Red Hat, 16%. Canonical, 1%.

(Warning: ill-advised rant ahead.  Read the apology here.)

In case you missed it: $SUBJECT is the percentage of contribution to the GNOME codebase.  Thanks, Dave Neary.

An upside of not working for Red Hat anymore: I can speak frankly about this kind of issue, since no one really cares what I think anymore.  I’m just another cranky dude with a blog.

If you doubt, for a nanosecond, that Canonical is a marketing organization masquerading as an engineering organization, then you’re either an unapologetic Ubuntu fanboy or you’re not paying attention.

One of the most irritating things about working at Red Hat was watching Canonical take credit for code that Red Hat engineers wrote.  Of course, Red Hat engineers, being the upstanding sort of chaps that they are, never said a word about it, because they’ve always been too busy carrying the load — and it’s really never made sense for Red Hatters to complain much about it anyway, because it’s not the sort of discussion that ever benefits the complaining party.  “You’re just mad because Ubuntu’s cooler than you,” the masses would say, and to be fair, there’s always been something to that.

But this figure is absolutely egregious.

I mean, I always knew that Red Hat put in a lot of work into GNOME, because I saw it every day — but until now, I thought that Canonical *also* put a lot of work into GNOME.  They’ve certainly given the impression, over the last several years, of having put a lot of work into GNOME.  They’ve been very successful at positioning themselves as the Eternal Champion of the Linux Desktop, and positioning Red Hat as the boring old has-beens who long ago abandoned the Desktop fight, and just do backroom server work that Real Linux People don’t care about.

So let’s call it plain.  Canonical has been riding on Red Hat’s coattails for years — not just down in kernel land, but also, we now learn, all the way up to the tippy tippy top of user space.  Not only that, but they then have the gall to suggest that Red Hat should change its release schedules to make it even easier for them to ride the gravy train (while at the same time making the spectacularly outrageous claim that Red Hat is actually a proprietary software company — LOLWUT???)

They’ve done an exceptionally good job with this sleight of hand, but the facts are the facts, aren’t they?

In the ONE area where Canonical claims to have the MOST customer focus and the MOST engineering expertise, Red Hat still outproduces them ***16 TO 1***.

OMFG, SRSLY?  SIXTEEN TO ONE!!11!!11!!one!!!!!

If anyone at Canonical even bothers to respond to this analysis (which I doubt they will), I’m sure it’ll be the same old song-and-dance about how everyone collaborates, and everyone competes, and everyone wins, and the strength of the open source model, and not a fair comparison because Red Hat is so much bigger, and distro wars are bad, and can’t we all be friends, and yadda yadda yadda.

Yeah, yeah.  I know.  Spare me.  The world is full of talkers and doers, and in the long haul, people are usually smart enough to figure out which is which.  Which probably explains why Red Hat has a billion dollars of cash in the bank, while Canonical is still continually reinventing itself to make any profits at all.

Red Hat, 16%. Canonical, 1%.

327 thoughts on “Red Hat, 16%. Canonical, 1%.

    1. I think this is actually a legitimate excuse. Ubuntu drives bug reports that make their way into Gnome’s trackers, and I wonder what percentage of Red Hat’s contributions were fixes for bugs reported by Ubuntu users?

      In the end, marketing Linux is a huge, scary, and danger-fraught undertaking. I absolutely celebrate Canonical for what they do to get Linux on the desktop. Perhaps Red Hat produces more patches for Gnome but, from my perspective and the perspective of the less-technically inclined friends I’ve gotten into Linux via Ubuntu, they’re not doing much to bring new users to Linux/FOSS.

      1. Which is more important …

        Users or contributors?

        And with distros that come and go, which is more important …

        Users or GPL code?

        And finally, what is more important from a long-term, GPL producing standpoint …

        Marketing in the red or producing GPL code in the black?

        I personally think the stimga on Fedora is that Red Hat is involved, and people mis-quote from third-hand comments back in 2002-2003. Because I’m running more and more into people who are running Fedora and saying, “I cannot believe I didn’t use this simply because I didn’t try it.” Luckily the stigma has died over the last 7 years, and most naysayers are largely ignored.

        I’ve been a Debian maintainer. I deal with Ubuntu desktops for my friends regularly. And, frankly, have you done Fedora v. Ubuntu when it comes to upgrades? Sometimes people argue Debian APT-DPKG, and the strong Debian Packaging Guidelines, not realizing that Ubuntu can’t leverage them. I’ll take Anaconda Preupgrade (and even Fedora’s YUM-RPM repos) over Ubuntu’s APT-DPKG repos any day when it comes to upgrading.

        It’s great that people install Ubuntu regularly. But how many people stick with it, through several upgrades? I see a lot of long-term users on Fedora. Every now and then you hear about people leaving Fedora, trying another distro. A lot of times I hear about people leaving Ubuntu and going back to Windows. I’m not interested on number of new installs, but people running applications.

        Because committed users often become contributors.

      2. masternetra says:

        @Bryan J Smith

        Both are important. Users can help find bugs, can come up with ideas to better the system, and might even become contributors themselves at some point. Contributors of course are what do the actual coding, artwork, etc. While users aren’t as vital to the actual system without them the OS/project amounts to nothing more then a hobby that usually benefits few, the few of course being the contributors themselves. After all if no one uses your OS/project what’s the point of it other then expanding your own skill in programing/artistry/whatever?

      3. joebob2000 says:

        I have a feeling (no flame intended) that a huge portion (dare I say majority) of the bug reports that Red Hat pays attention to come from CentOS developers… I suspect random Ubuntu users, unless it’s by way of automatic crash submission, spend much time creating bug reports.

      4. Nicolas Mailhot says:

        > I think this is actually a legitimate excuse. Ubuntu drives bug reports that make their way into
        > Gnome’s trackers, and I wonder what percentage of Red Hat’s contributions were fixes for
        > bugs reported by Ubuntu users?

        Of course Red Hat fixes bugs reported by Ubuntu users. And Ubuntu users report a *lot* of bugs. But is quantity any important ?

        I’ve seen several big-impact GNOME bugs, where the initial report was done by a Fedora user (when the changes were fresh and the report most likely to be useful). And then, six months later, when it was too late to do any good, scores of *duplicates* from Debian/Ubuntu users, because that’s when the same versions were exposed to their users.

        So I doubt Ubuntu contribution bug-wise is what people make it. And the more they fork GNOME behaviour the less their bug reports are likely to be useful upstream. In fact, if the current drift continues, upstream reaction may end up the same as xorg developper reaction to nvidia bugs (like Ubuntu, Nvidia “contributes” a lot of bugs reports xorg side)

  1. Jef Spaleta says:

    Careful, people will start thinking you are me.

    just for the record. I think last time I was able to pull the numbers in terms of staffing Canonical was ~ 10 time smaller than Red Hat. So even on a per employee basis.. the ranking doesn’t change. I’ve seen Shuttleworth try to make that argument that pound for pound Canonical is more engaged or something. Though its hard to get accurate numbers of Canonical headcount since they don’t have to do any financial reporting.

    1. It would be nice for Ubuntu fanboys to be helping fund Canonical, and not with sub-$100 landscape subscriptions. If people really believe in Ubuntu, they should be helping fund Canonical. The problem is that fanbase is breeding a number that do not. And that’s sad.

      Funding, via a sustainable model, is what breeds GPL. Distributions can come and go, but the GPL code is perpetual. The goal, for the community, is to see a commercial venture sustainable so GPL code can be produced. That is what I has always been most important.

      I also think people forget that one of the reasons why Canonical has not gone the way of Mandrake/Mandriva is because of the charity of Shuttleworth. That’s also a reason I don’t badmouth Shuttleworth, because it has been his charity at work.

      Unfortunately I wish some fanboys would see that. That their lack of funding is why someone’s money goes to subsidize their “free software,” instead of another, charitable cause. Which is why they need to help contribute funds to sustaining the salaries of GPL developers.

      And stop leeching off others who do. I don’t care if it’s Canonical, Novell, Red Hat or otherwise. People need to realize those who are funded are the ones out there producing a lot of GPL software. I’m tired of the “$1B to Linux” IBM (when that was almost entirely used to mature their proprietary product line) and Ubutnu being the common assumptions on who develops Linux.

      Okay, I’m going to shut up now, before I’m asking Greg for a job. 😉

    2. brian says:

      I’m glad you brought that up Jef.

      Canonical is no where near the size of Red Hat primarily because they’ve not been as commercially oriented.

      Although that is changing now and with more revenue hopefully Canonical can hire more engineers who can contribute.

      I think its a bit petty to in-fight in the Linux community over something like this though.

      1. brian says:

        I’d also like to point out that Red Hat’s commercial work often is project oriented where RH’s customers “need” enhancements etc.

        There is a difference in contributing when you are paid to do it… it gets done and probably sooner than later.

        When you are just doing it for the greater good it usually gets done when there’s time to work on whatever it is and your day job doesn’t interfere.

  2. > In the ONE area where Canonical claims to have the MOST customer focus and the MOST > engineering expertise, Red Hat still outproduces them ***16 TO 1***.

    I don’t think Canonical does not produce anything. It’s just that they work for themselves, not with upstream. You just can’t measure Canonical’s work by its upstream contributions.

    Which is a pity, really.

  3. Dear Greg,

    I agree with you… 100%.

    But I’m scared of being narrow-minded, not thinking out-of-the-box, when I tend to “compare” RHEL/Fedora & Ubuntu.
    I mean, I have used Fedora and RHEL on hundreds of computers for years, now…
    I tried Ubuntu, SUSE and other distros, but not as intensively as Fedora/RHEL…
    Maybe I don’t have enough data to support my opinion, which is: Canonical IS a marketing company (I don’t think it’s trying to masquerade as an engineering one) and its “attacks” are way not cool. They’re Open Source angels? Angels don’t bash others and say: “We’re angels, we good, others’ bad”. Angels protect, work and stay true to their nature… Reminds me of RHEL engrs ;-)…

    I read blogs, I check out articles, tweet, and so forth.
    So, to all out there, please, give me articles where, out of nowhere, RHEL is attacking Canonical, just for fun, or worse: for FUD…
    I sure would like to be enlightened, should I be dwelling in darkness…


  4. Also on the size argument – Mandriva, a company smaller than Canonical, with a lot less money, and which everyone considers to be a KDE distro, still is right in their ballpark (and actually right on the same number, if Pavol’s right).

  5. @Jeff: I like life in the great beyond. They have great parties.

    @Aurelien: I wonder if you’re right. One would think that Canonical *must* do a lot more work than they push upstream. Which is, one can reasonably argue, an even *worse* position — because in the former case they can at least try to claim resource poverty, but in the latter case they must own up to the notion that they continue to fork any way they damned well please with little/no regard for the health of the ecosystem that sustains them: i.e. the very exemplar of the free rider problem.

  6. That’s why I think it’s bad when Red Hat states that desktop it isn’t an area of interest for them. That usually ends in “what are they contributing to Linux desktop?”, and “linux desktop isn’t ready” alike BS.

    I’d like to think it’s misunderstanding and not plain FUD, but how can be Red Hat not interested in desktop and at the same time be the first corporate contributor to Gnome? I don’t get it. Someone is doing it definitely wrong.

    PS: I can’t agree that Canonical it’s just marketing, they’re doing it great in the packaging/integration arena, and as distributors that’s success, isn’t it?

    1. Paul hits this on the nose.

      Michael Tiemann, who co-founded the first, profitable open source company, was misquoted and demonized many years ago. There are still people who are speaking about that third-hand. Too many times the media does not quote correctly. Tiemann is brilliant. Tiemann is who Red Hat put on the stand in Texas over IP issues. Tiemann has never deserved to be so demonized, and it doesn’t matter how many people like I have been out there, circling his original quotes, there are too many people who don’t care third-hand.

      Although anyone still misquoting Red Hat, since the start of 2008 with a CEO that now hits this head-on, is just not stopping to read anything. Red Hat-Fedora has been continually dedicated to the desktop, and Jim Whitehurst has been quoted, re-quoted, stating and re-stating that he doesn’t know how to make money on the home desktop. Guess what? Everyone else has now said that as well. Yes, even Shuttleworth. So what do people expect Red Hat to do?

      Here’s what I know. Red Hat has a sustainable business model, one that generates GPL. What I also know is that Red Hat employees work on Fedora and upstream as a paid jobs function, because they are tracking upstream patches, contributing their own, etc… It’s the reality that the cost of sustaining engineering — what companies are willing to pay money for — is what funds leading engineering “as a bonus.” Hence why so many changes on trailing edge get fed back to Fedora and upstream, just like the same developer is tracking Fedora and upstream for the next generation, trailing edge Red Hat commercial offering.

      It’s a win-win. The cost of sustaining engineering, and the real-world funding of it by companies that understand the costs of it, is what is causing Red Hat to write a lot of GPL. And it’s areas where companies need solutions, yes, very much desktop solutions as well, that also funds those areas. The enterprise desktop drives adoption more than people realize. It will be the home user that has to run because their company standardized on it, not the other way around. Remember, a lot of what people do at home is defined on what they need to work on at home.

      Just my long-standing exposure, and a byte of reality.

  7. @Juanjo: Agreed, Red Hat has clearly brought a lot of this upon themselves. Maybe this is an opportunity for someone @ RH to step up and say something *public* about their position re: the Linux desktop.

    1. @Juanjo: Red Hat’s executives have said the company prefers to concentrate on being a company that sells big-dollar software to smaller numbers of customers, rather than the other way around. But even to do that, you can’t have a crappy desktop, which is why Red Hat has a big team working hard on desktop technology, contributing to a lot of the cool stuff you see in most Linux distros, not to mention the core technologies for GNOME 3, which is going to be a major head-turner next year when it’s released.

      There’s a difference between “We’re not interested in the desktop” and “We’re not interested in selling a desktop product.” What Red Hat bigwigs have said is the latter, but when the headline gets written up as the former, and repeated on enough blogs, it doesn’t matter. It’s true that Red Hat doesn’t spend a lot of time (enough time, perhaps?) talking about all the things it does for the desktop. Some clueful souls pick up on the facts regardless, e.g., but that’s not a great substitute for taking a front seat in the conversation, so I’m pretty much agreeing with Greg here.

      Side point: I can’t speak for any of them, but I suspect people who work hard on desktop stuff at Red Hat would also like to see such a public statement. No matter what, though, I don’t see Red Hat’s strategy changing as far as continuing to contribute upstream in a truly open source way.

      (Disclosure: I do work for Red Hat but I don’t speak for the aforementioned bigwigs.)

      1. Paul, I see this conception in a different light, have a loot at – we don’t have there any feature that would excite a “normal” desktop user, we don’t market our desktop.
        In contrast, when attending the Ubuntu 10.4 release party I saw them talking about features as the new desktop theme (!!!), the social desktop, with the ability to post to twitter directly from the desktop and their new online music store.

      2. @nicu

        I have been saying that for ages: Red Hat needs to actually *market* their desktop offering (ie. Fedora). Canonical is eating their lunch in this regard and for no apparent reason as we’ve just seen…

    1. When it costs money to do so when that money could be paying GPL developer salaries. Seriously. I have seen this over and over.

      Word-of-mouth is the best way. I just sure wish the “take it for free, don’t pay a dime” fanboys would realize it’s about “free speech” and not “free beer.”

      As I’ve mentioned, they are hurting Canonical as much as everyone else.

  8. Fred says:

    On the occasion that I visit Planet Gnome, it appears that most of the Gnome developers are all using Ubuntu. Yes, I know from the facts that this is not the case.

    Also, when I have a Gnome issue (really any Fedora issue for that matter), and I google for an answer, I get many more Launchad results then say Red Hat bugzilla or Fedora forum results. (Google seams to think Fedora is Ubuntu)

    Now for appearances sake only, Ubuntu seams to be much more involved with their users and the Desktop than Fedora.

    Fedora does not know who it’s user base is, and the CEO of Red Hat said that they are not interested in the Desktop, Full stop. That should say it all right there.

    I use Fedora. Have been with Fedora/Red Hat for about 8 years. And I’ll continue to use it. But the “market’ forces of Ubuntu are pretty strong.

    1. @Fred: “…But the “market’ forces of Ubuntu are pretty strong.”
      Yep, that’s true. I see Ubuntu driven by a marketing force that is very successful and appealing to the general masses.

    2. “… the CEO of Red Hat said that they are not interested in the Desktop, Full stop. That should say it all right there.”

      What he said in fact (you can look it up) is that no one should be paying for a consumer desktop, it should be no-cost and supported by the FOSS community, and that we have an excellent solution for that consumer – Fedora.

      This is a follow-on to the idea that Red Hat isn’t trying to be in the business of selling a desktop distribution of Linux. Clearly Red Hat is very active in making the no-cost desktop a wonderful experience for consumers, and does so without a direct business interest there.

      1. Fedora had a shaky foundation initially as the first releases of Fedora Core weren’t very stable. I can honestly see Fedora 14 becoming a popular desktop for workstations but I don’t know that it is there yet for consumers. It is pretty darned close though. I think with a slight refocus on building support for proprietary drivers (WIFI, etc) and a little less stringency on tools like Mono it could quickly become a contender.

  9. RedHat is a billion dollar company, Canonical doesn’t make a profit yet. How can you expect Canonical to invest as much as RedHat? Or even a fraction?

      1. If there was such a thing as this mythical “Linux desktop” which is a lie as much as any other bs from “Linux fans” that either are unaware or conveniently ignore that Linux is a kernel project and not an operating system.

        We’re still waiting for _appropriate_ branding for this mess of projects all thrown together in any random way. But I would suggest it doesn’t need one. Ubuntu is an OS and it markets it’s self as one, Fedora is an OS and it fails to do anything but moan to other contributors.

        Is Canonical a marketing company, well no, because if they where they’d be doing a damn site better job at it. I’d also note that this figure is erroneous as a whole bunch of Canonical code is not being accepted upstream because it doesn’t fit the rather terribly designed gnome shell direction. I predict a fork of the top levels and perhaps we should call Ubuntu’s gnome something else.

      2. Luke says:

        Ya, there’s a marketing campaign for you:

        “World Motor Company: We’re the best car company in the world! Well, kinda. We’re low on funding, and while our cars are the nicest looking and the smoothest running, most of the work is done by Ford. But buy ours!”

        No. Ase a company looking to push a brand, they are going to say they’re the best, for some reason. And for many people, they are the best. They’re wrapping up the work of others, putting their own touches, patches, and toolsets in, and calling it their own. Which, in the GPL world, is absolutely LEGAL and ACCEPTABLE.

        Go Ubuntu.

        Thank you Red Hat.

      3. Ndubi says:

        I have yet to see a press release or official marketing document in which Canonical claims that Ubuntu is the Linux desktop, exactly where has a Canonical employee or official spokesman said that Ubuntu *is* the Linux desktop? What fanbois tout in forums and blogs is not the official company position. From what I see from Ubuntu’s site and marketing material, they make it clear that Ubuntu is a collection of open source (not a term I’m exceptionally fond of) packages built by the community.

        While we cannot dispute that RH out-contributes Canonical codewise, that does not make marketing any less important. Since we all agree that the Linux desktop is awesome, I would think that the thing it holding it back from taking over the world is not its wanting engineering, but focused and successful marketing. (Not that development should cease, just that it is important for someone else to focus on that aspect) I’m not saying that Canonical has succeeded in this area, just that they’ve taken a step in the right direction, attempting to make GNU/Linux as a desktop platform viable for the common user.

  10. RedHat and Canonical are two companies which both have their own ideas about what’s perfect for the desktop. RedHat is older and has been contributing much longer. It has also got more money to hire people to do things that don’t directly contribute to increasing the profit margin.
    Canonical on the other hand is younger. RedHat may be contributing so much simply because it is. It was contributing a lot before Canonical was as large as it is now and that means it has much more influence on GNOME, either via official channels, or via the large group of its employees that contribute to GNOME. This means that the ideas of RedHat — which, being focused on business, are far less demanding of usability than Canonical’s — have a lot more chance to be implemented in GNOME, or at least influence the design.

    That discourages Canonical from using just the plain GNOME stuff, because their, admittedly, demanding plans are hard to be implemented they do it their own way.
    They chose to implement their plans themselves, because they don’t feel they can get what they need to improve the usability.
    Canonical is doing pioneering work for FOSS in the design area, with their usability researches and the new approach used for design. That is very valuable, but not something you can do in the conservative GNOME community right away.

    Also, I think Canonical is focussing on a slightly different part of the stack. Companies like Collabora focus on technologies like GStreamer and Telepathy, fundamentals. RedHat seems to be working a lot on the core GUI part, whereas Canonical seems to be working a little bit higher on the stack, in the presentation area, providing applications and themes and other additions to the presentation of the desktop to the user, all on top of next to the core.

    Of course, these harsh rants don’t help making Canonical feel welcome in the GNOME community.

    1. Why do you think Greg is speaking in behalf of the Gnome community?

      Red Hat it’s the biggest single corporate contributor to Gnome. That’s a fact, that happens to be wrapped in a rant. Don’t focus in the rant part.

      1. I don’t think he is speaking on behalf of the GNOME community, I’m just saying that because RedHat is much more present in the community it has much more possibilities to influence (consciously and unconsciously, by the company, its employees) the direction the project is taken. It is harder for Canonical, with a smaller presence in the GNOME community, to do this, and therefore also less attractive to try to get things done in the community.
        Both companies have strong ideas, only they don’t have equal influence in the GNOME community.

        The fact tht RedHat is the largest contributor to GNOME is indeed a fact, but that doesn’t have to be communicated with the outer world via rants. The fact that this blog does do that means something about the way the author views things and therefore I think it is naturally to focus on the rant part.
        Also, personally I find it pathetic if you are so unsure about (the position of) your favourite Linux distribution that you have to beat another one in the ground whenever there is something in the news that says your distribution is better than others in some area(s).

  11. I always knew Red Hat / Fedora developers contributed more to GNOME than Canonical, I believe they contribute more to the Linux kernel as well. If memory serves correct, though, Debian developers contribute more to GNOME and the Linux kernel still.

    1. Neo says:


      Nonsense. Debian is not a high profile contributor. Red Hat is the leading organization for Linux kernel, GNOME and very high on Xorg, largest in Glibc etc

      1. I’m not saying Canonical does not do a good job. I just believe they’re in a different business. It’s wrong to blame them for not contributing to GNOME, they have no such obligation.

    1. Greg’s comparing GNOME contributions because that’s what we have numbers for. When much the same numbers were printed about the kernel, Canonical apologists used the excuse that Canonical’s a desktop focussed company and they’re not working on the kernel, they’re working on the desktop! Now we have numbers for the desktop Ubuntu uses by default, and it appears they contribute very little to that either. If there were numbers for contribution to every F/OSS project in the world, Canonical’s numbers would likely be the same or worse. But there aren’t.

  12. pht says:

    I probably missed something, but what is Canonical actually saying about their contribution to Gnome ? Is there a place where they say “We are contributing a lot to Gnome” ?

  13. Valise says:

    Hum, let’s compare to each company (official) revenue :

    * redhat : 16.1 / 748.23M = 0.0217847453323176 percent of contribution per million dollar.
    * canonical : 1.03 / 30M = 0.0343333333333333 percent of contribution per million dollar.

    So, compared to the financing power, canonical contributes more.

    1. 80 says:

      You’re forgetting that RH is also a major contributor to the kernel, Xorg (posting the ratio signal/noise of Canonical there would be insulting for them), core system (gcc, glibc, etc …). Main projets that Canonical is supporting are: Launchpad, bazaar (who uses it nowadays ?), upstart (which was dead for a while), Landscape (proprietary).

      Even compared to Mandriva which has 1/30 of Canonical financing power, Canonical lose.
      The matter is not how much Canonical contribute, but the ratio boasting “we are number one in desktop”/contribution to the free software desktop stack. And it’s pretty miserable.

  14. David Kitchen says:

    A UX change may only constitute a few lines, but in terms of what it does to improve the experience of users and therefore the adoption of GNOME it’s immense.

    I think it’s pretty short-sighted to believe that a contribution can be measured in lines of code or number of commits.

    Both contribute, in the way that best fits their organisation.

      1. True i can’t find any of the themes in debian repository or any upstream (i like some of the dark themes like the one in ubuntu-studio)
        and many changes are in some random ppa’s that i can’t quite install easily in debian
        also with the fonts is the same : try to install droid fonts in debian

  15. @Sense Hofstede: I can’t reply you in the thread.

    Three points:

    – I’m Ubuntu, Fedora and CentOS user, for different reasons, and I love them three. I guess you’re being defensive, but I didn’t write the post (I don’t I agree 100%).
    – Please, check and lean how Gnome works. Obviously anyone can influence a project with code, but I think (no offense) you’re wrong.
    – Calling pathetic to the other part makes communication difficult.

    Don’t waste your time!

  16. Thanks for the read. In my opinion, Canonical is a joke. Anyone that has ever looked at a change log knows that Canonical doesn’t contribute much of anything and what they do contribute is of little value. Canonical “developers” seem to be responsible for more more of the bugs and instability than any other contributors. They carry the stigma of being the welfare distribution, yet they take the credit for all of the innovation in the Linux world.

    I am thankful that RedHat is profitable, it is unfortunate that they as a company have to silently carry the weight of this deadbeat distro. It really isn’t silent as you might think though.

    1. zekopeko says:

      Wow, you really do hold a grudge.
      Look, just because Mark Shuttleworth bitch slapped you in one of the comments on an Ubuntu bug doesn’t mean that you should generalize all the efforts that they make as a developer community with hundreds of members just because the head honcho called you out.

      1. There is no grudge zeko, however since your stupid comment remains intact I am compelled to reply (again).

        My distaste for Ubuntu began long before I ever exchanged words with Mark if you would spend 5 minutes reading my blog you would quickly find out why. My opinion is based on years of use of the platform, thousands of unfixed bugs in launchpad, as well as reviewing change logs for many core software products. Mark’s immaturity just solidified something that was already there. Grow up, and get over yourself.

    2. the_madman says:

      Canonical focusses on what they specialise in, which is user experience. It’s why the biggest changes they’ve made are basically the theme and default panel applets. But they also work on other areas that their current target market (Windows/MacOS users) are interested in, such as the Ubuntu One Music Store. Some changes, like the Ubuntu One cloud service and Ubuntu One Music Store, just aren’t sensible to send upstream (they depend on proprietary technology). Others, like themes and icons, may be too objective and based on opinion to belong upstream.

      Mostly, though, I think Canonical takes existing technology and builds on top of it. You know, like all those Android phones out there. Quite frankly, I don’t see the problem.

      1. Fair enough. If they only took credit for these things it would probably be a non issue for just about everyone. Wouldn’t you agree though that they take a lot of credit that they don’t deserve?

      2. guybrush says:

        Canonical works on the little bits they can afford to work on. Otherwise, they’d focus on more relevant aspects of user experience than a theme or where window buttons go.

        Maybe that’s why they’ve accomplished basically nothing in their 6 years of existence.

  17. Chris says:

    You can bitch about the supposed 16-to-1 contribution ratio all you like. If you’re someone who wants Linux to be successful, popular, and recognise that Canonical have done a great job in making Ubuntu popular, then you can’t deny they are doing *something* right. That ‘something’ may be technical (desktop customisation), non-technical (market) or a mixture of the two; I just don’t see how it’s helpful to the wider cause of making Linux popular, in this instance, to start drawing a comparison of who wrote the most code. Am I missing the point?

    1. Chris: “recognise that Canonical have done a great job in making Ubuntu popular”

      well, that rather depends, doesn’t it? They’ve succeeded in making it more popular than other Linux distributions, yes. Have they succeeded in making it any more popular, relatively speaking, with the general population than any other top distro did before them? Arguably, no. Measuring operating system installed base is notoriously tricky, especially when the numbers for Linux are basically the same as the margin of error, but I haven’t seen anyone even *suggest* that Linux in general is significantly more popular in the context of all computer use than it was before Ubuntu came along. They can’t use the excuse of being new any more, they’ve been the top distro, by consensus, for about five years now. What’s holding them up? Why isn’t the world using Linux? To summarize, why have they had bug #1 open for nearly six years, without making any visible progress on it? Is that really success?

      1. guybrush says:

        In short? Because they might admire OS X, but they’re not Apple. Surprise, surprise, giving away CDs is not a viable business model.

      2. Suman Alexandar says:

        What nonsense. So what if they have or haven’t succeeded for 6 years? What have you done of merit to judge others this way? Have you no decency?

  18. Raghavendra says:

    Nice discussion. People in linux circles are always aware that RH has much contribution to core libraries like glibc and to linux kernel, hence the gnome thing may not be a big surprise. It is just the external media and populace unaware of this fact. This is where marketing comes into play. Canonical marketed well by methods like free cd/dvd distribution among many. RedHat has also done but that has been in more enterprise circles.

    Another reason why people may not have used RH for desktop (again in non enterprise environs) is that RH desktop has not been as bleeding edge as others, again this is/was taken care of by Fedora. Again due to lesser marketing, not many are aware of Fedora-RH connection.

    Lastly, even though there may be differences here and there, all companies should strive towards better FOSS ecosystem, there are those outside FOSS who can take advantage of this fact.

  19. Of course Canonical is doing *something* right.

    That something is called “marketing”. And hey, that’s fine. There’s certainly a place for that, and shame on Red Hat for not doing a better job of speaking more directly to this issue in the past.

    But what Canonical is still *not* doing right — IN MY HUMBLE OPINION AS A NON-REDHATTER — is working with upstream communities collaboratively. We’ve seen this behavior for *years*. And it’s never been more obvious than it is right now.

    1. See above. Honestly, I’m not even sure that Canonical’s terribly good at marketing. I may write a (careful) blog post on the same topic I’ve been going off on here, but short version – let’s face it, owning the Linux desktop ‘market’ is not terribly hard. The competition’s not the greatest in the world. No-one’s really counted SUSE as significant since Novell bought it (not sure why, that’s a separate topic in itself), Mandriva’s basically bankrupt (and if you can dredge up someone able to write a press release in English with no grammatical errors you instantly appear more professional than Mandriva…), Fedora is really not aimed at being a for-Uncle-Charlie desktop, neither’s Debian, and that’s basically all the main players. Show up with a bit of money and you can win pretty much by default, and that’s all Ubuntu did. For them to really be *successful* even in their own terms – and that includes being ‘good at marketing’ – they need to demonstrate significant success outside the kiddy pool (the existing Linux ‘market’). Have they done that? Really?

      1. Numbers – .

        This is the OS share metric traditionally most kind to Linux (the overall average of a bunch of different sites’ metrics that Wikipedia tracks is around 1.3% right now).

        When Ubuntu bug #1 was filed in August 2004, Linux was at 3.1% share.

        After nearly six years, that has been driven up to…4.8%.

        Note that from March 2003 to August 2004, the number went up from 2.2% to 3.1%. So it doesn’t even appear that Ubuntu being the ‘leading’ distribution has increased Linux’s rate of growth (though that’s a short range to draw many conclusions from, and these are still all very low numbers and hence very susceptible to error).

        Is this success?

  20. Ian T. says:

    You should be glad that Red Hat does not get blamed for the crappy code that Canonical devs do write. The Ubuntu folks have a habit of re-arranging the furniture and making gratuitious undocumented changes. As a result we get shiny new bugs and irritated users. Meanwhile, things that really need attention get ignored, and if you complain you are told you are being negative.

  21. Simon says:

    The original data was looking at commits over the past two years

    So let’s look at ratios of employee number 2 years ago (the ratio at the beginning of the period should be more representative of what they’ve contributed over that period than at the end; both companies have hired a lot over the past year or two).

    So: Red Hat employees 2 years ago: 2200. Canonical employees 2 years ago: 130. (Source: WP history for both).

    That’s a ratio of about… ummm… Just over 16 to 1! What a coincidence.

    I suspect you already knew this perfectly well, given your pre-emptive attempt at discrediting of precisely this criticism of your attack (“…I’m sure it’ll be the same old song-and-dance about … not a fair comparison because Red Hat is so much bigger…”). It is self-evidently true that less people contribute less, and so that contributions per employee at the time of contribution, and not total contributions, is the relevant metric. So I’m not really sure what to make of the fact that you chose the latter one anyway, and tried to pre-emptively discredit the clearly valid criticism of this.

      1. @Zekopeko no, because Canonical doesn’t have nearly the diverse range of products that Red Hat does. Their portfolio is much slimmer so you’d expect a much higher percentage of their engineering staff to be devoted to GNOME than Red Hat’s.

  22. 70.27% of contributions are done “in my spare time” which leads to 23.45% of commits. I wonder how many of these can be attributed to Canonical’s marketing of the Ubuntu desktop.

    1. Canonical doesn’t market the Ubuntu desktop to contributors, it markets Ubuntu to users, so I’d venture to say (at a VERY high random estimate), at most a quarter. Unlike Red Hat and Fedora, who actively cultivate their users into contributors (I was brought into the fold by KDE SIG contributors suggesting I package some software I was missing), Canonical and Ubuntu are more interested in marketing a product rather than a community

      1. “[My impression is that] Red Hat and Fedora […] actively cultivate their users into contributors, [whereas] Canonical and Ubuntu are more interested in marketing a product rather than a community”

        Now 100% free of made-up “facts”!

        I mean, unless you’re a sociologist who is conducting a statistical survey of user/contributor turnover in the global Linux community… which I kinda assume you aren’t.

  23. @Simon, nice try.

    What you fail to take into account in your analysis is that Red Hat is also CARRYING THE REST OF THE LINUX STACK TOO, while Canonical is doing ALMOST NONE OF THAT.

    And even if this number were true, let’s look at the contributions of individual developers, because surely, if your aggregate-production-per-employee argument were correct, then there would be one or two superstars at Canonical helping to drive the GNOME bus, right? I would think, given the breathless praise of Canonical’s innovations to the GNOME stack, that they would have one or two disproportionately strong contributors — and that Red Hat, as the old Desktop fuddy duddy, would be ramping its desktop involvement down.

    So let’s look at the top 10 most productive individual contributors. Hmm, 5 from Red Hat, zero from Canonical.

    OK, let’s look at the next 10. Hmm, 5.5 from Red Hat, zero from Canonical.

    Oh, there’s Rodrigo Moya, at 29th. Hello!

    Seriously, Canonical? You tout yourself as THE LINUX DESKTOP LEADER, but YOUR LEADING GNOME DEVELOPER IS 29TH IN COMMITS???

    @Alan: your point about “spare time contributions” is potentially a fair one, and I never said that Canonical wasn’t good at marketing. Does that marketing account for 23% of commits? Bring some data to back it up. If it’s anywhere near as compelling as this data, it would be worth a listen.

    1. Nicolas Mailhot says:

      And let’s not forget that Red Hat has a *huge* investment JBoss-side, so a lot of Red Hat employees are doing free software in areas people do not count as “Linux” at all

      (and before someone claims Red Hat is diverting money from Linux, a lot of the revenues people like to quote come from the JBoss branch too)

  24. Peter says:

    Totally agree Greg. Canonical is definitely riding the gravy train and Shuttleworth’s appeal to sync releases was just a blatant attempt to grab even more innovation from others without investing a dime themselves (except in their self-serving marketing schemes off course). One way to make sure that credit is given were credit is due is by making sure Fedora releases are scheduled later than Canonical’s. And if you have your “time to set things straight” hat on then one could even consider to keep some of the new upcoming features in the next Fedora release off their radar until Ubuntu is too far into their dev cycle to integrate these features that are developed and paid for by Red Hat and others. After all Canonical has a history of keeping things closed themselves. Once Ubuntu starts to lag in features it will become quite obvious that Canonical is just leeching and not giving anything back. Let’s see if the user opts for the latest and greatest features in Fedora or still buy Canonical’s marketing and settle for yesterday’s news.

    On the other hand maybe one should accept that there will unfortunately always be takers who are not givers too. It’s obvious that it is up to Red Hat and the Fedora community to improve the marketing of Fedora.

  25. Rob says:

    Used Red Hat exclusively starting with about version 4. Was responsible for putting it into my enterprise (first linux machines in the data center). Dropped it like a bad habit when they dropped desktop support (continued to pay for server support) and started the fedora project, and now use ubuntu. Even ubuntu makes me install or compile wireless drivers for nearly every machine i’ve ever tried, and download a bunch of codecs to watch movies. They have made it less painful, to their (or someone’s) credit. If you want to know why Linux will never make it on the desktop, keep in mind that wireless networking support and multimedia support still largely suck (though they are miles better than they were when I was compiling wireless drivers in my graduate advisor’s office back in the late 90s).

    I don’t know if people at Red Hat understand the massive loss of confidence that happened when they discontinued RHL. Pretty much everyone I knew dropped them and moved onto other distros. A lot of people were really happy Canonical stepped up and provided a way off of RHEL. The sentiment was “well who knows what they’ll drop next.” I’m happy to give credit to Red Hat for producing a good product and providing good service to the community, but on the face of things they turned their back on a lot of people a long time ago.

    On a personal note, I don’t really care how much code they contribute to the gnome project, I’d really rather they nut up and fix the real problems like wireless and multimedia support that seem to always require some work. But oh, yeah, they aren’t interested in desktop linux. I forgot.

    1. Red Hat employs the guy who maintains the kernel wireless stack (John Linville) and the lead developer of NetworkManager (Dan Williams). There really isn’t any technical problem with Linux’s multimedia support, there is a legal problem with how much of it any distribution backed by a company with money can safely distribute, and that isn’t a problem Red Hat can possibly solve (besides succeeding in our lobbying attempts against software patents, which is tricky when the other side has far, far, far more money).

      1. Rob says:

        I’ll admit it has been a while since I tried Fedora, but if Canonical is taking advantage of their efforts then John and Dan sadly still have some work to do. I appreciate that there are some proprietary, closed drivers involved that complicate things but I really haven’t ever had a linux machine work on wireless out of the box. Ever. Neither has anyone in my social circle. The latest ubuntu at least prompted me to download the proprietary drivers, which was the least painful it has ever been (though they gave me two choices, and the one listed as ‘open source’ didn’t work). If John and Dan deserve credit for that then I’ll happily buy the beer, but I’d rather buy a case of 12-year-old scotch for someone who can just make wireless work reliably, consistently, out of the box.

    2. George Mitchell says:

      Rob, EVERYTHING you are seemingly blaming Red Hat for (wireless drivers/multimedia/etc) are IP issues, not technical issues. When hardware companies REFUSE to provide drivers for wireless, how is Red Hat supposed to fix that? Either an open source driver is there or it isn’t. Going around that with a wrapper gets into a shady IP area that large companies don’t like to risk. Multimedia is all about IP. If Red Hat ships you codecs, who is going to pay for them? Its really a whole separate business. If you (or anyone else) wants multimedia on Fedora, install Fluendo’s products.

      The Red Hat shift to RHEL happened because it was the only way that they could grow their business into a serious and sustainable enterprise. The same thing has happened with SuSE under Novell. A lot of free software types don’t like this, but it is these “impure” companies like Red Hat and Novell that are enabling desktop Linux to move forward by plowing the money generated back into opensource development which is the whole point of the above post in the first place. Whenever you appreciate CentOS, think of the Red Hat guys whole make CentOS possible!

      1. Interestingly: there is a company working hard to solve the multimedia problem. That company is Fluendo, and despite being an order of magnitude smaller than Canonical, they have twice as many contributions to GNOME as Canonical does. Another example of a small organization that tries to solve the hard problems the right way.

      2. Rob says:

        Be clear that I’m not blaming Red Hat for these problems, but I think if they want to be desktop Linux heroes their time would be better spent on these large, seemingly intractable problems than on whatever new icon set Gnome has in mind.

        I know I’m coming off as the bad guy here, but which would you rather read about:

        “Red Hat contributes 16% of Gnome source code, enabling 8% better performance on system login!”
        “Red Hat solves open-source multimedia codec problem, enabling people to finally watch video on the same day they install fedora!”

        It seems so obvious that the reason mere mortals don’t like linux is that the basic things they expect from a computer (networking, multimedia) are a pain in the ass. I don’t know who gets the credit, but my latest ubuntu install at least made an effort at trying to automate getting these things going for me.

      3. George Mitchell says:

        So Rob, how do you suggest that RH solve this codec problem. Ubuntu’s answer was simple. They just include the codecs in open violation of the law. It makes it all so easy when your operation is located in a country that doesn’t enforce these things and so you just do it. RH doesn’t have that option. But Ubuntu comes out as a hero in the eyes of people who want a given product and don’t want to have to pay a dime for it. Probably the same kind of people who don’t ask questions when they are offered bargains by shady characters on the street. The kind of people who don’t really care unless its their home that was burgled. The bad guy ends up being the corner merchant who isn’t finding a way to effectively compete with the street characters. That really sounds like what your are trying to say here and I think that this whole something for nothing syndrome is what this whole discussion is all about.

      4. Tim says:

        Ubuntu isn’t violating anything, US law simply does not apply to the rest of the world. I feel genuinely sorry for Red Hat being situated there and having to put up with abusive patent laws, but unfortunately this is entirely their problem.

      5. Rob says:

        Ah, George, thanks for trying to criminalize me for wanting to watch a movie.

        Can you point me to the place where I would buy a license to use the appropriate codecs on Linux? A web page where I could put in my credit card number, get a license, and download a codec package? Well, perhaps that would be a good start. I don’t think it is beyond any commercial Linux company who really cares about the desktop to pull off something like that. No one has tried, that I know of.

        Something for nothing? Perhaps, but given the amount of money I spend at amazon, netflix, and itunes its sort of disingenuous to think I wouldn’t be willing to pay if someone could make it easy for me. The real problem is open-source zealots sit up on their high horse and say “we won’t include these things for philosophical reasons” – well, then desktop linux won’t ever be successful, and when you wonder why, take a look at your philosophy. You don’t exist in a vacuum. Don’t begrudge people their features.

        Surely you have watched a movie on your linux box, or used some non-free driver at some point in your life? If not, I applaud you, but most of the folks I know want their computer to do more than provide a terminal window and Firefox.

      6. George Mitchell says:

        Rob, Legal codecs have been available for a long time from Fluendo ( It is just that most users would rather get them for free than pay for them. Ubuntu could pay for them and offer a supported version for a reasonable price just like Fluendo does. Why don’t they? That is the whole issue here. Other distros like Mandriva actually refers their US users to Fluendo for legal multimedia support.

    3. “I don’t know if people at Red Hat understand the massive loss of confidence that happened when they discontinued RHL.”

      Yeah, people know. Those of us who know make sure that new people know. It’s the dark part of the story of “RHL to RHEL and Fedora” that we tell. It was a bad screw up and the improvements in Fedora for contributors and users over the years are a direct result of knowing all about that loss of confidence.

      In fact, that set of lessons is part of the experiences that fueled this chapter from ‘The Open Source Way’.

      “… they aren’t interested in desktop linux. I forgot.”

      That is clearly a repetition of an inaccuracy. About six years ago former Red Hat CEO Matthew Szulik said some off-hand remarks about not focusing on the desktop, and that was the start of the misinformation. Current CEO Jim Whitehurst said, “Consumers should not pay for a desktop, it should be no-cost and come with community support. We have that, it’s called Fedora.” (Not a direct quote, but more accurate than the misinformation.)

      The facts are, Red Hat does care about the desktop, but doesn’t care to charge for it or wrap special services around it to charge for.

      1. Rob says:

        I will admit that I should give Fedora another try, and I appreciate that companies exist that will give back to a larger cause.

        The IP question is tough. I appreciate that codecs cost money and drivers aren’t always available. It is a massive undertaking to update drivers in particular. Red Hat doesn’t want to charge for a distro. I get it.

        Ubuntu prompted me to install proprietary wireless drivers that it downloaded and started for me. This isn’t perfect, but it is a start.

        Ubuntu prompted me to install a package of proprietary codecs for video and audio. This isn’t perfect (I’m not sure if I’m legally allowed to be using them or not, it wasn’t clear) but it is a start.

        Does Fedora do these things? Its been too long since I used it to know. Should I give credit to the Fedora project because Ubuntu provides these services? I don’t know. I will say I was sure happy when I was prompted to download those items instead of having to track them down and compile them myself. I am pretty sure I told others about it being more convenient to get my machine working this time around.

        I’m the kind of guy who knows to keep a USB wireless adapter around that I know runs off the linux kernel drivers, in case my built-in card doesn’t come up. I’m not the guy you need to be selling desktop linux to. Its the guy who doesn’t know who needs the help. If you really care about the desktop, you’ll think of him.

      2. It’s not just a question of money. The fact that the patent holders want to license the patents in such a restrictive way *in itself* makes them utterly incompatible with free software. Restrictive patent licenses simply cannot work with free software, since the very definition of free software is that it is freely redistributable; any patent license involved has to be at least that liberal. It’s not just that it would cost $A_LOT_OF_MONEY to license all those codecs, it’s that we can’t license them in a way which is compatible with free software. We just cannot, if the owners don’t want to do it.

      3. And no, Fedora does not prompt you to install proprietary software, because we do not think that’s the right thing to do. The FSF has several excellent reasons why.

      4. Rob: you have to realize there’s a difference between proprietary and patent-infringing software. When I say ‘proprietary software’ I am talking about NVIDIA and ATI graphics drivers and a few wireless drivers. I’m not talking about multimedia codecs. Multimedia codecs that are not available in Fedora are not available because they infringe on software patents. They are freely licensed, but due to the patent issue, not effectively freely redistributable in jurisdictions where patents have been issued, which includes the U.S.

        Proprietary software is a discretionary issue; there’s no legal problem with using NVIDIA or ATI drivers, we (Fedora) simply agree with FSF that supporting such software is not the right way forward. Patent-infringing software is not a discretionary issue. The advice Fedora has from its lawyers is that it is not legally safe to ship such software or even to help guide people to install it. We have no choice in this matter; if we provide such packages we can, and probably will, get our asses sued off.

        Canonical is working off more liberal legal advice, likely because of where the company is based. But as a user, there’s nothing discretionary about this issue for you either. If you live in the United States and you use patent-infringing software without a license, you *are* breaking the law. It’s not Fedora which is criminalizing you, it’s, well, the law (although I believe it’s a civil rather than a criminal matter, technically speaking).

        You can buy licensed versions of several codecs from the Fluendo store, BTW – .

      5. Rob says:

        Thanks for pointing me to Fluendo. I saw it mentioned in this thread but didn’t think to go check it out. I was unable to find a cut-and-dried statement that buying a codec license from a spanish company would put me in the clear in the US, but it is good to know someone is trying to address the problem. Not perfect, but it’s a start. I will likely buy something in order to at least have evidence that I tried to comply with licensing. I will say that it is hard to determine the difference between fluendo selling me a codec and Canonical licensing H.264 on my behalf, since they are both outside the US. Also, most DVD drives come with a license to MPEG2, but I wonder what the legal status is if you’re not using the bundled software? I just don’t know the answers to these questions so I take the codecs offered to me, and I’ll admit the legal status is questionable but I don’t have the energy to sort through it all, I just want to watch a movie. Someone interested in promoting Linux on the desktop might attempt to make some sense out of all this for the community’s sake. Canonical has done a good job in letting me just watch the movie, so I give them credit for that.

        As to proprietary drivers, it is simply a ridiculous position to not support that hardware because you don’t control the IP to the drivers. You think you’re taking a stand against the people who own that IP, but you’re not. You’re simply leaving users out in the cold. I suspect you recognize this at some level. Canonical certainly does, so they try to make those drivers easy to find even though they refuse to ship with them. They have found a compromise. It is not an ideal user experience, but it demonstrates that they have at least thought about their users.

        You say there are no legal problems, so you should also admit that you’re sacrificing your user experience in order to make an obtuse point to a bunch of hardware manufacturers. It’s a chicken-and-egg problem. NVIDIA isn’t going to care about your 4% market share, and your market share can’t grow significantly until you start providing a decent user experience, so which is it going to be?

      6. Rob: Canonical hasn’t licensed h264 on your behalf. The license Canonical purchased does not apply to people who just download Ubuntu and run it on their own systems, it related to a specific OEM deal. Canonical couldn’t afford to license h264 on your behalf, because it would have to pay for a license for every person on earth, since the license on the regular Ubuntu edition allows you to give a copy to anyone you darn well like. 🙂

        I don’t know whether having an MPEG-2 patent license through your DVD player or whatever covers using a different implementation on your computer either, I’m not a patent lawyer =). I suspect technically speaking the answer is no, but practically speaking, no-one is ever going to sue end users for patent infringement, it makes no economic sense. That doesn’t help the position of a US-based distributor with actual money, though; it is simply unavoidably legally unsafe for us to provide the codecs themselves or guide people to install them.

        The problem with your point about hardware support is that it’s demonstrably falsifiable. Working with manufacturers to develop open drivers works. Red Hat has been doing it for years. We’ve done it with network card vendors, wireless vendors, and graphics vendors; that’s why we finally have specs for ATI/AMD cards to write decent open drivers for them.

        Saying ‘oh, we’ll just ship the proprietary driver so people’s systems work’ is an attractive short-term solution, but ultimately we don’t believe it’s the right one. Take that position to its logical extreme and GNU/Linux would never have existed in the first place, because perfectly decent proprietary alternatives were available at the time. Where do you stop sacrificing freedom for the sake of convenience? You draw the line very close to ‘freedom’ or you don’t draw it at all, really.

        I can see both sides of the argument – I used to work for Mandriva, who ship the proprietary driver pre-installed in two of their editions. But over the long term, I do think RH’s position is the best. Especially after I’ve seen AMD/ATI unilaterally drop support for old cards from their proprietary drivers, and be very slow to update them for new releases, and seen NVIDIA say they won’t support X server 1.8 with their oldest proprietary branch (so no proprietary driver for FX 5xxx series cards or older with Fedora 13 or Ubuntu 10.10). We just don’t want to be hogtied to those kinds of proprietary driver vendor decisions. It’s not like we sit around waving our hands and doing nothing, either; who’s writing the radeon and nouveau drivers? Red Hat staff. Over the long term, the benefit to the F/OSS community of working free drivers for those adapters is going to be much greater than the benefit of packaging a proprietary vendor driver for them.

      7. George Mitchell says:

        Rob, Fluendo is openly selling those licensed codecs worldwide. Those are the kinds of issues that IP owners keep careful watch over. I suspect they have to report back to the IP owners how many copies went out and where to. Believe me, if this were not allowed, their license would be revoked immediately and you would be reading about a lawsuit against them. Why would a licensed Spanish company be selling their wares to the US via an English language web page if the wares were not legal. It just doesn’t make sense that they would. When they say their stuff is legal anywhere in the world, that is exactly what they mean. You buy the Fluendo kit and you are covered and it ties right in with g-streamer and makes stuff work just like Windows … painlessly. The video below helps to explain all of this:

  26. George Mitchell says:

    I have been a Mandrake/Mandriva user since 1999. As such, I deeply appreciate all the hard work that Red Hat has put into making all of this possible. There is no doubt among those of us who have been using desktop Linux for years that companies like Red Hat and Novell have done most of the heavy lifting. And now, of course, Oracle is putting a lot of money into btrfs which is going to outperform ZFS when it is out the door (see the latest Phoronix testing if you don’t believe that). The Canonical model reminds me of that of Lindows/Linspire. They like to be very high profile and then when the money runs out it all proves to be a flash in the pan. There major claim to fame is that they have such deep pockets that how they manage their business really doesn’t seem to matter a whole lot UNTIL the money runs out. In the mean time they live in a dreamworld. The reality is that it is tough being a “Linux company” (or even a non-Linux company, like Novell or Oracle, that increasingly invests in Linux), and Red Hat and Novell are shining examples of how it can be done successfully and with integrity. And its not just limited to development. Factor in all the resources expended on the SCO fiasco to which Canonical and others have contributed not a dime. So hats off to Red Hat.

    1. Jason says:

      Factor in all the resources expended on the SCO fiasco to which Canonical and others have contributed not a dime. So hats off to Red Hat.

      Not to knock Red Hat here, but how much did Red Hat actually contribute financially (or otherwise) to fighting off SCO? I know Novell contributed a lot (it was their lawyers who won, afterall), but I didn’t think Red Hat contributed to that. Maybe I’m wrong…?

      I just get the feeling you are giving credit to Red Hat for something Novell did.

      That said, my hat is also off to Red Hat for their history of string Linux contributions.

      1. Jared says:

        Red Hat has successfully defended several patent litigation trials (some alongside Novell), and is currently fighting Red Hat vs Bedrock. Check out Groklaw and a few other sites for more info.

        RH has definately pulled it’s weight in the ring of the justice system.

  27. Perhaps I don’t know all the facts, perhaps I’m wrong, but I think I know one reason why this might be this way.

    Red Hat usually stays as a bleeding edge distro, Debian (inserted here only for this explanation) usually is a little bit outdated in favor of greater stability. Ubuntu is based on a testing Debian branch, and is thus a middle ground between stability and Bleeding edge. Remember that in Ubuntu some packages might introduce changes that disrupt it’s flow and some are not the latest version but work flawlessly.

    So, by staying on the bleeding edge, Red Hat needs to work out a lot of bugs in newer software releases to be able to release a good distro, while Ubuntu doesn’t need to do all that, and focuses on usability changes and configuration.

    I think we are all grateful for the wonderful job Red Hat does in that bug hunt, and yes, a lot of other distros benefit from that, but it all comes down to position in the life cicle of software development as distro.

    1. The argument isn’t ‘Red Hat contributes much more code therefore Red Hat’s distros must be better for a user’, the argument is that any company claiming to be a key part of the F/OSS ecosystem ought to be contributing to it properly.

  28. whocarse says:

    The whole point of a contribution is to not care where it came from or speak ill of people who don’t contribute as much. You give what you can when you can. In defense of ubuntu, they provide a great product that will always be free of charge. You cannot deny they put together one great desktop. And finally, I definitely give respect to redhat and I would suggest it in the server room to companies any day.

    1. Dulwithe says:

      I deny that they put together a great desktop. And many others do, too. And it is not much “they” who put it together, so much as “they” who have repackaged stuff that has already been put together by many others.

      But who can argue sense into an ubuntu apologist…??

  29. bjr says:

    Redhat should probably say that they aren’t interested in the Consumer market rather than say they aren’t interested in the Desktop market. The Enterprise market requires a GUI to support the enterprise environment and that GUI looks exactly like a consumer desktop but it’s purpose is to support the whole, not as an end in itself. Mathematically there can’t be a viable consumer desktop market. Prices in the consumer space are so low that only a company with Microsoft’s volume (i.e. hundreds of milliions of copies/year) can make money. Do the math, at $30 a copy you would have to sell 24 million copies a year to generate Redhat’s revenue, that can’t possible happen. To make any money at all your support costs per user would have to be essentially 0 but that’s only possible with a community support model, but people who have paid expect more than that. Also the whole value proposition of a Linux distro is that it’s infinitely customizable, runs on a huge array of hardware and includes a staggering amount of capability. The cost of supporting that is huge which is why the community model is the only viable means of supporting a full desktop distro. When you want to make money out of Linux you have to forgo that proposition. RHEL is only supported on a finite set of certified platforms. The supported packages are a limited subset of the packages in Fedora, it has a glacial upgrade cycle and finally it’s expensive which leaves room for a reasonable profit. In the consumer space the only viable model is Android which doesn’t look like Linux at all. Specifically the consumer has no control over any aspect of the OS, you get what your phone company wants on the phone. At most any specific phone will only get a couple of OS upgrades before the manufacturer stops supporting that model. And any particular edition of Android will only run on a small set hardware platforms.

    1. Well, that’s exactly what Microsoft is doing. They do not sell to endusers, but to OEMs. So does Google with Android, they do not really target end users, but phone maker.

  30. Not Ted Ts'o says:

    P.S. Next thing for Ubuntu to learn — how to pay their engineers
    well enough, and how to give them enough time to work on upstream
    issues, that once they gain that experience on Ubuntu’s dime and
    become well known in the open source community, they don’t end jumping
    ship to companies like Red Hat or Google. 🙂

    On the other hand, if Ubuntu management doesn’t learn, that’s also OK.
    Google is hiring. 🙂

      1. Mackenzie, your comment makes absolutely no sense. What are you trying to insinuate? Just because someone used to work for Red Hat doesn’t mean they went to Canonical for better pay/treatment or that they receive better pay/treatment. In fact, it doesn’t even mean they left Red Hat of their own choice. Furthermore, out of any software company I’ve talked to ex-employees of, Canonical has an extremely poor reputation for how it treats its employees.

        From the results of this study, it seems Red Hat employees have far more of a commitment to upstreaming their work than Canonical employees.

        Haha. Tee hee. Funny.

  31. Frank Earl says:

    Uh… Going and doing trash-talk about this stuff isn’t going to impress many when you get right down to brass tacks on things. Ever wonder if it’s not part of why Canonical’s where it is, considering that they’re doing the trash-talking themselves?

  32. Dulwithe says:

    This is buried at the bottom so maybe no one will read it, but a key point I think is “marketing” versus “branding”.

    Successful marketing leads to business revenue. Unsuccessful marketing leads to company closures and bankruptcies.

    Successful branding leads to greater popularity of a product. Unsuccessful branding yields no greater popularity or possibly less popularity.

    IMHO, Canonical’s branding of Ubuntu is very good. I am a linux desktop non-guru USER of about 10 years (remember corel linux??). I tried Ubuntu/Kubuntu when it first came out and several times at intervals since then. I have never stayed with Ubuntu/Kubuntu because it was fraught with too many problems for me. However, I really liked the Ubuntu/Kubuntu image (brand) created by Canonical, with the promo video showing Nelson Mandela video where he describes the concept of “ubuntu” – meaning community or village if I remember correctly – and describes the old African days when a traveller coming into a village would be treated with respect and have his needs taken care of.

    The powerful image/brand they created sucked in a lot of people and gathered a lot of “mind-share”, which as you all know is an important term and concept in the FOSS world.

    I just wonder how much better off we would be if another desktop-user oriented distro like PCLOS (or another similarly strong and more-trouble-free-than-ubuntu distro) had attracted the mind-share that the Ubuntu brand has attracted.

    BTW, the Mandela ubuntu video can be seen here:

    I didn’t see anywhere in Canonical’s literature where they asked and received permission from Mr. Mandela for using this video to promote their brand. (Perhaps they did receive permission, but I have my doubts…)

  33. lpbbear says:

    For me its issues with both distros/companies and a lot of others Linux based companies. It isn’t so much a question of who contributes more to the kernel but instead who actually FIXES the desktop issues that have plagued the Linux desktop for FAR too long.

    Does the distro not only recognize the sound card correctly but actually (gasp) not mute it at the first system startup thus leaving the user wondering why the sound isn’t working?

    Does the so called “NetworkManager” actually work out of the box without requiring hand editing of conf files and uninstalling NetworkManager completely?

    Does the distro come with a application that actually provides working VPN connections….again…..without ridiculous hassles? (please don’t say KVpnc, just got done restoring my backup copy of resolv.conf again after using it.)

    Does Samba actually work correctly out of the box again without having to screw around with hand editing conf files?

    Does the distro base their so called production releases on promising, but totally not ready for primetime, software such as KDE4 was initially?

    Linux could be a great desktop but Linux distros/companies seem to spend more time being concerned with fixing core issues rarely seen by desktop users while generally ignoring basic desktop usability applications and functions in Linux that are used daily by desktop users.

    Linux Companies, like Ubuntu, are basically recreating the same wheel every 6 months without really addressing the bigger issues of the basic desktop functions actually working….flawlessly.

    RedHat is also at fault for not taking a more public prominent positive stance in support of Linux based desktops. Their support has wavered back and forth for years. Put the weight of those “billions” behind actively promoting and supporting Linux as a desktop.

    Both companies are at fault in different ways.

    Until Linux companies begin to work together with their main focus being on finally addressing and SOLVING these kinds of core desktop needs Linux desktops will, sadly, continue to play third fiddle to less deserving operating systems such as Windows and Apple/Mac

  34. I would be much more interested if this was limited to contributions made in each company’s first 5 years (or in Red Hat’s case, since it’s older than GNOME, in GNOME’s first 5 years). It seems a bit unfair to compare the 14-15 years that GNOME & Red Hat have coexisted to the 5 years that GNOME & Canonical have coexisted, nevermind the relative sizes of the companies (that is, Red Hat was already sizable by the time GNOME started).

      1. Well, more level than I thought it was. Thanks for filling in the missing info. The bit where RH has 10x as many butts-in-chairs could also have an effect, though. In which case, you’d expect more like 2% for Canonical 😉

        I tend to think of Ubuntu’s version of GNOME as a fork of real-GNOME (which is what’s in Fedora). I haven’t heard much of plans to move to GNOME 3.0 in Ubuntu-land, so it seems to me like Canonical’s going to be trying to maintain a fork of GNOME 2 down the line. I wonder whether the code that makes up the fork-bits was submitted to upstream and rejected or never submitted at all. If the former, I’m not sure I can fault the patch-writer for a patch-rejection in cases where the code works but upstream doesn’t like the direction. If the latter: 😦

        But hey, I’m a KDE user. The fewer contributions to GNOME the better, right? 😛 (Yeah, I know, in a similar analysis of KDE’s codebase, Novell would be the clear winner.)

      2. PS: that bit about GNOME2/GNOME3 is pure speculation. I’m a Kubuntu person. The Ubuntu Desktop Team may have plans that I haven’t bothered to find out about.

      3. mackenzie: Greg already answered the point about relative size with two points. One, Red Hat also does a whole bunch of other upstream engineering which Canonical doesn’t do. Two, if it were the case, then Canonical’s total contribution should be small, but it should have just one or two people making disproportionately large contributions – say, in the top ten – right? But no, it doesn’t, Canonical’s top contributor is at #29.

  35. Steve L. says:

    Wow, what a screed. Sounds like an ex-grunt from big, bad Red Hat is mad that Canonical is doing something to advance the Linux desktop by making it easier and more accessible – in a manner that RedHat either wouldn’t or couldn’t.

    Are you working for Microsoft now?

    1. Dulwithe says:

      @Steve L.

      Sorry, Steve, but you’re a knob!

      @ The Moderator – Feel free to remove this comment if you wish.


    2. George Mitchelll says:

      Except for the little detail that they haven’t made it “easier” and “more accessible”. They have simply copied what a lot of other people have created and marketed it in a way that makes uninformed people think that they invented it. Either Open SuSE or Mandriva is easier to use than Ubuntu, not to mention PCLinuxOS or a host of other smaller names. And for all the knocks it gets, Fedora itself isn’t that bad. But these companies don’t spend a lot on marketing, branding or implying credit to themselves for other people’s work. I recently tried Ubuntu only because Ubuntu supports a specialized program out of the box which I would like to have. But I found it primitive compared to what I am used to with Mandriva. But it is designed for novice users in a way that PR makes up for any lack in technology. Similar to the Microsoft approach, actually.

      1. Peter Webb says:

        > Either Open SuSE or Mandriva is easier to use than Ubuntu

        Have you got any usability studies to prove it, or did you just pull it out of your ass?

        Any details about this specialized program? How come the same program was primitive in one distro but not the other?

        Any major technology failing of any distro is a linux failing? But then you would realise that if you were not blinded by jealousy and hate?

        Perhaps you should ask why Fedora does not have a major mindshare despite their good work elsewhere. You should also ask how important is to have 100% code contribution to provide a perfectly engineered desktop that does not do what the user wants?

        To me it seems to be a difference in philosophy. One group wants to provide a perfect desktop, that is based on pure ideals, but does not care if nobody else ever uses it. The perfection of the desktop is an end to itself. Others would like to be more pragmatic. There is no point having a good thing if you don’t share it with others. And the more that can share it the better.

  36. Aleve Sicofante says:

    I just don’t get it. Here comes again the legendary contradiction: “The advantage of open source is that you are free to change it yourself and create something new. Then, if you don’t like the way “upstream” works and don’t collaborate with them, you’re a bastard.”

    Listen: RedHat DOES NOT care for the desktop user. Sending desktop users to use a permanent beta thing (Fedora) and ask them to search for help in a typical Linux-way RTFM community means the company doesn’t give a damn about those users. (BTW, the Ubuntu community is not better. No business should rely on community support at all.)

    Canonical just doesn’t like upstream. So what? They want to make things their own way. So what? They DO provide commercial support for their desktop users. They DO care about their desktop users. If I wanted Linux for my company would I go RedHat or Canonical. Would I ask for support to a hundred youngsters making fun out of my ignorance or would I pay for a help desk to guide me?

    Sadly, Canonical is just a billionaire’s toy and most of their design choices are just capricious showing they (he, really, it’s just the billionaire) haven’t got a clue on user interaction design. But they are entitled by the very GPL to do that if they wish and if they didn’t contribute a single line of code upstream, blaming them for that is the typical hypocrisy of the open source world. THEY DON’T HAVE TO. Period.

    1. George Mitchelll says:

      Thats right THEY DON”T HAVE TO. It certainly not illegal to talk about sharing all the time and then not do any of it yourself. In the case of Gnome (without which Canonical would not exist) the sharing they continual prattle about only goes in one direction. THATS the point. That is the very definition of “hypocrisy”, check the dictionary. Of course they also take the liberty to share commercial codecs without permission. They are VERY BIG ON SHARING. You’ve got that right. Nothing wrong with providing commercial support for desktop Linux, but at least do it in an ethical way.

      1. Peter Webb says:

        This is such a myopic view. The code changes they make is available for sharing (GPL no less) but is just not shared upstream. What you are saying is that “sharing counts only if you share it my way with my rules”. Nice definition of ethics.

        FYI, there is a very large world outside the US that is not stupid enough to accept US IP rules without question.

  37. exploder says:

    Canonical receives way too much credit and press when they produce some of the most bug and regression filled releases in the Linux world. Canonical is not doing us any favors. Build something that has decent hardware support and update the popular apps, then the press will have something worth reporting.

  38. dart says:

    Why the fuck u r so jealous Greg. I c u r pissed off from ubuntu’s success, its ability to connect to its users , its strong community base, n i know every red hat employee or ex employee is.

    You qouted one fact. Now can u quote another fact about the desktop market share of ubuntu n red hat, but i know u wont. Just because ubuntu is in desktop market n RH was not able to keep a foothold in desktop market and eventually withdrew from it u guys are seriously pissed off. I know u wil say that i have already left red hat n i have nothing to do wid it but there is lots of heat up ass.

    P.S. Type linux in google. The first link is of and second is of ubuntu,NOT RED HAT. (forget about SEO crap)

      1. I know, thats why my comment was tongue in cheek. 😀 Sorry for the ahem .. strongly worded post by the way. That guy seems to follow me all over the internet posting silly crap like that, and I lost my temper with him a ‘lil bit this time.

    1. “Why the fuck u r so jealous Greg. I c u r pissed off from ubuntu’s success, its ability to connect to its users , its strong community base, n i know every red hat employee or ex employee is. You qouted one fact. Now can u quote another fact about the desktop market share of ubuntu n red hat, but i know u wont. Just because ubuntu is in desktop market n RH was not able to keep a foothold in desktop market and eventually withdrew from it u guys are seriously pissed off. I know u wil say that i have already left red hat n i have nothing to do wid it but there is lots of heat up ass.” — dart

      Red Hat(R) focuses on sustainable, commercial offerings. Fedora(TM) is for everything else. The developments overlap, sometimes even transition from being Fedora-only to Fedora and Red Hat, but the supported, sustained products are only going to be offered by the former. If anyone shows Red Hat a sustainable, commercial offering for the consumer desktop en masse, Red Hat will be all over it. Business desktops are another story, because businesses actually spend money. Red Hat has never, ever left the sustainable, commercial offerings in the desktop subscription area.

      Until then, most “loss leader” desktops area is regulated to those that pay money and/or support hours to PC OEMs as, essentially, charity. Instead of paying charity to PC OEMs and others, those dollars would be better spent paying GPL developer salaries in my view. I would rather fund a company to spend money on GPL development than one that has to do such for PC OEMs. This has traditionally been the model that Linux companies have had to adopt to “get in the door.”

      And I’ve worked with many at several different Linux vendors in prior years that have been a party to such. So take that reality and feed it back to what you think Greg is really talking about. It’s about GPL, not adoption, because we all benefit from GPL. This is the open source world, code rules, for everyone to use. Adoption is about marketing and brand name … I left that rat race when I left the IP-locked commercial software world.

      “P.S. Type linux in google. The first link is of and second is of ubuntu,NOT RED HAT.” — dart

      Try Bing. They buried Red Hat on the 4th page from day 1. Even Novell, Microsoft’s partner, was regulated to page 3. Search engines are marketing engines, with a clearly defined theme that is not about popularity. Gentoo was even beating out Ubuntu on the first page … interesting considering who founded Gentoo and where they work now. 😉

      I also personally love it when people like to use Fedora and Red Hat statistics separately (or rebuilds like CentOS for that matter), where they work against them, instead of focusing on the duality. Then again people don’t differentiate between Ubuntu and Ubuntu LTS either, causing themselves their own issues too.

      Just goes to show that people focus more on brand names than substances. 😉

  39. Andy says:

    No surprise here.

    The sad thing is that Fedora for instance isn’t growing quickly only because it doesn’t ship with illegal codecs like Ubuntu does (just wait till they get sued, I checked popcon, and 70%-100% of users are using ubuntu-restricted). Ubuntu also took off because of the perception RPM’s suck compared to debs.

    You must also remember that Redhat is being used in many places where people don’t even realise they are using it. Many “pro-Ubuntu stats” are from linux fansites, which clearly misrepresent corporate machines. Ubuntu is dead in the corporate market.

    1. The RPM vs DEB perception is unfortunately a very good point. Even though they have the exact same limitations, and DEBs are 10x more difficult to build RPM still has the stigma.

      I wonder where we would be today if yellowdog update manager was integrated in RedHat earlier on.

      Meh. Anyone with half a lick of experience can shred the RPMs suck DEBs are teh awesome argument with their eyes closed.

    2. “The sad thing is that Fedora for instance isn’t growing quickly” — Andy

      What is this based on?

      Fedora is much stronger than people assume. It gets dismissed, until people see some of the real statistics on long-term installations, upgrades, etc… It’s a real eye opener.

      And I’m not just talking about Smolt statistics, but several others. Although it is funny that Smolt garnered so much attention from other distributions. 😉

      My God, Fedora is actually capturing _real_ statistics of not only its deployment, but use, updates, etc… 😉

  40. Red Hat’s 16% would be code, whereas Canonical’s 1% has to be all the bugs.


    Cheers from a happy CentOS user (on servers and desktops).

  41. guybrush says:

    So what? Red Hat is targeting the enterprise. Canonical is targeting the Linux fanboy, and the fanboy is about noise and ideology, not technical merit.

  42. Dave says:

    Its interesting to note that a much smaller company, and one that only lasted from 1999 – 2001 is still a bigger contributor to GNOME than Canonical. Good job Eazel!

  43. Daniel says:

    16 % of people called Nelson Mandela support Ubuntu, but only 1 % of them wear red hats. At least RHEL is not a rebranded build of Debian Unstable.

  44. Folks, this is the wrong battle to be fighting. We’ve lived through the wars of Vi versus Emacs and Gnome versus KDE, etc. We’re all on the side of Free Software. Let’s come together and stay focused on building great products that are open. The enemy is proprietary. The enemy is closed.

    1. “Let’s come together and stay focused on building great products that are open.”

      A sentiment I thoroughly endorse — and upstream projects are the very means by which this “coming together” is supposed to happen.

      1. Great, so we’re in agreement there. The next question is: Why is Canonical not upstreaming more code? That question should be followed with: How can we help or encourage Canonical to exhibit the desired behavior. You catch more bees with honey than vinegar.

      2. Jef Spaleta says:

        What makes Fluendo contribute disproportionately more than Canonical? Perhaps because Fluendo isn’t invested in setting up a parallel set of intrastructure tools they control and use exclusively for their in-house work.

        A lot of what is holding Canonical back.. is Canonical’s unwillingness to step outside of their own infrastructure and their own copyright control boundaries and participate in processes they don’t have ultimate control over.

        It’s like watching an agoraphobic build a playground inside their own house and inviting everyone to come inside and play with them, instead of going out the door into the wide open and playing with the other children in the municipal park.

        Canonical wants the safety and luxury of being in control of as much of what they produce as they can. They don’t want to be a peer among equals. They want to control projects instead of having to deal with the process of consensus building among peer contributors or share control with other corporate interests.

  45. aL says:

    well, i started in the linux world with ubuntu 2 years ago…

    And i am really interested in this world now… i even programed an small application for gnome to listen music with gstreamer and dbus in c and a frontend with pygtk …

    it may had happened with another distro (i tried several before) but it happened with ubuntu…

    I am not confident enough to change my home distro right now (even tho ive configured computers with debian with no problem) nor i have the time…

    but i guess you could tell i started in this world thanks to ubuntu marketing or whatyoucallit… my gf and aunt are using it now too with no major complain for 3 and 11 months now (im not even around my aunt to help her and i didnt have to help my gf once yet)

    Anyway, ill take an eye in RH eventually… coz i am interested in LPIs 😉

    Maybe in january if i have the time…

    I dont know what you call my case…

    1. aL — the most important choice in a distro is … the one you’re already running! Furthermore, if you’re new to Linux, pair up with someone that does similar to what you do, and run that distribution. Why? Because they already know what you’ll need and you’ll get the best technical support.

      Running Linux is most important from adoption.

      If you’re interested in LPI, understand it covers both Debian-based (including Ubuntu) and Fedora-based (including Red Hat), since they are very popular. If you know Linux, and go through the LPI objectives, you should do well on their exams. Also know that when it comes to training, Red Hat is the most highly regarded in the industry for its approach — exams and training — and not just in the Linux world, but IT industry-wide.

      Anyone who considers themselves a “master” on tuning Linux should take RH442 from Red Hat Global Learning Services. It’s the class I absolutely recommend the most. If you have any training budget in your organization and want to bring an instructor on-site for your team of experienced, senior-level sysadmins, that’s the class, even if you don’t believe in certification at all.

      1. aL says:

        Thanks for the advice…

        No i dont have any training budget since i am currently unemployed :S (I couldn’t have learned so much in the last 6 months without the time anyway :D)

        Im taking a CCNA +CCNA security course right now. I plan to take a look at LPIs when i finish this task. 😉

      2. aL says:

        I forgot… the point of my post i wanted to stress… I think i owe this to ubuntu… even if ubuntu builds on top (or ride on, as greg puts it) of many, many work from others…

        That must be worth something… 🙂

  46. Not Ted Ts'o says:

    It is quite simple , ubuntu is a great desktop since they do not give a dam about patents/IP, or stock holder getting sued for laps in ip, you want dvd to play on ubuntu out of the box it is there in the repo off shore. Oh that new spiffy nvidia/ati cards has a repo from the manufactures so that is why dell chose ubutu to sell and support. But I would never have a ubuntu box as a server at a real company, why ? There is no focus on the server, aside from the Amazon E2c work they did. Ex-redhat ers who have been interviewed said so in with Canonical said the suse and redhat have corporate market locked up. Ubuntu has no plans getting btrfs to where redhat has already a btrfs transactional rpm , so you can undo the under lying file system. As an old systems admin who has seen the pain that can be left behind when rpm transaction are unrolled leaving the box a pile of ash.
    With ubuntu 10.04 one can not remove snapshots in btrfs 😦

    As a laid off ex-redhat I have no love for management at rh, it pains me to say they have a better corporate server. It was the people that made RH great as more people like Greg , Alan Cox, Alex Maier, David Burney , Chris Van Hoof, Bob Young , Matt Szulik leave RH is one more corporate company. Where one has to use the American with Disabilities Act to keep a job. Where they pink slip of the transformation department department when the eco tanked, for this department was in charge of building out a SOA enterprise back end systems vs the hodge podge of silo systems. For the shadow man is not in charge but the corporate sales man is. I look forward to deploying centos and Oracle un breakable linux, as the corporate man trained me well there is no loyalty, but to the bottom line.

      1. Erle says:

        mine .. my brothers .. my father … somehow, it was the marketing of ubuntu that got us into using linux.

        That must be worth something.

        Is it a great desktop? For me it is. Great is a relative word isn’t it?

  47. I’ve learned to ignore Shuttleworth. I’ve learned to ignore the Ubuntu fanboys that badmouth Red Hat as well. In reality, I feel sorry for Shuttleworth and Canonical. After all, for all those fanboys that talk about how great Ubutnu is … they aren’t funding Canonical at all or just a tiny bit.

    Which is why I eventually ask them, “If you really believe in Ubuntu, why aren’t you paying them like Red Hat?” Most of them respond, “Oh, I only pay for mission critical and Ubuntu wants too much, like Red Hat, so I might as well go Red Hat.”

    Let’s see here … something about that grassroots/fanboy approach isn’t paying dividends … or more appropriately, paying GPL developer salaries so we can have more GPL code. Regardless of distribution. L-<

    I don't claim Red Hat has the "best model." I just point out Red Hat is still the *ONLY* 100% open source Linux vendor that has a *SUSTAINABLE* model, one that keeps GPL developers employed, and keeps things moving.

    Red Hat(R) and Fedora(TM) is not a company of two distributions, but two trademarks. If Red Hat offers subscriptions and support behind it because it's sustainable, it gets the former. If not, there is Fedora(TM). Doesn't mean Red Hat employees are not working on the latter.

    In fact, Red Hat employees are quite overly involved — compared to any other 'distribution" — and that says something. I also don't like seeing Novell badmouthed alongside Red Hat, because they too do contribution. I may not believe Novell's models are as sound as Red Hat's in the open source and Linux world, but I'm not going to badmouth their contributions.

    Because in the end, it's about GPL, which anyone can use. Why Shuttleworth and Ubuntu fanboys want to continually badmouth GPL producers is beyond me, from a standpoint if minute proportion. But those who know, who have been around, know. Red Hat is here to stay, and continue to work with the community, and fund as much GPL development that is sustainable.

    And that's why I've defended Red Hat for the past decade. Not because I think they are "better." But because they have built a sustainable model. And I limit my rebukes to people who want to badmouth Red Hat, and even then I bite my tongue most of the time.

    1. Tobias says:

      > I’ve learned to ignore the Ubuntu fanboys that badmouth Red Hat as well

      I only ever see badmouthing in the other direction.

  48. bochecha says:

    Wow. I could have expected that from anyone, Greg, but not from an ex FPL who, I admit, was until this very post one of my models as a Fedora ambassador.

    I’m the first one to think that Canonical doesn’t contribute as much as they could, and that « we are smaller » or « we are younger » are no excuses.

    I know they carry downstream patches (and an almost forked Gnome desktop!) instead of working directly upstream.

    But this is just too much of a rant, and it’s way too easy to just bitch about how they are mean and don’t put their money where their mouth is, and it makes me even more sad that it’s coming from someone I had so much respect for.

    Let me just take the example of the one Canonical developer I personally know. I’ve seen him go to fd.o propose a standard, post patches in Gnome bugzilla, discuss on the devel lists about changes he (and Canonical behind him) would like to implement,…

    Most of those were refused, often being dismissed in a very condescending way. Guess what? He is still trying as hard as he can to push everything he does upstream.

    So in my opinion, this guy has a very low number of commits (maybe even 0, I don’t really know/care), but in my opinion he is doing everything right.

    Now, you might think that he must be damn incompetent to have so much of his ideas/patches refused. Fact is, almost every time, initial (and sometimes vehement) criticism came from Red Hat/Fedora folks.

    Could there be some rivalry between the 2 companies that would make Red Hatters be more inclined to refuse what comes from Canonical? I don’t know, and saying so would just be FUD.

    Add to that the fact that Red Hatters are the primary maintainers of most of the Gnome modules (and most of the Linux ecosystem really), and that makes newcomers in a tough spot to make their own room and take a slice of the « we are the ones behind Gnome » pie.

    Is Canonical marketing more than what they do? Definitely. Are they failing to be as good community players as possible? Probably.

    But please, don’t make it a black vs white statement. This is just bullshit and is plain stupid (and again, disappointing)

    PS: I’m a proud Fedora user and contributor who just tries to have some perspective and not be a dumbed-down fanboy.

    1. I’m sorry, Bochecha, to disappoint you by having the temerity to express my personal opinion. And if it you’ve lost any respect for me, so be it. I’m no saint.

      The facts are the facts. Canonical has, at this point, *lots* of engineers on staff. It is now evident that not *one* of them does serious heavy lifting in the upstream GNOME community. I’d never really understood that until today.

      And now that I do understand it, it pisses me off. You don’t have to like it. That’s fine, and your opinion.

      My opinion is that it’s long been time for Canonical to recognize their responsibility to step up and participate more directly in the upstream communities that sustain them. “It’s hard and we don’t have to” is a lame excuse.

      1. Peter Webb says:

        But you don’t address his main point.

        He contends that it is much harder for a Canonical developer to get an accepted commit because of pre-existing bias by Fedora developers. It seems to me that this a far more serious problem than the relative commits by two commits.

        *If such a bias exists*, and if you really cared about linux, that is what you should be pissed about.

        It may be easy to get that statistic; may be a summary of rejected commits by company?

      2. bochecha says:

        > « But you don’t address his main point.

        He contends that it is much harder for a Canonical developer to get an accepted commit because of pre-existing bias by Fedora developers. It seems to me that this a far more serious problem than the relative commits by two commits. »

        That’s not really what I meant (but not being a native english speaker might be why this wasn’t clear).

        I’m not saying Red Hat/Fedora is actively pushing back Canonical contributions. I sort of implied it unvolontarily, but what I meant was to say that *IF* there is such thing, which I clearly have no idea of and don’t event think it’s true, then that would be a huge problem.

        No, my main point is that when you are the primary maintainer of a module, then you can just commit whatever you want to it. When you are an outside contributor (which everyone is at the beginning), you have to propose your changes, defend them, have them accepted, and so on.

        Right now, Red Hat is the primary maintainer of lots of Gnome modules (and given the number of red modules on the picture in Dave’s report, I could even say Red Hat is the main maintainer og Gnome, period), and Canonical is the newcomer who has to justify every change they want to make before they get in.

        That certainly helps pushing the numbers towards one direction.

        Now, is that an excuse? Certainly not. Smaller startup companies have proved this is possible (see how much Collabora, Litl and the others perform in the report).

        But still, flaming away and making one particular contender look like the big bad wolf is way too much in my opinion.

        Others said it already in the comments, you catch more flies with honey than with vinegar.

        We should be praising Red Hat as an example kudos by the way), congratulate those small startup companies for their awesome participation, and encourage Canonical to go on and grow this 1%.

      3. Craig says:

        I’m sorry Greg, you failed to respond to he comment/question that perhaps the lack of Canonical contributions is at least in part attributed to the difficulty in getting maintainers to accept patches – those maintainers for the most part being RedHat employees. If your frustration is any indication of how other RedHat employees feel, then perhaps RedHat is creating a bit of a walled garden to contributions?

      4. Craig says:

        (hmmm… OK I forgot to refresh since from when I started reading this and see bochecha clarify / still… while anecdotal and perhaps other companies don’t have the same contribution issues – it’s not unreasonable that sour feelings and Ubuntu’s popularity would make it difficult for Canonicals their ideas/patches to be accepted. Why invite a popular kid, who has “ridden on your coat tails” to the party?

      5. craig: greg isn’t a Red Hat employee any more (which is why he wrote this post), and even when he was, he wasn’t involved in maintaining GNOME components (AFAIK).

      6. Peter Webb says:


        I did not mean that the bias is intentional or malicious.

        It might be easier to quantify. What we need to know is the commit proportions for each module along with the company affiliations of the module maintainer? Is that easy to pull from the data?

    2. bochecha —

      I think it’s best to focus on the results, not the arguments. I regularly talk about the harm “fanboys” are doing to the bottom-line of Canonical, or any other distribution that is held up for the “free beer” and is the “cool distro of the year.” GPL (free speech) is king.

      As far as contributions being rejected, I see many happen and it’s not a distro-alignment detail in many cases. Many times they are patches that have already been considered, and rejected for sound reasons. However, I do understand the need for Red Hat and former Red Hat employees to be “above things” and never let these details into decisions … ever … regardless of what is going on.

      I can’t speak for Greg, but there is a huge backlash against the “fanboys” that aren’t pushing the truth out there. I just hope people realize that good, helpful, contributing Canonical employees and upstream contributors from Ubuntu are caught-in-the-middle. And we should praise them for fighting through the fanboys, the backlash, the counter-backlash, etc… as they contribute more code.

      That’s the greater message I think.

    3. was this trying to patch ubuntu one support into nautilus? cuz as a GNOME foundation member i certainly don’t want to see support for a proprietary service (yes with open APIs. when i invite people over for dinner i don’t give them recipe cards and tell them to cook it themselves) built into GNOME, sorry.

      1. bochecha says:

        > “was this trying to patch ubuntu one support into nautilus? cuz as a GNOME foundation member i certainly don’t want to see support for a proprietary service (…) built into GNOME, sorry.”

        It’s funny how you went from my « he proposed a spec at fdo and does everything right » to « he just cares about his in-house proprietary crap ».

        Like I said, I’m a proud Fedora contributor and ambassador, which should mean something about my adhering to our four foundations (you know, one of those being « Freedom »). So when I say « in my opinion he is doing everything right », I’m obviously *not* talking about anything related to a proprietary service.

        To suggest the contrary is quite insulting of your fellow community member.

      2. Sorry bochecha, I was not trying to be insulting at all and I sincerely apologize. I obviously don’t know the person you’re speaking of, and am merely listing the one complaint I know of showing Canonical-developed code not being accepted upstream. I’m glad your contact is continuing to try to upstream despite the obstacles faced.

    4. jonathon says:

      >He is still trying as hard as he can to push everything he does upstream.

      And how many of those patches _increase_ the number of bugs in the software?

      Increasingly, I’ve been experiencing bugs in Ubuntu that simply are not found upstream.

      From my POV, Windows Vista and Ubuntu 10,.4 are equally dysfunctional, broken operating systems that should never have been released. They have too many bugs to be considered anything but pre-alpha. Canonical is starting to look a lot like Microsoft — pre-alpha code is being shipped out as if it were general release quality.

      Canonical is marginally better than Microsoft in hiding its faults by increasing the amount of eye-candy. However, such things do nothing to improve the functionality, or productivity of the operating system.

  49. Toxie says:

    Install Ubuntu. Go ahead, I’ll wait. OK now install Fedora. Notice a difference? No mere mortal ever completes a Fedora install, while the Ubuntu guy is already reading email or watching porn or whatever. This is Canonical’s value. Idealism gets in the way of practicality with the RH guys (and who are we kidding, Red Hat == Fedora). I thought we were over this religious crap. It’s 2010 people.

    1. It’s *NOT* religion …
      – Liability gets in the way (e.g., Fedora *IS* deployed in corporations)
      – Mitigating indemnification issues gets in the way (e.g., downstream distros from Fedora)
      – Not putting open source at the mercy of proprietary licensing gets in the way (e.g., now having a dependency on a commercial company)

      And for those that think it’s an US-only, please do *NOT* advise people (let alone corporations) on such. If the sole argument is on including unlicensed, proprietary software, then that’s rather poor. E.g., there are unofficial Fedora Respins like Omega.

      1. Windows and MacOS X are proprietary platforms that both provide their own as well as license (as costs passed on to the consumer) proprietary IP various support. These will never be free beer (much less free speech) because of such.

      2. elric says:

        So what? Those are the platforms desktop Linux has to compete against for adoption, and mere mortals don’t care much about free as in beer, much less about free as in freedom.

    2. Gabe says:

      I did as you suggested. The Fedora 13 install took about the same length of time as the Ubuntu install. Figuring out how to install illegal codecs and flash took 15 minutes, same as it does in Ubuntu.

      I ended up with a desktop that was pleasantly close to stock GNOME. I really dislike Ubuntu’s indicators.

      I may not be a mere mortal when it comes to operating systems, but I know easy when I see it. Neither Fedora nor Ubuntu has an advantage here.

  50. Peter says:

    While I see that Red Hat contributes much more to free software projects than Ubuntu, and that it is a bigger, more profitable organization, I’m not sure how that effects me. Ubuntu is a much nicer distribution. Not a little nicer. Dramatically nicer. Ubuntu is more user-friendly than Windows now. Red Hat/Fedora is hell even for advanced sysadmins. I honestly do not understand how Red Hat is so bad given the calibre of people Red Hat employs. But it is.

    I’m really not sure what your point is. Ubuntu is parasitic on Red Hat? Possible. People will somehow install a distribution that barely works over a superpolished one because of that? Unlikely.

    If I were Red Hat, I’d try to spend a tiny portion of what is being spent to develop Gnome and the kernel, and try to find some way to fix RPM, and to revamp the distro as a whole up to Debian or Ubuntu levels.

    1. First, Ubuntu is not more user friendly than Windows. Second, any “advanced” sysadmin that believes RedHat to be “hell” is not much of a sysadmin at-all. There is nothing wrong with RPM, just as you do with a Debian package you manage it with a utility that resolves dependencies. Sounds like you have been drinking too much of the Ubuntu kool-aid.

      1. Peter Webb says:

        > First, Ubuntu is not more user friendly than Windows.

        Opinion. As it is worthless without proof. Are you saying that the few UI tweaks that Ubuntu does reduces the usability of a vanilla gnome desktop to that of windows?

        > Second, any “advanced” sysadmin that believes RedHat to be “hell” is not much of a sysadmin at-all.

        Opinion about opinions. No facts either way.

        > Sounds like you have been drinking too much of the Ubuntu kool-aid.

        Opinion and ad hominem, unjustified by your comments. What are your motives?

      2. > Opinion. As it is worthless without proof. Are you saying that the few UI tweaks that Ubuntu does reduces the usability of a vanilla gnome desktop to that of windows?

        No, it is not simply an opinion. The market speaks for itself. There is no “consumer” distribution that is on par with Windows to date. Enterprise distributions however do. This could be due to the level of effort put into open source by RedHat vs the lack of effort by Canonical.

        > Opinion about opinions. No facts either way.

        Not really. Even more “advanced” topics like packaging and performance tuning are relatively easy with RedHat.

        > Opinion and ad hominem, unjustified by your comments.

        Not at all.

        > What are your motives?

        Why must I have a motive? My only “motive” was to respond to the inaccurate information posted by the op.

    2. Red Hat works on more than just the kernel or GNOME.

      An overwhelming majority of the components I hear people talk about how Ubuntu does “cool” were not only written by Red Hat or Novell, but many were in Fedora first, and worked very well. I honestly cannot believe how often people ignore this.

      But they do.

    3. Luya Tshimbalanga says:

      “[…] and try to find some way to fix RPM, and to revamp the distro as a whole up to Debian or Ubuntu levels.”

      Nothing to do with the package management itself but the policies do make difference. That kind of statement only shows most who praise about Deb mainly use apt-get or aptitude but never tried dpkg. When they will be asked to only use the latter i.e. low level of package management, let see if they will make that sort of statement.

    4. “I’m really not sure what your point is. Ubuntu is parasitic on Red Hat? Possible. People will somehow install a distribution that barely works over a superpolished one because of that? Unlikely. ”

      It’s funny how everyone assumes that any time anyone related to RH/Fedora posts about Ubuntu, we’re trying to make people use Fedora instead. We’re not. We probably don’t really *care* what you use. What we care about is the development of the overall F/OSS ecosystem. When there’s a company which presents itself, and which is generally perceived as, a leader of that ecosystem, and yet contributes almost nothing to it, there’s a problem.

      Let’s be clear: the optimum solution Greg’s looking for here is not ‘Ubuntu magically goes away’, it’s ‘Canonical steps up and starts making significant contributions to the wider ecosystem’.

  51. As a developer of FOSS I once carried the flag of RH and only installed RHEL on servers, fought through driver capability issues, etc. But than one day RHEL 5 came out and something funny happened….one isty-bitsy file that was needed to get Tomcat and Apache connected was only accessible if you had a PAID subscription to RH. The majority of my customers didn’t want to pay $700 a server for this one little file. (I actually put it up for free download on my site and pissed of RH a little)

    I decided to give this new distro called Ubuntu a try. Everything worked day 1. My hunt for drivers got cut by as much as 98%. It just worked with little thinking which freed up more of my time for my software.

    Regardless of who is doing the work, Canonical does a better job of bundling/packaging/whatever so I put my vote with them.

    1. John — I don’t understand your comment at all. How were you installing and receiving support on Red Hat Enterprise Linux release 3 or 4 at your customers at all without a subscription or other, valid entitlement?

      As far as “pissed of RH a little,” trademark law is very unforgiving (especially in Germany and select, other locales more than the US). I know many people wish it was otherwise, but some select commercial companies really abuse and test the charity of GPL companies.

  52. brian says:


    tried to read most of this thread but many comments were like being back on the school grounds and two kids pushing each other saying the other is ugly.

    No one wins with those kinds of rants.

    1. Then focus on the ones with technical meat, critical analysis and the realities of the industry.

      In a perfect world, consumers would pay for product, and the providers of the product would be able to hire more developers. It still works very well for GPL-centric companies too, although recent articles show that — as a reguarly quroted example — Red Hat replaces every $10 of proprietary, lock-in software with $1 of open source subscriptions.

      But the sad thing is that too many people want the “free beer” or I’ll call it “free lunch” plan. They don’t just expect free software at home, but want to base a business on one where they don’t fund the developers of a significant portion of the software they use — especially long-term software with real, sustaining engineering costs — any money. And PC OEMs expect Linux vendors to give them not just free software, but support and other things at reduced (often loss leader) costs, possibly even give them money just to ship. How sustainable is that?

      Business needs in sustaining engineering funding leading edge developments that home consumers care more about … now that’s what sounds feasible. Guess what? There’s a company, with a project, that focuses on that. And guess what? That company products significant portions of major projects regularly. How Red Hat got known for only the kernel is beyond me. Heck, even GKH over at Novell-SuSE has done his best to dispell that as well (although I can’t condone his tact towards Canonical as well, the data is still beneficial).

      Oh well, I’ve spent way too much in this thread, regardless if it’s on my own time.

      1. Jef Spaleta says:

        It’s important to note that GKH’s initial “attack” was actually aimed at Google while standing on Google’s campus trying to actively shame them into contributing more to the kernel. The only reason why the remark about Canonical was made was because someone in the audience of Google employees asked where Canonical was on the list. He made an off-the-cuff remark about the lack of contribution from Canonical, that Canonical employees got publicly defensive about so he obliged them with the hard numbers…publicly.

        So lets cycle back to that and widen the criticism. Out of all the vendors who are leveraging the linux kernel and an open source stack that sits above it in the consumer space who is the best example of doing it right? Litl? is tiny Litl the shiny example of a consumer product oriented company doing the work the right way?


  53. Luya Tshimbalanga says:

    Not surprising about Canonical action. What I found sad is how easily gullible majority of fanboy went to praise Ubuntu as their savior.
    I think it was about time that some ex-RedHatter expressed what they saw with Canonical taking credits on work not being done by them.

  54. adante says:

    I’ve similar comments by GKH about about Canonical’s (lack of) contribution to the Linux Kernel [1], or Michael Scarf on Eclipse [2]

    I’m not trying to be a jackass here, but could you elaborate on what basis Canonical is obliged to contribute back?

    I get the impression that it is implied as a matter of decency that if one were to gain benefit (not necessarily commercially – as you are quick to state Canonical has not done so, except for street cred) from upstream projects, one should reciprocate.

    But how does this compare against other downstream users? Most end-users generally do not contribute back to the FOS-software they use (I have lost the link, but the ratio of contributions to users in OS projects is much closer to a 80-20 pareto principle ratio than 100%).

    Of course most end-users are not _capable_ of easily making contributions back. But consider Google, who are one of the Linux systems largest corporate users (and profieteers) – they have made significant modifications to the kernel and are certainly technically capable of giving back[3].

    Obligations wise, how are these guys different to Canonical? Or should they also be taken to task?


  55. trunks says:

    Wow, kiddo galore. Adults degenerating into small kids once more.

    Not to say those figures aren’t a big difference, 16 to 1 is huge alright, but then again, there’s more to Ubuntu than direct contributions, too.

    Much of Ubuntu’s focus is on what Joe User wants: He wants an out of the box working OS on the PC he just bought. For that, Canonical focuses on stuff that directly impacts this problem. They’re trying to get OEMs on board, improve the system for them, enable Ubuntu to be shipped with all those little proprietary bits that are necessary to make the OS enjoyable by Joe User.

    What Ubuntu managed until now is no small feat: With their marketing and their focus on smaller details they actually managed to create a huge community. A community with many new and willing “opportunistic developers” who start up [random project x] that wouldn’t have ever started without Ubuntu and might just have major impact at some point.

    Well, so much for that stuff.
    What I actually wanted to write down is what keeps me sticking with Ubuntu rather than other distributions…

    Launchpad is a great collaboration platform and especially great when you have all those ppa’s for daily builds of projects that you can selectively enable and get updates on.
    It enables a lot of same-day bugtesting, more bugreports and just in general a lot of exposition for products-in-development. No longer do I have to keep many scripts and schedules for me to build my own development snapshots of software I’m interested into using the latest, even if more buggy, versions.

    That’s likely the single biggest reason I stay with ubuntu, the ease of PPAs for daily builds for testing.

    This project started by Canonical actually does a lot to help usability doing everyday work. There’s a ton of little annoyances and quirks in everyday software that, while not big enough to be regarded as “important” to most developers, can turn using the software into a hassle.

    Software Center:
    It’s a great simplified way to Package Management that even the most nontechnical users that I know have easily adapted to. Sure, there’s the “Add/Remove” thing that is still default in Gnome (afaik) but it just doesn’t have the ease of use and ‘pretty’ factor that the average user wants.

    Indicator applets:
    Indicator applets are damn nice, they might be more restricting on the system-tray things, but having things that belong together grouped together like that (and take less space) is already looking very promising.

    Willingness to focus on design:
    This is something that’s often completely ignored in OSS circles, so I’m really glad that a company finally steps up to the task of getting design to be more mainstream ready.

    Canonical might not (yet) contribute as much code upstream as other companies, but they are doing their part.

    They got new people to the OSS environment, they manage to bring ‘opportunistic developers’ into the community and they’re making it far easier to distribute your software to many people *and keep it updated*.

    They’ve never made extremely huge leaps, but you can see a steady improvement. Is it bad to focus on things that enrich the OSS environment in other ways than code?
    If you believe it is I have to ask that you try to get a good look at the overall. An ecosystem is not only about one part (code) it’s also about community, spreading the word (to fix bug #1 just using word of mouth won’t work), doing the stuff nobody else wants to do.

    Red Hat, Novell and all the other players who contribute a lot of code rock. They are great companies for their contributions. Canonical is definitely not without faults, they can definitely still improve their code contributions (there, I said it!), but please… It doesn’t help anyone to be childish about things like this.

    Canonical is trying its best to push Ubuntu to the desktop, especially OEM. What do you suppose they should do?
    Give away a thick book of all the contributor names and percentages of contributions every one of them makes to any software shipped in the Ubuntu repositories?
    Isn’t the credit given in any about dialog or copyright notice? Isn’t that how it usually works?

    I really don’t understand all the hate for Canonical. They’re trying to focus more on projects to bring Launchpad bugs&patches upstream, they’re trying hard to improve the status quo. Why the hate? Why not constructive criticism or even lending a hand?

    Is this all you can do when someone doesn’t do what you think is right? Is it so hard to make yourself distanced from the argument, take another look around, take a deep calming best, rethink your course of action and adjust it to something more constructing than a flaming rant?

    These are the moments where vocal individuals make me feel like a jack-ass for thinking that the OSS community is above such petty things.

    1. “enable Ubuntu to be shipped with all those little proprietary bits that are necessary to make the OS enjoyable by Joe User.”

      That’s exactly the wrong approach, and is often actively damaging to the F/OSS community.

      Ubuntu ships and encourages people to use proprietary graphics drivers; then ATI and NVIDIA decide to drop support for old models, and Ubuntu is screwed. They can’t patch the drivers for newer because, hey, they’re proprietary. If ATI/NVIDIA decide they’re not going to maintain them any more, no-one can. This *has already happened*, for both ATI and NVIDIA. Ubuntu had to do a rapid backpedal and switch to the free drivers instead for older cards.

      And don’t get me started on the proprietary driver Ubuntu shipped in their custom respin of UNR for Dell netbooks using Poulsbo graphics chips. That’s just a nightmare.

      Shipping ‘little proprietary bits’ is not, at all, the right way to do things.

      1. trunks says:

        So, users using an HD5xxx from ATI for example are supposed to just be pissed off that they don’t even get decent basic support, thus have to stick to windows where nobody cares about whether it’s proprietary or not?

        Now that’s an useful attitude.

        Many things *just don’t work* without proprietary software currently, so what’s so wrong about shipping it?

      2. Peter Webb says:

        “We will provide you a with a robust, well engineered, sleek desktop, which will not work with any degree of usability with any hardware less than two years old. Did we also mention that you cannot play pretty much all of the audio and video content out there? Now you should not only use and like this desktop, but also be thankful that we were gracious enough to provide you worthless worms to play with it. And don’t forget you can’t comment or criticize anything we do, unless you belong to our inner circle, contribute, think and behave like we do”

  56. Y’know, I honestly couldn’t give a shit who commits the most code to upstream GNOME. Canonical are making advances in bringing Linux to the mainstream by actually treating users as _people_ and not basement-dwelling men with neckbeards carrying satchels.

    The signs for the future are positive as Canonical works on Ubuntu Light and pushes for OEM deals so we can actually get Linux tied to hardware, so that when someone walks into a shop they can buy a computer with Ubuntu on it. Like what real people do.

    As far as the statistics, in my short time being involved in open source thus far, it appears that GNOME have these ridiculous design guidelines and their own ideas for Gnome Shell. They think that what Canonical is doing downstream doesn’t fit into these guidelines and so they don’t accept their work.

    Just look at the recent disapproval of Zeitgeist integration – stupid, stupid decisions made by people who have no idea what a real person wants out of their computer, and what sort of development will actually enhance the Linux desktop as a whole.

    1. Jef Spaleta says:

      Ubuntu Light…..

      Canonical has tough competition in that space already. DeviceVM has an instant-on solution called Splashtop..its linux-based. if you think DeviceVM is going to roll-over and play dead while Canonical goes after the same OEMs that DeviceVM is already providing services for…think again. Linux is tied to hardware already…just not Canonical’s variant of linux. If you haven’t used should try it.

      And about those up and coming Canonical OEM contracts. What OEMs exactly? The most popular Ubuntu OEMs are self-servicing niche entities who have no need for Canonical’s enginneering services. And for all the big players right now everybody and their brother is rushing to produce Android based ARM devices….while it seem’s Ubuntu’s ARM effort is rebooting/retooling into Linaro. We’ve already seen HP jump ship from Canonical after paying to have the Mi interface produced. I’m not so sure the future is so bright. It’s one thing to be hopeful, its another thing to be delusion. DeviceVM and Google are real _linux_ players in the OEM space that Canonical will have to compete with. It’s not clear at all that Canonical can execute OEM contracts well enough to win major contracts away from either.

  57. To me, it seems to be partly a business-model issue. Canonical’s desktop business model is built on differentiating themselves on the desktop *now.* It’s hard to differentiate yourself when you’re using the same vanilla GNOME that everyone has access to. So you differentiate yourself by tweaking the interface. Some people call this forking the interface.

    Red Hat doesn’t have to concern themselves with differentiating themselves on the desktop. They get their moolah from supporting servers. If people want to use Red Hat as a desktop workstation OS – they will support it, but it’s not their bread and butter.

    To me, this gives Red Hat more breathing room. They don’t have to make their desktop sexy (Mark Shuttleworth has actually said that he wants the desktop to have lustful, skin-deep beauty – ). Red Hat doesn’t have to worry about this – they can contribute to upstream GNOME and catch the benefits in a few years when they release their next version of RHEL. In the meantime, folks who want to can run Fedora for free.

    Those are my two pennies on the upstream GNOME contribution stuff I don’t think we should discount the bazillion bug reports made by Ubuntu users, though. Surely a good chunk of that helps out a wide range of upstream projects – not just GNOME.

  58. Albert says:

    I feel like I’m back in the 80’s and my spectrum is better than your C64.

    I thought the true nature of Open Source was that you could do what you want with no obligation or guilt. Strange that most of the posts I am reading here say the opposite.

    On the specific question of which is better users or developers I think this is not a valid question.
    No users, the code sits on the shelf and does nothing, so maybe having somebody bringing in new users is important.
    No developers – no code – so maybe having somebody committed to the developer community and investing there is important.

  59. NB45 says:

    I have been introduced into Linux 3 years ago by MS Windows. Yes, I typed that right. But that’s another story.

    My point being, it was Ubuntu 7 that did it. Of all the distros available then, this was the one that I could handle right from the get go. It’s just more intuitive. What did I knew then about all these FOSS contributions? Nada! I was just happy I found an alternative to Windows.

    Fast track to 2010 and my distro of choice is still Ubuntu, now at 10.04. Why? As a user, I had a feeling it belongs to me. It speaks to me. It understands what I wanted. It keeps me updated. It informs me. It calls me back when I wander. Go on and laugh, it’s all right. Sure it’s got some bugs. But I’ve used Fedora and it’s not like it doesn’t have any bugs either.

    Ubuntu simply has captured the user’s heart. You might say, “at the expense of other developers.” Ubuntu IS the user’s distro. Do not just blame Canonical’s marketing. Everywhere you look, you see write-ups about it. It’s being talked about over and over again. Journalists make it the de-facto benchmark of other distros. It is because of Ubuntu that I was even aware there is such a thing as Fedora or Mandriva. When enough media exposure has propagated Ubuntu, how do you think the world in general will view Ubuntu, if not look at it as “The Linux Desktop?”

    So yeah. From the point of view of FOSS developers, maybe Ubuntu/Canonical is a leech. But not when the non-technical user is concerned. Canonical knows its market and they hit the home run.

    You want more proof? This blog is one more fine example why Ubuntu and Canonical will continue to enjoy being the by-word when it comes to the Linux desktop.

  60. I used to be an Ubuntu fanboy, back when I was a GNU/Linux fanboy. Eventually, I grew out of those pants, and have settled into a different pair altogether. After being on every side of the fence possible, and many fences on that, here’s my take on your topic:

    I would love to see the GNU/Linux desktop market flourish. However, watching the market share stats for the past 8 years, it’s not going to happen. We’ve been at the bottom of the barrel for years, and if Canonical’s marketing can’t pull that number up, then who can? Red Hat (yes, it’s “Red Hat”, not “RedHat”. Red [space] Hat) has shown they don’t know how. Novell hasn’t made any headway after getting in bed with Microsoft. Mandriva? Nope. Linspire? Definitely not.

    I want to use Fedora. I really do. We run RHEL servers 100% at work. We purchased the RHEV software, on their beta program before it went GA. We are RHEL through and through, and it works wonders. RHEL is truly an amazing operating system.

    BUT, every time I try to run Fedora, so I can get the “heads up” on what the next RHEL will be like, I’m fighting with it all the way. I’m fighting with video drivers. I’m fighting with wireless drivers. I’m fighting with Yum and RPM. I’m fighting with a lack of packages. I fight with high loads regularly, high RAM consumption, and just about everything you can think of. Not just me. I would watch coworkers at a previous employment fight with Fedora weekly, even though we were primarily a Red Hat training contractor (and they’re all RHEL/Fedora fanboys).

    I want Fedora to succeed. I really do. But, from where I’m sitting, I don’t have the time to deal with all this breakage. I guess that’s what you get for running a bleeding-edge distribution, but I really, really need to get my work done. And I have submitted my bugs, as I’ve come across them, if I can’t fix them on my own. You can find them on Red Hat’s bugzilla.

    Then there’s the decisions to change your init system more often than your underpants. So, RHEL 5 is SYSVR4 (which is broken, and has never gotten fixed). RHEL 6 is Upstart. Will RHEL 7 be Systemd? As a RHEL system administrator, and RHCE, this is just silly. It means changing and updating thousands of documentation, creating errata out your ears, and pissing off your admins.

    Now, with all that said, I’m not running Ubuntu either. Stuff used to Just Work. At least back in 2005 and 2006, when they were mostly just a Debian clone. Now, their releases are broken more often then they are fixed. I don’t subscribe to the LTS philosophy (I blogged about why), and I don’t understand why wpa-supplicant would be needed in a server install. I have a lot of gripes with Ubuntu, but I keep them to myself, as I have signed their Code of Conduct, and even become an Ubuntu Member, which I value as much as my RHCE, and I don’t want to lose.

    So, where am I? I’m not running Fedora, because it breaks. I’m not running Ubuntu, because it breaks. I’ve tried openSUSE (yes, “openSUSE”, not “OpenSuSE”, or any other convoluted mixed case), but it’s far to overengineered to be productive (plus the packages I care about are too out-of-date). Right now, I’m on Debian GNU/Linux. I’m there because stuff works. I’m not fighting with my operating system, Debian Sid provides a near-bleeding-edge functionality, without risking the great breakage Fedora and Ubuntu ship. It’s contributed massively to the larger GNU/Linux ecosystem (ext3/ex4, alternatives, etc), it works closely with upstream, as it’s bugtracker reports bugs to the upstream developer when they are created, and a number of other reasons.

    While you bring up good points about Canonical riding off the coat tails of Red Hat, I just can’t run Fedora. Good for the large number of contributions to GNOME. Good for the large contributions to the Linux kernel. Good, good, good. But, as I’ve stated, Fedora just breaks too much for me to be of any value. Fedora fanboys will disagree, of course. “I’ve never had a problem” will be the reply, but if it works for Arch (which I’ve run), who doesn’t do _squat_ for distro-specific patches, or Slackware, then why is it broken in Fedora/Ubuntu, who has massive corporate entities behind it?

  61. Milan says:

    I agree with all that Greg said, although I knew this already, Red Hat devs contribute a lot to kernel for example, what about Ubuntu, anyone ever contributed something to kernel? Red Hat employed one of main nouveau dev to continue working on nvidia driver and later you have news articles about Ubuntu including nouveau driver and how good it is, like they have something to do with it, no word about Fedora.

    Now, I don’t have anything against Ubuntu, I am glad there is some distro that atracts people to Linux, I personaly use Gentoo, and when I am to install something quickly I use Fedora, or PCLOS and I don’t see a difference between those distros and Ubuntu, except, Ubuntu has those window buttons on wrong side, cripled notifications and non standard names/locations for some libs and utils, that is enough for me to install something else…

    Now, what they actually did, beside marketing, themes and wallpapers, maybe that new notification daemon which was realy, realy necessary, just what Linux needed, playing and moving things around, indicators?! There is no project that upstream wants to implement this hack, and finally what is cooking, a brand new font! Now, how cool is that 😉 And they have no plans of implementing gnome shell, so how exactly are they related to gnome upstream?

  62. Rob Falkner says:

    Let’s say that Ubuntu getting publicity as The Linux Distro isn’t just stealing mind-share away from other distros but that it actually gets Linux-based systems installed by OEMS and in school programs etc etc etc. Then people who are used to using a Gnome-based system come into the work force, doesn’t that open up a lot of space for Red Hat growth in the business/corporate sector? People have used a similar system, but it’s buggier/not as well designed as Red Hat when a company is researching who they want to go with for support. Isn’t that the perfect storm for Red Hat? If Ubuntu was less buggy, it wouldn’t be an opportunity for RHEL.

    1. A more technically accurate phrase would be “not as regression avoiding” and “does not maintain a long-term ABI/API.” Such things are important in organizations that want to deploy for 5+ years (sometimes as long as 10 years).

      Such sustaining engineering is not done cheaply, and that’s why organizations pay for it. The side benefit for the community is that more GPL software gets written as a result.

      I don’t condone everything Greg said in this article, and I think it’s a mistake to go tit-for-tat with “fanboys” of any kind, let alone project leaders that feel the need to bash the efforts of companies like Red Hat regularly. However, it is tiring when people are stating falsehoods about contributions and keeping the community’s long-term sustainability in focus.

      Why people feel the need to bash Red Hat is something I still haven’t figured out. Then again, it’s similar to what I see in the non-open source world — marketing and branding. Why are we dragging that over?

  63. It’s always fascinating to see that people who claim to promote the idea of Open Source always have a problem when others are actually taking their code and use it in their own products. If you’re so insistent on getting something back, you should write closed source software and sell licenses. If you’re serious about Open Source, you better stop the whining because you’re only discrediting yourselves.

    From a user’s perspective, I’m pretty sure that Canonical has done MUCH more for Linux, Gnome and everything else around that ecosystem than any other Linux project, except for maybe openSUSE. Why? They do the marketing and try to attract users that are NOT the geeky Linux stereotype. They want normal people to use that software. And Canonical is actually doing a great job at getting average users to give it at least a try.

    In case you’re still not getting it: Canonical is shipping and advertising a Linux desktop system that is usable by somebody who doesn’t give a damn about Free Software, source code contributions or any technical aspects whatsoever. This somebody could be your mom or your grandfather or your niece who just needs to surf Facebook but doesn’t give a damn about the technology behind it. But in order to attract such users, your product also has to look good (which is why fonts and fancy notification systems ARE a big thing, but the typical Linux nerd fails to realize that – the guys at Canonical instead actually grasp that concept).

    If you want to put it that way, then Canonical is the Apple of Linux Land. Red Hat is not. Like Microsoft, Red Hat is in the corporate business. So it shouldn’t be surprising that nobody is using Red Hat or Fedora on a desktop or notebook. Canonical went and took the market that apparently nobody else wanted.

    So does it matter whether Canonical actively contributes to the Gnome project or anything else? If you want to have the code that they’ve written, download it from their website. It’s there, they don’t mind you taking and re-using it. But they certainly don’t have to butter up and please any “community” developers and their arbitrary code contribution policies. Instead, they’re busy working on the next version of Ubuntu, which certainly is much more important for them and the rest of the world.

    1. “From a user’s perspective, I’m pretty sure that Canonical has done MUCH more for Linux, Gnome and everything else around that ecosystem than any other Linux project, except for maybe openSUSE. Why? They do the marketing and try to attract users that are NOT the geeky Linux stereotype. They want normal people to use that software. And Canonical is actually doing a great job at getting average users to give it at least a try.

      In case you’re still not getting it: Canonical is shipping and advertising a Linux desktop system that is usable by somebody who doesn’t give a damn about Free Software, source code contributions or any technical aspects whatsoever. This somebody could be your mom or your grandfather or your niece who just needs to surf Facebook but doesn’t give a damn about the technology behind it. But in order to attract such users, your product also has to look good (which is why fonts and fancy notification systems ARE a big thing, but the typical Linux nerd fails to realize that – the guys at Canonical instead actually grasp that concept). ”

      One, this is not a problem-free way of doing things. If the only way to make Linux ‘win’ is to ship it with a bunch of proprietary code and divorce it from the community concepts that get it written in the first place, that’s a pretty damn Pyrrhic victory. You’ve replaced one proprietary black box operating system with another. How has that helped anyone?

      Two, it’s rather arguable whether Canonical is, in fact, any good at the things you claim it’s good at.

      1. George Mitchelll says:

        Adam, The reality is, as you well know, there is a LOT of hardware out there, ESPECIALLY in the realm of wireless, that will ONLY work with proprietary code. That is not the fault of Ubuntu or any of the rest of us in the opensource world. But it is REALITY. Thus people, especially those with notebook type gear, need access to an in between world where they pay for part of it and still enjoy many of the advantages of opensource. My point would be that it really needs to be clearly defined as to which is which. This is why I like the Fluendo model. This is something that Mandriva has done to some degree with PowerPack, but I would really prefer that it be done third party like with Fluendo. The distro would then have their free download edition, but also a pay for download edition that contains the third party proprietary toolkits. Then the user could make a choice according to their needs and the distributor would not be compromised by having their devs working on proprietary stuff. That is how I feel about this issue as a user.

  64. This whole article is pathetic. Why such hostility?
    It does not matter which Linux distribution contributes most, or which distribution is most used or cool.
    The most important thing is that it is LINUX and it is FREE SOFTWARE.

    1. Dulwithe says:

      “It does not matter which Linux distribution contributes most”

      Well, it really does, because if the ONLY contributors to the linux/foss world were distros like Ubuntu, then there wouldn’t be much of an operating system and software ecosystem for anyone to market.

      Just think it through again and I expect you will agree…


  65. Joe says:

    Just thought I’ll put in my 2p in here.

    I’m not a developer. I’ve contributed nothing to Gnome or anything else. I’m just a plain ordinary user.

    I have used Linux in the past – namely Red Hat, Mandrake but found them a bit difficult so went back to Windows.

    I went back on Linux again when I was told about Ubuntu (Warty Warthog) and for once I found that I liked it enough to stick with it, dual-booting with Windows at first.

    I found myself spending more time on Ubuntu than Windows and about a year and half later, I now run Ubuntu only, no Windows at all.

    Did whoever contributed more to coding, be it Debian, Novell, Red Hat influence my decision here? No. It was Ubuntu Forums that helped the most – where I was able to get answers and I do like command line (in the early days it helps!)

    Frankly – community is what win users over and Ubuntu does have this. So while Red Hat have contributed the most code, I already knew that they contributed a lot, not just on Gnome but others such as the kernel – but this did not influence why I use Ubuntu.

    I went with where the community is. As an end-user this is important too. At the end of the day, that’s all that matter really isn’t it? Getting more users onto Linux? However I always knew that it was going to take a long time to switch users from Windows to Linux. As an end-user I tell people about Ubuntu. I show it to people and offer to help install and show the rope and so on. I do my little part in showing people about alternatives to Windows.

    That’s a good thing, no matter whatever choice of distro you use, be it Ubuntu, RedHat or anything else. I’m glad that RedHat is still spending money on coding and making Linux more user-friendly than before and better. I’m glad that Ubuntu/Canonical is doing marketing and supporting a community that allow people to get help with problems within their distro (Ubuntu forum for example).

    I guess I’m just saying that coding is not the only thing the benefit Linux. Hosting a community and making it as friendly as possible as well as helpful is just as important. If it wasn’t for the Ubuntu forums and more I might have given up on Linux a long time ago.

    But do allow some users to have some intelligence too – yes there are idiots everywhere, we all know that! 🙂 but quite a lot of users do realise that a lot of people/companies contribute to Linux overall, and quite a lot do know that RedHat contribute a lot, more than Canonical. It doesn’t prevent me from using Ubuntu though. I like it. I also like Fedora. I’m just comfortable with Ubuntu in the end, but you won’t hear me saying that Ubuntu is better than all others and that they invented everything, although they still can’t butter my toast.

    So thank you RedHat and many others that spend time developing – we might not say anything, but don’t think you’re unknown! 🙂 I just hope I made some sense above. Cheers!

    1. yusuf says:

      Well MR Joe. Although I agree with you. This sort of explanation doesn’t count by Gregdek. Since they a not intrested in consumers but in contributers.

    1. It’s not good for Canonical to profit off of either people’s work without any effort to put back what they took.

  66. Tomorrow, I’ll wonder someone will blame Debian to not contributing to Gnome. Stop percentage war. It is all about Linux folks!

  67. Paul Smyth says:

    This kind of biased one-sided article is precisely what Windows and Mac users laugh at GNU/Linux users for. “My dad’s bigger than your dad” etc.

    The GNU/Linux community, like a large corporation, each contributes different things to the party. In any corporation you will find Sales, Accounts, Services, Engineering etc. Each plays it’s part. As for Ubuntu’s contribution, Joe’s post is spot on.

    For a very long time I considered Gnome to be the most god-awful looking interface ever produced. I was convinced it was designed for babies. I used KDE exclusively for a long time. Redhat’s Blue Curve around Redhat 8.0 (pre-Fedora) was the first time it looked slightly better but still ugly. Novell did some great work with the computer menu and slab on improving it but still not enough to turn me from KDE. When I installed Ubuntu 8.04 to give it a try I went straight to Kubuntu, so determined was I that I wouldn’t touch Gnome but then with KDE 4.0 being such a pile of pants I gave gnome a go on Ubuntu. First of all I installed the computer menu and made it look like SuSE gnome and tolerated it. Upon upgrading to 9.04 I ended up reinstalling using the standard Ubuntu desktop. I left Gnome how it was and got used to it. Gradually with 9.10 it got much better looking then finally with 10.04 it was beautiful. Never before have I considered Gnome to be that but beautiful it is. My point? It doesn’t matter whether it’s 1%, 10% or 100%. If that 1% is of the highest quality then it deserves the highest praise. It doesn’t mean we’ve forgotten just how much hard work Redhat, Novell, IBM, SGI, Intel, Debian and all the others have put in. In fact it’s a compliment to them. It shows that their work is now so good that we’ve stopped noticing it. That in itself is high praise if you remember what went into getting a usable Linux desktop connected to the internet 15 years ago.

    Enough bickering, let’s just get on with it and respect each other’s choices and contributions. Not every contribution is code.

    1. elric says:

      I’d say Windows and OS X users are more likely to laugh because while Linux fans fight between them, desktop Linux is still going nowhere.

      1. Paul Smyth says:

        They point is there are members of the community who don’t WANT desktop linux to go anywhere because they want to be able to say that they’re guru’s who can configure this really complicated system that requires editing of text files rather than ticking a box in a gui.

        Ubuntu embraces newbies and as such has built a huge user community instead of aggressively flaming them when they ask basic questions that seem obvious to a Windows user. My 66 year old Mum asked me to put Ubuntu on her PC because she was so sick of Vista and has been over the moon with it. If she’d had to ask someone for help other than me she would’ve found a helpful community on the Ubuntu forums. I’d dread to think what would happen should she have found certain other mailing lists and forums.

        Just think of what could’ve been achieved with all the energy wasted in this debate. Year of the Linux Desktop? Never going to happen and that disappoints me because it’s there for the taking.

  68. George Mitchell says:

    Paul, Some of us here are just making comments as outside observers. I have never really used EITHER Red Hat or Ubuntu (other than the fact that I have tried them both) and as such really don’t have a dog in this fight OTHER than the fact that I want to see Linux move forward and not be hobbled by the actions of one particular player. I don’t “hate” Ubuntu at all. I rather wish them the best of success. But some of the things they are doing really bother me and remind me of the whole Linspire debacle. But it is impossible to discuss any of these things without getting a defensive and angry response. In short, I think that Greg has brought up a good point here which Mr Shuttleworth would do well to think about. He would also do well to think about emulating Mandriva and SuSE with a decent system control center. Not having that feature is the number one failing of Ubuntu as a desktop Linux OS at this point.

    1. Paul Smyth says:

      IMHO implementing Yast in Ubuntu would be the killer feature to win over Windows sysadmins and users. But I suppose they’d be accused of being parasites again then.

      I do have the same “Lindows/Linspire” concerns but then that was about mainstreamism too. Hmm, Michael Robertson – Mark Shuttleworth, is there a likeness? Yep, same as there is one to Gates, Balmer, Jobs & Ellison! Worth thinking about.

      1. George Mitchelll says:

        Paul, I am a Mandriva user since 1999 when Linux Mandrake was in its infancy. So thats where my loyalties lie. But I wouldn’t for one minute hold it against Canonical if they were to adapt the comprehensive Mandriva Control Center for their distro. I WOULD expect that they would work to improve it and build on it by adding new config tools AND, of course, share those improvements with Mandriva. This is how distros can work together and mutually benefit as a result. That kind of collaboration is not being parasitic. Its like bittorrent, you take your free download, but you give a bit back in return.

        My concern with Shuttleworth, is that like all of the others you mentioned, he tends to put marketing over technology. That achieves a whole lot in the short term and a certain amount of it is necessary. But, in the end, technology is what brings home the bread. This is especially true in the opensource world where vendors can’t lock in customers with IP claims. This is exactly why, if you want a distro with legs, you need little icons when you boot up that lead you to simplified tools that will get your system configured fast and easily, and a way to do that all in one place.

  69. George Mitchell says:

    More advice to Mr Shuttleworth would be that he needs to tell his Ubuntu users in the US which codecs are licensed out the door on his distro and which are not AND he needs to point them to Fluendo as a legitimate supplier of multimedia tools for Linux. OF COURSE he can not control the actions of individual users. But HE CAN point users in the right direction with regard to multimedia tools for Linux. A lot of Linux users are violating IP law unnecessarily because they don’t know that there is a legal option. That is too bad. It is the moral responsibility of the distros to point them in the right direction.

      1. George Mitchelll says:

        Thats a good thing. Unfortunately, it seems that some Ubuntu users aren’t aware of Fluendo. This is not only about codecs, its about other Fluendo products as well, like a fully licensed DVD player for Linux that works out of the box.

  70. Craig says:

    23% of contributions – more than any company – are coming from volunteers. What distribution are these volunteers using? It’s very likely that a notable portion of them are using Ubuntu, it’s a popular distribution. OK, perhaps a large number of the Ubuntu population are “users” but even technical people like myself want stuff to just work, so it’s not unreasonable to expect that some portion of that 23.5% are Ubuntu users.

    [I’m also curious about the 16% gap called “unknown”… seems pretty large and begs one (me) to read it again to try and understand what is going on there]

    If these volunteers are using Ubuntu, or draw in because of it, does Canonical not deserve some recognition here for packaging a distribution into something desirable and accessible, and for all the infrastructure they have put in place, and the love of the media and popular word?

  71. Craig says:

    Also, reading through all the comments, it seems there are suggestions of Canonical/Ubuntu having a different vision, of them making concessions that purists don’t find acceptable (like closed video drivers until the open source ones work, or questionable but not underhanded choices to distribute portions of windows drivers [rightfully paid for by hardware purchases] to make WiFi work), of their work not benefiting upstream, of their patches and ideas being rejected upstream… upstream being very much RedHat in this case.

    From where I’m standing, it’s not really clear whether the issue is that Canonical is not contributing or not working well with up-stream, or if up-stream doesn’t work well with those who have a different vision or approach to solving the issues, or simply that Canonical is doing a lot of indirectly related work that up-stream doesn’t feel there is a benefit to.

    If you’ve got a problem with they way doing things – then do something else – and do it better. Right now – people are happy with the value added that Canonical is adding on top of Linux. If this value added is desired — then up-stream better get better at finding a way to integrate these ideas into their work (whether by accepting their patches, or moving the desktop components in their direction). If you feel it is completely wrong — then another distribution better step up and start delivering the value that Canonical supposedly is with the values you think are more sustainable and push that.

    1. George Mitchelll says:

      Craig, I think it is very clear that Greg’s point is that upstream is Gnome, NOT Red Hat. The reality is that Ubuntu as a distribution owes its life to Gnome. Ubuntu uses it ubiquitously. That would indicate that they should be taking more of an interest in it. IF they are submitting patches that are being rejected, then at least they are trying. But if they are simply not taking a serious interest, then that is a problem, for themselves more than for anyone else. In the opensource world, as the saying goes, “what goes around comes around”.

      As for the wireless drivers issue. That is a huge problem that Ubuntu should also be taking more of an interest in than just taking a short cut by taking the wrapper route. ALL of the desktop distros should be looking for a way to address these ridiculous hardware issues and Fluendo provides a pretty good model for dealing with those things. There really needs to be an arms length shop that provides native Linux drivers for proprietary hardware under NDA with the hardware vendors. This software needs to be kept rigorously separate from the GPL core product. In that way the purists and the utilitarians can live in peace with each other. Nobody detests software patents more than I do, but they are a fact of life and we have to learn to live with them until the point at which, hopefully, they are eventually thrown out. Mixing things up the way Canonical is doing, once again, is just taking a short cut rather than really getting out there an grabbing the bull by the horns and solving the problem once and for all with legitimate contracts and agreements.

      1. WTF!?

        Ubunut owes its life-blood to Debian, no ifs or buts about it…just like a dozen other wannabe child-distros. GNOME or any other half decent DE is just the icing on the virtual cake. Without Debian providing the backbone, where would Ubuntu be?

      1. Dev says:

        Nop, i am not Mark.

        You read the article?, is only butthurt, and that is bad. Is unusual that a member of Fedora attack other distro, Fedora is great, but this guy… is only butthurt.

  72. People. Red Hat, Canonical, Fedora, Mandriva, GNOME, KDE… all contribute to make a better Linux experience.
    Percentages? attack other distros? a dude anger with canonical? My god, sound like “Apple vs Windows”.

    Relax, all create great things.

    1. In this case, Canonical isn’t creating anything and is profiting off other people’s work. There’s a fine line between “they create great things” and “you work for free and I make money, mmk?”

      1. Luke says:

        Welcome to the GPL.

        You’re blind if you think Ubuntu hasn’t contributed to the Linux ecosystem. Like Cont3mpo said, there are different ways to contribute. Ubuntu works damned hard to be as user-friendly and good-looking as it can. It MARKETS LINUX WELL. It has just about the largest user community of all the distros.

        One cannot deny that Ubuntu has brought many users and developers to the GNU/Linux world.

  73. […] (16%), Novell (10%), Collabora (5%), Intel (3%) e Fluendo (2%). A Canonical foi objeto de algumas criticas porque sua contribuição é de apenas 1%. Fonte Barrapunto This entry was posted in […]

  74. […] Greg DeKoenigsberg пишет: I’ve been a Debian maintainer. I deal with Ubuntu desktops for my friends regularly. And, frankly, have you done Fedora v. Ubuntu when it comes to upgrades? Sometimes people argue Debian APT-DPKG, and the strong Debian Packaging …. Also on the size argument – Mandriva, a company smaller than Canonical, with a lot less money, and which everyone considers to be a KDE distro, still is right in their ballpark (and actually right on the same number, if Pavol’s right). … […]

  75. With the risk of pouring gas on the fire: why would some use GNOME, anyway? If one wants an interface created on the “I know better, here is what you’re allowed to do” principle, one would better go to Windows, this this way one can play good looking games.

    1. Veneficium says:

      Why would some use KDE, anyway? It’s a bloated, messy and cheesy desktop created on the “I wanted to copy Windows but failed miserably” principle, one would go better to Windows, then this way one can use applications that doesn’t start with K.

  76. Canonical is in a dangerous position with such little upstream presence. I discuss this more on my blog, but if a key project with a fast moving development timeline moves away from what Canonical considers “ideal”, they’re pretty much in a hole. Canonical does not have the upstream presence to push a project back to what they need in the case of one shifting goals.

    For instance, if the Linux Kernel moves away from what Canonical considers ideal, there’s really only two solutions and each are equally as messy. The idea solution would be use their upstream presence to shift the goal, but since Canonical has no influence, they are at the mercy of those developing above them.

  77. Here’s a thought on the ‘oh, but Canonical contributes in other ways’ argument.

    At Red Hat, I’m a QE Engineer. Red Hat also employs Software Engineers. Those dudes get paid more than me. Know why? We’re a software company. Their contribution is more important than mine.

    Take a big software project. It’s got developers, QA people, artists, writers, and support people. Now, take out all the QA people, artists, writers and support people. What have you got? A software project. Maybe it has more bugs than it otherwise would have, maybe it doesn’t look as good, maybe it isn’t as well documented, but you still have a software project. Keep the QA people, artists, writers and support people, but take out the developers, and what have you got? Jack shit is what you’ve got. Everyone’s making a contribution, but the code is the most *important* contribution.

    Want an analogy? Take a Yankees game. It gets played in a big stadium. There’s the players. Then there’s a couple of managers, umpteen assistant managers and trainers, the bat boys, the water boy, the stadium announcer, the ticket takers, the hot dog vendors, the guys in crazy costumes, and all that jazz.

    You can take away the managers and the water boy and the announcer and the hot dog vendors and all that jazz and put the players in a corn field somewhere, and you still have a Yankees game. If you take away the players and keep the stadium and the managers and the water boy and everyone else, you have jack squat. Everyone’s contributing to the event, but the players make the only truly vital contribution.

    1. But Adam 🙂 You speak of artists but not designers. You take designers out of the equation, and you have software, and it runs, but maybe it does nothing that does anybody any good. Or it does something useful but in such a painful manner you’d rather scoop your eyeballs out with a spork than use it (oh I can think of a few examples 🙂 )

      Of course I’m going to defend designers; sorry I can’t help it!!! Back to your regularly-scheduled program.

      1. sorry, mo, but people wrote in *edlin*. if you need to do a job, and there’s a tool that does the job, you’ll use it, even if it looks like a cactus, comes with no manual, and feels horrible to use. If someone gives you a picture of a perfectly designed tool to do the same job, what can you do with it? It’s a picture. 🙂

      2. Jef Spaleta says:


        You just described every piece of code I write for work. I am a cactus farmer.


  78. Greg says:

    Excuse my ignorance, but out of all the Linux companies producing distributions, Red Hat, Novell, Mandriva, Canonical, i wont add Oracle cause they dont produce a commonly used distribution & i hope im not forgetting someone, are there any others that produce proprietary software besides Canonical?
    And how can a company like that have demands from the FOSS community, especially from a GNU project?

    From my point of view whats wrong here is:

    a) Bug #1 which states on “Mark’s” own words: “Non-free software is holding back innovation in the IT industry, restricting access to IT to a small part of the world’s population and limiting the ability of software developers to reach their full potential, globally. This bug is widely evident in the PC industry.”
    While at the same time his company produces non free software:

    b) Ubuntu has a large userbase.

    1. Don’t forget landscape.

      btw this is an interesting discussion:

      “Ubuntu should de-emphasize freedom and openness (because it’s not delivering that), and emphasize usability (because it’s trying to deliver that). What I don’t like is that everything about Ubuntu marketing emphasizes “Humanity to others,” help the downtrodden, uplift the oppressed, use Free Software etc., when they’re trying to make money by selling proprietary software / services. I feel like that’s kind of cynical of them.”

  79. Nice Post With Huge Comments I like People That Post Lagit Comment On Blog And I Found Here That Every One Post The Real Comment No Spam Or Stupid Comments Here Good Blog, Post, And Also Comment Defenatly Worth Of Read..

  80. I wanted to point out this conversation which is really interesting:

    It seems Jono Bacon, who is a community leader for Ubuntu and who has spoken up in Canonical’s defense in response to this blog post, was under the misconception that contributing to something you are the ultimate upstream for is ‘contributing to upstream.’ It was pointed out to him that this is not the case as you cannot contribute ‘upstream’ to something you are the upstream of (just as you really can’t ‘contribute to charity’ when the charity is actually for your own benefit).

    Maybe part of the bad feelings here is because Canonical earnestly believes they are contributing ‘upstream’ by a definition of ‘upstream’ not shared by non-Canonical folks. Perhaps if Canonical would clarify which codebases they contribute to for which they are not the upstream, it would help them understand the complaint here a bit better.

Comments are closed.