something else, ASK for it. DOES IT DOCUMENT THE DIFFERENCES BETWEEN THE TWO??? YES, then what’s your damned problem.
Even so, the end user DOESN’T CARE – why should Canonical. The only people that DO care are those using Debian, looking across the pond at all the things Ubuntu does out of box or with a minimum of fuss, that they are spending (as snozzberry put it) “nearly an hour manually configuring in Sarge”
The bit in the article about ‘The “universe” community that gathered around Ubuntu’s core surprisingly often does not know about Debian, or do not know how Ubuntu and Debian relate‘ misses the point – it’s not that the Universe doesn’t know, it’s that it doesn’t CARE. It’s hard enough contributing to one distro, much less being expected to handle a second that you likely aren’t even RUNNING.
Top to bottom the article seems more like the same type of annoying naive whining that launched “free (as in freedom) software” in the first place. You gave something away freely, then somebody else ran and made it successful without slapping your name in lights – you gave it away dumbass, what did you EXPECT?
The term ‘sour grapes’ comes to mind.
Quite frankly, “Monolithic Packages” – big deal – if Debian’s devs want something else, ASK for it. DOES IT DOCUMENT THE DIFFERENCES BETWEEN THE TWO??? YES, then what’s your damned problem.
If they are arguing about it, I’m pretty sure they did…
As for the rightness of monolithic packages, tell that to the KHTML developers who received patches from Apple’s WebCore in huge lumps. Size does matter when you want to evaluate the quality of a set of patches.
Even so, the end user DOESN’T CARE – why should Canonical.
Because they are basing their work on Debian Sid? Unless they want to do stabilisation work all by themselves, having better collaboration with the Debian folks will improve both camps.
You clearly do not understand the core of the issue.
And sure, “sour grapes”, but they’ll only be sour until we’ve had enough. And it’s not the alleged “stealing” that pisses many off, it’s more about “lost potential”.
o On the ubuntu.com front page, Debian is not mentioned at all; if you actually follow the About Ubuntu link, you can find Debian mentioned once in the first paragraph (albeit without a link), and if you did notice the right hand menu expanding, you may even decide to click on Ubuntu and Debian, where the relationship is finally illustrated from the Ubuntu point of view. To many, this is just not enough.
Update (+6h): the About Ubuntu page now links to the Debian homepage. Thanks!
Ubuntu is not working enough with the Debian community and not trying to give credit where credit is due? Right…
Update (+6h): the About Ubuntu page now links to the Debian homepage. Thanks!
Ubuntu is not working enough with the Debian community and not trying to give credit where credit is due? Right…
I had raised this issue several times during the past year. It took public exposure to cause a change…
I seem to remember that when Gentoo was born, it was marketed (wrong term, I know) as a base for other distributions to build derivatives on. But this never happened on the scale originally envisaged. Most users of Gentoo-based derivatives probably are… Gentoo users.
On the contrary, Debian ended up having few desktop users relatively to the several popular distributions built on top of it, while not having this idea in the first place (based on my recollections, at least)
This of course generates a number of problems, such as those described in this article.
Ubuntu at least is doing something. It can improve, agreed. And I, as a user, am comforted by the fact that Mark Shuttleworth seems to be willing to talk to the Debian community.
But what are various Linspire, Mepis (now Ubuntu based, but still…) and Xandros doing? In particular, Xandros, which was bragging just the other day about being profitable and fast growing… is Xandros hiring Debian maintainers and contributing patches in a sane way (I have never heard anything to this effect, but perhaps someone will be able to shed light on the subject)?
The GPL argument made earlier on this thread is not very good, as Debian has to be in good health to assure a prosperous future to its derivatives.
A more co-ordinated and responsible approach (and probably a stronger financial committment) by the derivatives might be necessary.
Ubuntu at least is doing something. It can improve, agreed. And I, as a user, am comforted by the fact that Mark Shuttleworth seems to be willing to talk to the Debian community.
But what are various Linspire, Mepis (now Ubuntu based, but still…) and Xandros doing? In particular, Xandros, which was bragging just the other day about being profitable and fast growing… is Xandros hiring Debian maintainers and contributing patches in a sane way (I have never heard anything to this effect, but perhaps someone will be able to shed light on the subject)?
This is a very good argument, and you are absolutely right that Mark/Canonical/Ubuntu stand out. Other distributions do contribute, but their impact is far lower, and their contributions are few, the cooperation could be improved. This is, btw, something which the Debian Derivers Council (in formation, still; see http://yukidoke.org/~mako/copyrighteous/pyblosxom.cgi/2004/Sep/24)
wants to address.
However, not everything that shines is gold. One of the main criticisms about Canonical/Ubuntu is that they claim a lot which just doesn’t square with reality. So, what’s better, someone who talks a lot and does less, or someone who talks less and does less? Ask that question from your perspective, but also try to answer it from the Debian perspective.
One of the main criticisms about Canonical/Ubuntu is that they claim a lot which just doesn’t square with reality. So, what’s better, someone who talks a lot and does less, or someone who talks less and does less?
I’ve heard that “criticism”. But I certainly have seen no evidence to support it.
Seriously this is ridiculous.
Even if Ubuntu is based on Debian, the two distributions are way too different to be “linked” to each other like this.
Don’t get me wrong but Debian is the home of some zealots (not in a pejorative way). The distribution mostly focus on stability and security. Thus said, Debian is irrelevant for most people that want a modern linux desktop. But is really doesn’t matter. Here comes Ubuntu.
At the end, everyone should be happy.
PS: Alot of people talk about patches being submitted etc. Well, I think this is a major problem unrelated to Debian, Unbutu. There are way too much patches floating around. This is crazy actually. Except distro-specific patches, I think all patches should be first submitted to the developers/maintainers of package X. I see too many patches everyday that dont get into original packages. It’s a nightmare for people maintaining some little and specific distributions.
“The distribution mostly focus on stability and security.’
And what is wrong with that?
“Thus said, Debian is irrelevant for most people that want a modern linux desktop.”
That is really unbelievable. And why? The installer is a lot easier than it used to be, testing is more bleeding edge than many other distros (and it soon overtakes Ubuntu as well)…
So I wonder what is wrong with Debian as a desktop distro.
That is really unbelievable. And why? The installer is a lot easier than it used to be, testing is more bleeding edge than many other distros (and it soon overtakes Ubuntu as well)…
So I wonder what is wrong with Debian as a desktop distro.
There’s nothing wrong with added stability and security. But most people would trade a bit of stability/security for bleeding edge packages.
Anyway, you seem to be biaised torward Debian 😉
There’s nothing wrong with added stability and security. But most people would trade a bit of stability/security for bleeding edge packages.
Well, one of my machines has been running (through several hardware changea and replacements) Debian sid for about 4 years. I never ever felt I don’t have the newest packages. I ran into package problems sometimes, but it wasn’t something I couldn’t resolve this way or the other. You could argue that the average user won’t do that, still, you said you wanted bleeding edge, and there’s no distro that gives you more bleeding edge _constantly_ than debian sid (besides gentoo ~x86/amd64 of course) and there certainly isn’t any other distro that gives you that level of bleeding edge with less hassle (for me it’s not, but I’m no fool to think it wouldn’t be for average users).
Anyway, you seem to be biaised torward Debian 😉
I note the smiley, still, that’s hardly an argument, it’s something one says when there’s nothing else he could come up with.
There’s nothing wrong with added stability and security. But most people would trade a bit of stability/security for bleeding edge packages.
http://backports.org
PS: Alot of people talk about patches being submitted etc. Well, I think this is a major problem unrelated to Debian, Unbutu. There are way too much patches floating around. This is crazy actually. Except distro-specific patches, I think all patches should be first submitted to the developers/maintainers of package X. I see too many patches everyday that dont get into original packages. It’s a nightmare for people maintaining some little and specific distributions.
This is pretty much exactly what Canonical is actually trying to address with Launchpad.
This is pretty much exactly what Canonical is actually trying to address with Launchpad.
The problem is that it’s not the right tool. You cannot force a large group of people to use a tool, however good it may be.
And you certainly can’t get Debian developers to come to rely on a proprietary tool.
It might come as a surprise to you but No Debian == No Ubuntu.
Having Used Ubuntu, I decided to switch back to Debian because Debian == Upstream and using ‘Debian Unstable’ there is no ‘release’ just a continuous stream of packages so there’s no need for me to dist-upgrade. That said. Ubuntu is also starting to push patches back to upstream.
There needs to be improvement from Ubuntu. With Ubuntu’s next release talking about taking drastic and rather neat new approaches to Linux (bootup and other things) I don’t see how Debian fits into that plan of ‘drastic/rad’ new changes. I’d like Ubuntu to let Debian participate in those discussions but right now, I don’t see that otherwise Ubuntu will maintain a fork of patches that Debian will/would not accept. (If someone from Debian / Ubuntu is reading this, correct me If I’m wrong).
Shawn.
Edited 2006-07-07 02:26
For years I have used Woody as my main desktop/laptop system. I liked it and the underlying Debian philosophy.
But as soon as Ubuntu came out I switched because I was tired of living with a lot of backports, which weren’t even uptodate (please remember that until one year ago Debian stable alias Woody came with Gnome 1.4…).
Ubuntu offered what I (and apparently many other users) wanted: A polished desktop on a stable foundation with uptodate software and a very reasonable release cycle.
It seems to me that the success of Ubuntu has had a good impact on Debian in the sense that it provided new ideas, showed what could be achieved in terms of a debian-based distro and with GNU/Linux at all. In many ways Ubuntu has taken over the leadership in the desktop GNU/Linux market – ahead of Redhat and Novel/Suse. To me this clearly shows the strength of the Debian platform.
I hope Ubuntu is seen as a positive challenge to Debian: “Look, this is what many users want – why can’t we do it something like this”. This is the type of challenge which makes OSS so successful. If Mozilla is too big and bloated – fine: Firefox comes a long, takes a chunk of the code, enhances it and becomes a big success. Same thing with the “competition” between KDE and Gnome. I believe we all have a better desktop because of such competition.
Therefore: I hope that both Debian and Ubuntu will continue their success. Debian is a great platform and I hope the two will find a way to make it work for both of them.
Thanks for your balanced post.
Ubuntu offered what I (and apparently many other users) wanted: A polished desktop on a stable foundation with uptodate software and a very reasonable release cycle.
Do note that I have never criticised this, nor do I personally see a problem along the lines of “Ubuntu is taking users away from Debian”. My blog was about the developer perspective, and about a great deal of lost potential.
By all means, the user should use what they want, and I agree that Ubuntu serves many much better than Debian, just as Debian serves many much better than any other distro.
It seems to me that the success of Ubuntu has had a good impact on Debian in the sense that it provided new ideas
Absolutely. And it stirred up many deadlocks. I had a paragraph pertaining to this in the orginal post, but my reviewers deemed it to be too “greasy”, meaning that they thought it was better to post criticisms without the mandatory first paragraph that starts out listing positive points before switching sides. I agreed in the end, but I don’t otherwise pass up the opportunity to get people to realise when and where Ubuntu has done good. As said, many thought I was a Canonical employee…
hope Ubuntu is seen as a positive challenge to Debian: “Look, this is what many users want – why can’t we do it something like this”.
The simple answer: because we define “users” in a much broader sense. To Ubuntu, it’s Desktop users. To us, users include the advanced system administrator team in charge of thousands of machines.
“The simple answer: because we define “users” in a much broader sense. To Ubuntu, it’s Desktop users. To us, users include the advanced system administrator team in charge of thousands of machines.”
Maybe you should take a look at Ubuntu Server and the plans for “stacks” (such as the “Install a LAMP server” option available in the current release of Ubuntu Server. I think that Debians main problem is that they insist on being everything to everyone and hence fail to shine in any single role except maybe platform independance where they are doing well for being a Linux distro (take a look at NetBSD to see my point).
If Debian longs for a huge userbase, they either need to reconcile with the fact that a bunch of specialized derivates WILL “steal” all the users or they need to put out multiple distros (Debian Desktop, Debian Server, Debian Supercomputer etc) based on what’s today known as Debian.
Given that Debian really haven’t got any big bucks behind it I don’t see the second option happening and hence we’re back to square one: living with it.
As stated earlier and elsewhere, the best Debian and Ubuntu can hope for is to streamline the way patches are handled. Given that Ubuntu is based on snapshots of Debians repositories, the patches may very well be outdated and obsolete when they are sent upstream.
Apart from sending a shitload of modular and specific patches, like you wrote in your article, I don’t see what more the Ubuntu devs can do to please the Debian project.
Well, I could see Ubuntu being made the official Debian Desktop (see above) but that will sadly not happen.
Maybe you should take a look at Ubuntu Server and the plans for “stacks” (such as the “Install a LAMP server” option available in the current release of Ubuntu Server. I think that Debians main problem is that they insist on being everything to everyone and hence fail to shine in any single role except maybe platform independance where they are doing well for being a Linux distro (take a look at NetBSD to see my point).
No “stack” or other wizardry functionality can replace full control given to an administrator, and that’s where Debian shines. Or do you know another distro that gives you full control and still scale to the most complex needs? That is Debian’s main selling point. And I’ll leave the Debian project if we ever deviate from that.
If Debian longs for a huge userbase,
We do, but only in the sense that we want feedback, not necessarily the users. We want to make Debian better (and not do whatever it takes to have every computer on earth run Debian), which is why we want stuff back from Ubuntu and others. Otherwise, we are going to run too close to becoming a supermarket (http://www.kitenet.net/~joey/blog/entry/the_supermarket_thing.html), which could potentially ruin the whole project.
Yes, this sounds far-fetched but it’s really not. Debian lives because it’s the best distro when it comes to integration of its components. If we let others speed away with their own integration without helping to consolidate all these efforts, then Debian will lose its purpose as a universal base for derivatives and other things. It will lose its pioneering position — Debian has a track record of guiding the distro world, one of the reasons that it’s always behind everyone else. But when you let the other distros be dogs you take for a walk (for the purpose of this comparison), running back and forth and showing their speed and agility to the world, Debian is the dog’s owner, walking steadily in one direction and never turning back. And it can only do so, because the dog’s owner has a broader goal than the dog.
So I’ve heard some people say that my post is “whiny”, but that was obviously not my intention. I am personally not at all threatened by Ubuntu, nor affected by it. I continue to earn my money with Debian and have never had a single client approach me about Ubuntu. I do Debian because Debian is exactly what I need, and because it’s a fascinating project. But I do see the potential in Ubuntu and all the potential that Debian can harness from it, and I do see fellow developers demotivated by Ubuntu, so that’s why I even bother…
Edited 2006-07-07 10:32
“We do, but only in the sense that we want feedback, not necessarily the users.”
And just how exactly are you going to get feedback without users??????
>> And just how exactly are you going to get feedback without users??????
Uhm, by uhm, by… dirty hippie idealism and design by comittee?
“Uhm, by uhm, by… dirty hippie idealism and design by comittee?”
Damn those pinko commie hippies!!
The simple answer: because we define “users” in a much broader sense. To Ubuntu, it’s Desktop users. To us, users include the advanced system administrator team in charge of thousands of machines.
Better be carefull.
1) targeting desktop users does not neccessarily makes Ubuntu less significant than Debian, if not more.
2) I don’t think Ubuntu will limit it’s target to only desktop users. From the progress we see so far, it seems Ubuntu will be focusing on the server side too.
Ubuntu will be focusing on the server side too
Thing is, with Debian you don’t have a “focus” in your sense. If you want a desktop os, Debian is good for it. If you want a powerful server os with all and every freedom to control every aspect of it, Debian is good for it. You want a mix, a desktop os with several server functionality, Debian is good for it. Once tried, the good feeling won’t easily go away. You’re a skilled user of admin, a base install and apt-get can get you a totally working system easily – one of my favourite things in Debian -, you want package managing apps, here you go pick one from the bag.
In my book, Debian is great especially because it doesn’t give you a desktop install/lamp install/server install/etc but it gives you a bag of tools and the largest package bag on this planet so that you are able to make your own system of out it.
Ubuntu is good, and with Ubuntu you also can achieve the above, but the average users won’t know anything about that, they just know that Ubuntu has a nice install and is less hassle for them to install. That’s good and ok, but won’t make it better than Debian, just differently targeted.
If you’d like to make your suggestions happen, then I’d suggest lobbying to have them included in the specs for the next release of Ubuntu:
https://launchpad.net/distros/ubuntu/edgy/+specs
In particular, you might want to contribute to this spec:
https://launchpad.net/distros/ubuntu/+spec/easier-motuing
Keep in mind also that the differences between Ubuntu and Debian would be a lot less if Debian Unstable could be stabilized and frozen for 6 months *within the Debian process*, because as was illustrated in the post you were responding to, Unstable is too unstable and Testing is too stale for many packages. Even if this happens, there still may be some differences (e.g. Ubuntu was ahead of Debian unstable in the X.org changes) but those would be a lot smaller and easier to manage. And as pointed out in my previous email, there are things Debian can do to help offshot distros be able to help Debian:
http://lists.debian.org/debian-project/2006/06/msg00296.html
IMO, it was eventually going to happen that Debian derivative was going to dominate, and from history, derivatives tended not to give anything back. In this sense, Debian is fortunate to have in Ubuntu a derivative distro that *wants* to give back and work with Debian. There’s a bit of a culture difference, but it’s not as if Debian hasn’t had its share of crises in the the past. It’s always come out stronger.
The point is, there are issues on both sides and it would benefit both Debian and Ubuntu if they’d talk through the issues.
In particular, you might want to contribute to this spec: