Competition is tough on the contestants, but it gets great results for everyone else. We like competitive markets, competitive technologies, competitive sports, because we feel the end result for consumers or the audience is as good as it possibly could be.

In larger organisations, you get an interesting effect. You get *internal* competition as well as external competition. And that’s healthy too. You get individuals competing for responsibility, and of course you want to make sure that the people who will make the wisest choices carry the responsibilities that demand wisdom, while those who have the most energy carry the responsibilities for which that’s most needed. You get teams competing with each other too – for resources, for attention, for support elsewhere, for moral authority, for publicity. And THAT’s the hardest kind of competition to manage, because it can be destructive to the organisation as a whole.

Even though it’s difficult to manage, internal competition is extremely important, and should not be avoided out of fear. The up side is that you get to keep the best ideas because you allowed them to compete internally. If you try and avoid it, you crowd out new ideas, and end up having to catch up to them. Usually, what goes wrong is that one group gets control of the organisation’s thinking, and takes the view that any ideas which do not come from within that group are a threat, and should be stopped. That’s very dangerous – it’s how great civilisations crash; they fail to embrace new ideas which are not generated at the core.

In Ubuntu, we have a lot of internal competition. Ubuntu and Kubuntu and Xubuntu and Edubuntu and *buntu-at-large have to collaborate and also, to a certain extent, compete. We handle that very well, I think, though occasionally some muppet calls Kubuntu the blue-headed-stepchild etc etc. It’s absolutely clear to everyone, though, that we have a shared interest in delivering ALL these experiences together with as much shared vision and commonality as possible. I consider the competition between these teams healthy and constructive and worth maintaining, even though it requires some fancy footwork and causes occasional strains.

The challenge for Gnome leadership

The sound and fury writ large in blog comments this week is all about how competition is managed.

Gnome is, or should be, bigger than any of the contributing individuals or companies. Gnome leadership should be in a position to harness competition effectively for the good of the project. That does, however, require very firm leadership, and very gutsy decisions. And it requires vigilance against inward thinking. For example, I’ve seen the meme reiterated multiple times that “one should not expect Gnome to embrace ideas which were not generated and hosted purely within Gnome”. That’s chronic inward thinking. Think of all the amazing bits of goodness in the free software stack which were NOT invented in Gnome but are a part of it today. Think how much better it is when goodness is adopted across multiple desktop environments, and how much harder it is to achieve that when something is branded “K” or “G”.

When we articulated our vision for Unity, we were very clear that we wanted to deliver it under the umbrella of Gnome. We picked Gnome-friendly technologies by and large, and where we felt we needed to do something different, that decision required substantial review. We described Unity as “a shell for Gnome” from the beginning, and we have been sincere in that view. We have worked successfully and happily with many, many Gnome projects to integrate Unity API’s into their codebase.

This is because we wanted to be sure that whatever competitive dynamics arose were *internal* to Gnome, and thus contributing to a better result overall in Gnome in the long term.

We’ve failed.

Much of the language, and much of the decision making I’ve observed within Gnome, is based on the idea that Unity is competition WITH Gnome, rather than WITHIN Gnome.

The key example of that is the rejection of Unity’s indicator API’s as external dependencies. That was the opportunity to say “let’s host this competition inside Gnome”. Even now, there’s a lack of clarity as to what was intended by that rejection, with some saying “it was just a reflection of the fact that the API’s were new and not used in any apps”. If that were the case, there would be no need for prior approval as an external dependency; the rejection was clearly an attempt to prevent Gnome applications from engaging around these API’s. It’s substantially failed, as many apps have happily done the work to blend in beautifully in the Unity environment, but there has been a clear attempt to prevent that by those who feel that Unity is a threat to Gnome rather than an opportunity for it.

Dave Neary has to his credit started to ask “what’s really going on here”?

In his blog post, he quoted the rationale given for the rejection of Canonical’s indicator API’s, which I’ll re-quote here and analyze in this light:

it doesn’t integrate with gnome-shell

That’s it – right there. Remember, this was a proposal for the indicator API’s to be an *external* dependency for Gnome. That means, Gnome apps can use those API’s *optionally* when they are being run on a platform where they are useful. It has NOTHING to do with the core Gnome vision. External API’s exist precisely BECAUSE it’s useful to encourage people to use Gnome apps on all sorts of platforms, including proprietary ones like Windows and MacOS and Solaris, and they should shine there too.

So the premier reason given for the rejection of these API’s is a reason that, as best we can tell, has never been used against an external dependency proposal before: “it’s different to Gnome”. At the heart of this statement is something deeper: “it’s competition with an idea someone in Gnome wants to pursue”.

What made this single statement heartbreaking for me to see was that it spoke clearly to the end of one of Gnome’s core values: code talks. Here we had API’s which were real, tested code, with patches to many Gnome apps available, that implemented a spec that had been extensively discussed on This was real code. Yet it was blocked because someone – a Gnome Shell designer – wanted to explore other ideas, ideas which at the time were not working code at all. There’s been a lot of commentary on that decision. Most recently, Aaron Seigo pointed out that this decision was as much a rejection of cross-desktop standards as it was a rejection of Canonical’s code.

Now, I can tell you that I was pretty disgusted with this result.

We had described the work we wanted to do (cleaning up the panel, turning panel icons into menus) to the Gnome Shell designers at the 2008 UX hackfest. McCann denies knowledge today, but it was a clear decision on our part to talk about this work with him at the time, it was reported to me that the conversation had happened, and that we’d received the assurance that such work would be “a valued contribution to the shell”. Clearly, by the time it was delivered, McCann had decided that such assurances were not binding, and that his interest in an alternative panel story trumped both that assurance and the now-extant discussions and spec.

But that’s not the focus of this blog. My focus here is on the management of healthy competition. And external dependencies are the perfect way to do so: they signal that there is a core strategy (in this case whatever Jon McCann wants to do with the panel) and yet there are also other, valid approaches which Gnome apps can embrace. This decision failed to grab that opportunity with both hands. It said “we don’t want this competition WITHIN Gnome”. But the decision cannot remove the competitive force. What that means is that the balance was shifted to competition WITH Gnome.

probably depends on GtkApplication, and would need integration in GTK+ itself

Clearly, both of these positions are flawed. The architecture of the indicator work was designed both for backward compatibility with the systray at the time, and for easy adoption. We have lots of apps using the API’s without either of these points being the case.

we wished there was some constructive discussion around it, pushed by the libappindicator developers; but it didn’t happen

We made the proposal, it was rejected. I can tell you that the people who worked on the proposal consider themselves Gnome people, and they feel they did what was required, and stopped when it was clear they were not going to be accepted. I’ve had people point to this bullet and say “you should have pushed harder”. But proposing an *external* dependency is not the same as trying to convince Shell to adopt something as the mainstream effort. It’s saying “hey, here’s a valid set of API’s apps might want to embrace, let’s let them do so”.

there’s nothing in GNOME needing it

This is a very interesting comment. It’s saying “no Gnome apps have used these API’s”. But the Gnome apps in question were looking to this very process for approval of their desire to use the API’s. You cannot have a process to pre-approve API’s, then decline to do so because “nobody has used the API’s which are not yet approved”. You’re either saying “we just rubber stamp stuff here, go ahead and use whatever you want”, or you’re being asinine.

It’s also saying that Unity is not “in GNOME”. Clearly, a lot of Unity work depends on the adoption of these API’s for a smooth and well-designed panel experience. So once again, we have a statement that Unity is “competition with Gnome” and not “competition within Gnome”.

And finally, it’s predicating this decision on the idea being “in Gnome” is the sole criterion of goodness. There is a cross-desktop specification which defines the appindicator work clearly. The fact that KDE apps Just Work on Unity is thanks to the work done to make this a standard. Gnome devs participated in the process, but appeared not to stick with it. Many or most of the issues they raised were either addressed in the spec or in the implementations of it. They say now that they were not taken seriously, but a reading of the mailing list threads suggests otherwise.

It’s my view that cross-desktop standards are really important. We host both Kubuntu and Ubuntu under our banner, and without such collaboration, that would be virtually impossible. I want Banshee to work as well under Kubuntu as Amarok can under Ubuntu.

What can be done?

This is a critical juncture for the leadership of Gnome. I’ll state plainly that I feel the long tail of good-hearted contributors to Gnome and Gnome applications are being let down by a decision-making process that has let competitive dynamics diminish the scope of Gnome itself. Ideas that are not generated “at the core” have to fight incredibly and unnecessarily hard to get oxygen. Ask the Zeitgeist team. Federico is a hero, but getting room for ideas to be explored should not feel like a frontal assault on a machine gun post.

This is no way to lead a project. This is a recipe for a project that loses great people to environments that are more open to different ways of seeing the world. Elementary. Unity.

Embracing those other ideas and allowing them to compete happily and healthily is the only way to keep the innovation they bring inside your brand. Otherwise, you’re doomed to watching them innovate and then having to “relayout” your own efforts to keep up, badmouthing them in the process.

We started this with a strong, clear statement: Unity is a shell for Gnome. Now Gnome leadership have to decide if they want the fruit of that competition to be an asset to Gnome, or not.

A blessing in disguise

Aaron’s blog post made me think that the right way forward might be to bolster and strengthen the forum for cross-desktop collaboration:

I have little optimism that the internal code dynamics of Gnome can be fixed – I have seen too many cases where a patch which implements something needed by Unity is dissed, then reimplemented differently, or simply left to rot, to believe that the maintainers in Gnome who have a competitive interest on one side or the other will provide a level playing field for this competition.

However, we have shown a good ability to collaborate around FD.o with KDE and other projects. Perhaps we could strengthen and focus our efforts at collaboration around the definition of standards there. Gnome has failed to take that forum seriously, as evidenced by the frustrations expressed elsewhere. But perhaps if we had both Unity and KDE working well there, Gnome might take a different view. And that would be very good for the free software desktop.

189 Responses to “Internal competition is healthy, but depends on strong and mature leadership”

  1. Flopsy Says:

    Oooh, Good read.
    Am enjoying testing Unity and am looking forward to what it brings in the future!

  2. luca Says:

    Aaron’s blog post made me think the same, cross-desktop collaboration is a vital aspect for the foss ecosystem, should be something more tangible and on top of all interests.
    What bother me about Gnome3 is conservatism and obstructionism towards external contributions, I hope, at least, they will not break Gnome Hig guidelines (like in Nautilus3).

  3. Alex Says:

    Interesting read!

    Mark, how about the licensing (LGPL / CCA) complaints?

  4. sect2k Says:

    From all of your responses in the last few days it seems to me you suffer from the same syndrome I’ve seen in a lot of CEO/owners, the illusion that their company is the absolute best and that their employees are the cream of the crop and can do no wrong, and when something doesn’t go as expected it’s naturally someone else’s fault.

    Now here is the hard truth, your employees are human, just like the rest of us, they make mistakes, not all of their ideas are the best and sometimes they are wrong and the same applies to you as well.

    The first step in resolving this situation is to acknowledge the fact that there were mistakes made both sides and stop pointing the blame towards GNOME exclusively. So they don’t agree with everything you do and that is fine, we don’t always have to agree on everything, you will just have to live with the fact. I personally don’t agree with lots of your (as in Canonical) decisions, and from what I’ve read I’m not the only one, but I’ve gotten used to the fact that when you make up your mind, no amount of reason will change it.

    Let he who is without sin, cast the first stone, it’s just you are not that person, and I say that with utmost respect for what you have done.

  5. Andrew Ampers Taylor Says:

    A good approach and what you say makes a lot of sense It seems like Gnome is falling into the pitfalls of so many corporates and conglomerates in the business world. Of whom the senior staff all seem to go by the idea of “Do unto others as they might do unto you – but do it first!”

    You definitely need to sit down and think this through, “‘n boer moet ‘n plan hê” springs to mind!

  6. pt8 Says:

    “The first step in resolving this situation is to acknowledge the fact that there were mistakes made both sides and stop pointing the blame towards GNOME exclusively.”

    Would you mind telling that to gnome people too who love to pick up Canonical everytime and put blame on the latter exclusively? I did not see this comment of yours on any of the blog posts from gnome people which they keep blaming canonical. They are also humans, they also make mistakes.
    I also dont agree with many decisions, which doesnt mean I will blast on them.

  7. Jeff Waugh Says:

    Mark is deliberately misleading the community about the way GNOME works, and the way he has attempted to engage with GNOME. I am writing a post about all of this at the moment, and will demonstrate that what Mark says in this post is essentially a lie.

    Despite liking Mark and loving Ubuntu, I find his current behaviour and attitude towards collaboration incredibly disappointing.

  8. mark Says:


    I don’t think there are any GPL / LGPL issues. We’ve happily relicensed stuff LGPLv2.1 recently, for example, to facilitate its inclusion in ZeitGeist and Gnome. The fact that we could do that speaks to the value of contribution agreements, which I’ll articulate more fully in a long-overdue blog post, soon.

  9. Peter Says:

    Ubuntu is not Debian. Sometimes there are valid reasons to reinvent the wheel.
    Also, to most users, this kind of conflicts are meaningless. They are very relevant to developers but, correct me if I’m wrong, the focus of Ubuntu is the USERS’ experience on Linux. It’s “Linux for Human Beings” not “Linux for developers with egos”

  10. daniels Says:

    ‘When we articulated our vision for Unity, we were very clear that we wanted to deliver it under the umbrella of Gnome.’

    That might be your problem right there: the messaging (at least from my uninvolved point of view) seems to have been ‘here’s a fully-formed piece of software, take it and call it GNOME’. For all gnome-shell’s failures in this regard, of which there have no doubt been more than a few, I don’t think you can really blame GNOME even primarily, let alone single-handedly, for not relying on a fully-formed, externally-developed[0] brand new UX and using that to execute a 180° turn in GNOME UX.

    Oh, and if I read the Unity page right, it requires copyright assignment to Canonical. For obvious reasons — see Michael Meeks’s excellent talk from Plumbers last year if you haven’t done so already — this is an absolute non-starter. It’s also the sole reason I’ve never touched GRAIL or GEIS beyond their architecture documents while working on multitouch, and will not unless their policies magically change to exclude copyright assignment.

    It’s sad that a company I left with a strong ‘upstream first’ commitment — I believe the phrase of choice was ‘upstream is our rock’ — and ethos has degenerated into this, where productive engagement with upstream has been replaced by ‘take it or leave it’, and then blaming upstream if the decision isn’t ‘take it, immediately, no questions’.

    Oh well.

    [0]: Just to be clear here, I don’t mean ‘not by Red Hat’, I mean developed outside of GNOME fora, using external infrastructure, with a single distribution as its primary/sole target, etc.

  11. Arnoldas Says:

    I’m not a fan of Unity, but I won’t kill myself if I have to use it in some cases, but Gnome-Shell is a big disappointment to me. I feel like something good reached the end. I don’t know the reasons why it happened, but I guess it’s because of lack of vision to make something usable to others not only to fulfill some egocentric ambitions. I hope that I am wrong in my thoughts and Gnome will continue to be a great Desktop as it was for years. I am, like Fedora (and Ubuntu) user, really hope that, because I don’t want to see Fedora dragged down by choosing crazy GUI.

  12. eet Says:

    With all due respect to Ubuntu; yur doin it wrong!

    Your proposals for the future direction development of GNOME have failed. As you are now pursuing your ideas ALONE instead of adopting GNOME 3, what you are doing is building a FORK; it certainly is not WITHIN GNOME.

  13. Phil Says:

    > Competition is tough on the contestants, but it gets great results for everyone else. We like competitive markets, competitive technologies, competitive sports, because we feel the end result for consumers or the audience is as good as it possibly could be.

    Yes, I did read (most) of the article, but I’m only going to respond to the first part: Please acknowledge just how much of a generalisation that first paragraph is.

    1) Competitive markets very often fail to provide the best results for all or even most people; the current economic state of the entire western world is proof enough of that. Competition allows for things like biggest > best, which are of no use to anyone.
    2) “We” often don’t like it anyway. Most people realise competition has its uses, but to just state that “we” like it is ridiculous, particularly when talking to a section of society that has explicitly chosen cooperation as its preferred means of progress.

    That sort of rhetoric really does not help when talking to an educated audience.

  14. mark Says:


    The problem you’re referring to is a problem of regulation, not competition. It’s the job of the regulator to set the terms of competition and the playing field on which it happens. Regulators across the Western world failed to ensure that financial system participants had sufficient buffer, and thus the financial system failed to handle a shock. The same is true in many cases where there’s a market failure. Regulation, like leadership of internal competitive dynamics, is very hard. But the alternative (central planning) has all the difficulties and none of the benefits.

    You say “a section of society that has explicitly chosen cooperation as its preferred means of progress”, but I think you mistake the situation. You often see statements like “let the best code win” which is essentially competitive. We do like the idea of collaboration AND competition, or coopetition, but there’s definitely competition in free software, and that’s a good thing.

  15. Seb24 Says:

    Unity like a Freedesktop project it’s a really good idea.

  16. mark Says:


    I am not misleading anybody.

    We spoke at length about this a few weeks ago and it was clear you had no knowledge of any of the background whatsoever. Please don’t let that stop you from expressing an opinion 😉


  17. eet Says:

    @Mark: (Note that I’m not part of GNOME development in any way.) I think that nobody would mind competition, we do have competition between KDE and GNOME for example. But competition must happen _between_ project-teams (or companies, if we want to stay in your parable) not _within_ project-teams.

    If you want to compete, call your desktop environment Ayatana, not GNOME, and market it as a competitor. Quit the rhethoric.

  18. mark Says:


    Different companies operate under different constraints. We’ve done some work in public from the start, and others in confidence, with customers and partners. That’s the nature of the game, and to deny that is to deny what’s interesting about a diverse ecosystem. I’ll note that you left Canonical to work at Nokia and Meego, both of which have done some work first in the open and some work first behind closed doors.

    I’m also explicitly NOT suggesting that Gnome should do a 180 degree turn, or drop work they’ve done. I’m saying that Gnome will be a stronger organisation in the long term if it finds ways to embrace ideas and work that start outside the core Gnome cabal.

    As for contribution agreements, I’ll address that in a separate blog. In short, though, the fear and loathing of such agreements is misplaced and weakens the ecosystem of contributors and investors in free software. You’ll note that none of the champions of that fear and loathing have a clue about what attracts investment, and many of the top projects that have sustained their pace of development against their competitors – Qt, MySQL, Firefox(thanks Dave), have such agreements. I note that you contribute heavily to a project which is not under a copyleft license, X, so it can’t be that you’re so totally opposed to the existence of proprietary versions of free software. And I’m sure you’re aware of RMS’ commentary on dual-licensing, which makes the clear point that such work is not unethical and is a net win for free software.

  19. Jeff Waugh Says:


    If you define “knowledge of the background” as “acceptance of Mark’s version of events”, sure. But the root cause of your inability to collaborate with GNOME is still a huge toss-up between incompetence and malice. The spin and flagrant mischaracterisations both you and Jono have spat out over the last year, all the rationalisations, just make it easier and easier to believe it’s malice.

  20. Jeff Waugh Says:


    If you understood the background to Canonical’s community interactions, you would not be attributing negative intentions to quotes from GNOME developers that you clearly do not understand. To read the libappindicator module proposal discussion as “purposefully blocking a Canonical contribution” is absolutely laughable.

    Truly independent communities don’t operate on your terms or whims, Mark.

  21. mark Says:


    No, I define “knowledge” as “knowing the facts of the matter”. I’ve laid out the facts as I see them and your response is to allege misleading spin and mischaracterization on my part and that of Jono.

  22. Mark not Shuttleworth Says:

    I’m a relative new-comer to Ubuntu, and have used it exclusively on my netbook a little over a year now. I’m completely agnostic to the Gnome/Unity debate other than the tidbits I’m picking up here and there.

    When Unity first installed on my netbook I was immediately struck by how much real-estate I was *losing*. Which is strange because all that I’ve read seems to indicate that I’m supposed to get more. But what I got was a side bar with relatively large icons that wouldn’t budge. It feels like it’s designed for widescreen displays.

    I also found that the overall experience was much slower than the standard Gnome Desktop – perhaps that’s entirely subjective, but that’s how it felt.

    In the end, it simply didn’t work for me on the netbook. But that’s ok; I switched to Gnome Desktop and installed Cairo Dock (which I love), so I’m happy again.

    I’ve seen some preview screencasts of the forthcoming Unity, and there’s a feature which I’m sure has been adopted to save real-estate, but which I feel takes away from an application’s usability. And that’s putting all the application menus at the top of the screen, and making it context sensitive depending on which application currently has focus.

    There are two reasons I think this is unfriendly: it makes you move your attention to another part of the screen – on a large screen that can be a big distance; it only appears when you move your mouse over the bar – so you’re looking for something that’s not there, unless you know it’s going to appear. I never liked this menu/app separation on the Mac, and I don’t really like it here either.

    Thanks for putting your side Mark, but I don’t think I’ll be switching to Unity any time soon, but then, thanks to this being a Linux based system I can choose how I want it to work for me.

  23. frustphil Says:


    “But the root cause of your inability to collaborate with GNOME is still a huge toss-up between incompetence and malice.”

    Ah so your GNOME is collaborative now :-)

    Why don’t you address the real problem huh, Jeff?

    Your hate(or whatever it is) against Mark is disturbing. Did he dump you? *Kidding aside*

    Seriously, why are these people denying their fault when it is very obvious already. Everyone is expressing the same concern. Do they really think the community are idiots?

  24. Daeng Bo Says:


    I know that people are up in your shorts about “competing with” GNOME: don’t sweat it. Stick with FD.o specs (and contribute where they are lacking) and make the best desktop your team can. Ubuntu has its problems. I’d guess you know what they are better than anyone and have a problem to solve them using Unity. If Unity ends up being significantly better than GNOME 3, people will switch in droves. From the looks of both, I expect that to be the case.

    I know you get lots of unsolicited advice: I’m going to give you a bit more. Make Ubuntu a development platform. Make it easy for new devs — give them an IDE with Launchpad integration, a language, and a toolkit to use with copious tutorials using Unity and the core libs. Look at QT’s dev website (or Apple’s or MS’s) for inspiration.

    A user since Warty’s first days and regular critic

  25. Daeng Bo Says:

    … and have a /plan/ to solve them using Unity …

  26. eet Says:

    @frustphil: Contrary to your impression, Mark isn’t the community…

    The problem is one of Mark trying to imitate Steve Jobs, not one of the GNOME community.

    If you want to follow your leader, please do so and hype evey whimsical Ayatana-‘innovation’ Ubuntu comes up with. But again: What Mark does has nothing whatsoever to do with GNOME, and if he doesn’t like GNOME and how it works, he got all the right in the world to run his own business (Ubuntu) the way he sees fit.

    Ayatana isn’t GNOME.

  27. eet Says:

    Just read a good blog-post about all this:

  28. daniels Says:


    Yes, Nokia under OSSO/Maemo/MeeGo (though I left Nokia two and a half years ago, and haven’t worked on MeeGo for several months) produced a mix of open and closed work; some of the latter was later destined for upstream, some was not. When we had submissions rejected by upstream, due to not having communicated our intentions well enough and having given upstream an opportunity to influence our work, we bucked up and worked on it until it was acceptable to all concerned. As opposed to simply blaming upstream for being unwilling to accept our obviously perfect ideas and/or code.

    That’s why you find Nokia in the top kernel contributors and widely respected for punching well above its weight. Contrast with Android, which doesn’t bother engaging, occasionally throws ‘perfectly designed’ code over the wall, and then throws its toys out the pram when they don’t get accepted. Not that rejection is hugely surprising when you haven’t contributed or engaged with the community: how can you possibly know what they want?

    As for copyright assignment, you’ll note that these are three non-community projects. It’s borderline pointless even trying to contribute to Firefox if you’re not employed by MoCo, and likewise working with Qt and MySQL is tough. In fact, the latter was so tough it spawned an increasingly-popular fork that I’m sure you’re heard of: Drizzle. Qt also has no real community to speak of. So if your goal for Unity is to end up as a Canonical project with zero community involvement or influence (beyond the odd few patches whose main benefit to the project is to back up the illusion of a community), then I wish you the very best of luck with that, and the very best of luck in recouping the associated payroll costs with proprietary licensing, but surely you can see why GNOME didn’t accept it.

    X is indeed licensed under a license that allows proprietary derivatives: a sad fact of history. The Xhp/Xsgi/Xsun/… forks were hugely damaging, in that a massive amount of incredibly valuable code was lost forever, not to mention that most of the reason for the brain damage all through the X codebase is to support insane proprietary forks we reaped no benefit from.

    Cheerfully, krh hasn’t made the same mistake with Wayland, nor has he made the same mistake as Unity/GEIS/GRAIL/Qt/MySQL/Mozilla.

  29. Alexander Rojas Says:

    Well, that just strengthen the position of Linus Torvalds towards Gnome. Well, there will always be KDE

  30. Jeff Waugh Says:


    I worked for Mark. I love Mark. But I also understand his faults, and have high expectations for Canonical’s collaboration with the GNOME (and broader Open Source) community. I want Canonical to succeed at doing so, because I want Ubuntu and FLOSS to succeed. I am angry because Mark has screwed up.

  31. Luke Says:

    I found this useful to read for background, but I still need some help understanding.

    So although Gnome 2 applications will work without modification with Gnome shell or Unity, enhancements are required for both to take advantage of the new environment’s features?

    Does this furore basically comes down to Gnome people fearing apps will make Unity integration enhancements but not gnome shell integration enhancements?

    Did the Gnome people talk about their shell API enhancements with, for example, KDE API people? So KDE could take advantage of the new enhancements? So in theory, this is the role of

    I’m trying to imagine a world where desktop environment choice is no more complex than changing a theme skin.

    Taken to extreme, I’m trying to imagine a world where Banshee, Amarok, Rhythmbox and even VLC are nothing more than skins for the same underlying app. It seems to me that it’s far better for people to waste time on cosmetics than waste time on fundamental workings.


    Guys, calm down. This is just a conflict and everyone is explaining their positions.

    @Jeff: I have the impression, that you think that Mark hasn’t correctly reproduced the way things happened. Is that correct?
    @Mark: Can you imagine what might be the cause for Jeff’s thinking that you did not reproduced things like they happened?

    After stating your positions, you should go hunting for the interests that build these positions. That being said, I’m a mediator. It could actually be helpful to mediate here, as I see it.

  33. eet Says:

    @Luke: GNOME development is a very open process:

    Anyone can participate in it. Even Mark.

    But Mark has decided to go the way alone.

  34. eet Says:

    @Luke: I misunderstood you. You can perfectly well use KDE-Applications in the GNOME environment and vice versa NOW.

  35. Conzar Says:

    Why do you (Mark) state that competition is “Healthy”? What health are you referring to: Physiological health, physical health, or both? Who’s health are you referring to, those of the competitor where there are winners and losers, the end user, or society?

    What are the disadvantages to competition?

    Is the GPL and other Free Software licences designed to be competitive or collaborative? Is Free Software competitive or collaborative? Is proprietary software competitive or collaborative? Is Canonical’s ultimate goal to be competitive or is it to provide a collaborative environment for the world’s people to share?

  36. luca Says:

    Althought I don’t think about software made in Canonical and not accepted upstream, also as external dependencies,
    the problems inside the Gnome Board remain the same:
    fossilization, willingness to keep the status quo and no signs of collaboration.

  37. eet Says:

    @luca: If you think that, why don’t you join up and change things? Because you can, you know. Work on the project, run for the board. Or, put the other way round; if the board is bad, why did you vote for it? Because you can vote, you know?

  38. Seb24 Says:

    @eet : Because you can vote, you know?
    I know, but i don’t understand how… Other big problem for gnome project : it’s missing a good documentation and better communication. it’s really hard to find information and understand how you are working.

  39. luca Says:

    ” If you think that, why don’t you join up and change things? Because you can, you know. Work on the project, run for the board. Or, put the other way round; if the board is bad, why did you vote for it? Because you can vote, you know?”

    I already contribute to foss where I can, spending a lot of efforts.
    I feel I can express my disappointment even if I’m not directly involved in Gnome project.


    @Conzar: Competition is a tool to get information you do not know. You cannot know e.g. what is best for the users, unless you let them chose what they like best. In that for competition is „healthy“, because the most „fitting“ (or used as a common synonym: healthiest) solution for the users is chosen by all users individually.

  41. Dean Says:

    > That’s it – right there. Remember, this was a proposal for the indicator
    > API’s to be an *external* dependency for Gnome. That means, Gnome apps
    > can use those API’s *optionally* when they are being run on a platform
    > where they are useful.

    Mark, I think you’ve misunderstood about external and optional gnome dependencies.

    External dependencies are dependencies that are required by a certain application and require sign-off by the Gnome release team.

    Optional dependencies are dependencies that are optional and do not require sign-off by the Gnome release team.

  42. Jackie McGhee Says:

    There is some good stuff in here , Mark. I’ve been an Linux user since the first public release of Ubuntu and the events are not unconnected. I wish GNOME had got on board with libappindicator, but I have read somewhere that to do so would require assigning copyrights to Ubuntu.

    Is this true?

    (I’ve yet to see a good reason why copyright assignment is a good thing; why a developer should then give up their rights to their own code, effectively losing the time *and* fruits of their labour. As a designer, if a client wants me to transfer copyrights, it bumps the price which is as it should be.)

    If this isn’t true, then it makes GNOME’s position all the more baffling, but if it is true then all it does is re-inforce the (by now, somewhat entrenched view) that this is all a play for power and influence on Canonical’s part.

    In the end, it leaves me (a user) unsure what to do. I won’t be transitioning to Unity (to me, both it and GNOME-Shell are *horrible* on screens with large resolutions), GNOME is moving away from how I want to use my machine and as much as I love the ideas behind KDE, I don’t like how everything feels cramped inside KDE windows (I like my padding, what can I say?).

    And although it isn’t earth-shatteringly significant, I’ve helped a couple of handfuls of people transition from Windows and Mac OS X to Linux via Ubuntu – if I don’t know which way the wind is blowing, how am I supposed to advise others in all this mess?

  43. Barbie Says:

    Speaking of ‘health’………
    Shuttleworth, Seigo: GNOME’s Not Collaborating
    OS News – Thom Holwerda
    “We’ll have to see what this all leads to, but I’m hoping this ruffles a lot of feather and causes a great stink, since a re-evaluation of what kind of cooperation is important could be just what the doctor ordered.”

    All those in favour of mediation…say ‘‎I’!

  44. mark Says:

    Adopting the API’s does not require copyright assignment to Canonical. Apps can use the API’s with no assignment or contribution agreement. When Canonical (and Mozilla and Qt and MySQL and the FSF) are the anchors in complete codebase, we prefer to maintain copyright integrity, so we ask for contribution agreements to be signed. Those agreements only cover patches to the actual core code, not apps which adopt it.

  45. [mp] Canonical vs GNOME « Idl3's Blog Says:

    […] rea di mettere i bastoni tra le ruote a Unity, ultima creatura di casa Ubuntu. Nel post “Internal competition is healthy, but depends on strong and mature leadership” Mark dice: “When we articulated our vision for Unity, we were very clear that we […]

  46. Fanen Says:

    I see Ubuntu trying to become the Linux distro that every one programs against, sorry everyone else.

    Seems like a natural move, to take advantage of your critical mass. To me, it removes the reasons why I liked linux in the first place. The choice. I am now getting forced to choose between Ubuntu and Gnome… not the way I imagined it.

    Time to start evaluating the options.

  47. Dean Says:

    Fanen, you can still use the ‘classic’ gnome experience. In 11.10 you should be able to use gnome-shell (you can use it in 11.04 via a PPA.

  48. robert renling berencreuz Says:

    There’s a point both Neary and Shuttleworth are making.
    There is a big barrier of entry in regard of contributing to the GNOME project, avid apologetics and developers related to GNOME will probably tell me to put up or shut up.

    There are more than a few people who got tired of beating their head against the wall while trying to contribute to something they viewed as a fun and interesting project, andrew johnsons smooth comes to mind.

    But the bickering needs to stop, one of the cornerstones of collaboration is being able to work together despite personal differences.
    your professional ethos is not the equivelant of your personal ego.

    I do understand disagreeing with a project like Unity because believing they work in secret, but that is also the commercial reality of an open source projects possibilities to make inroads vs. its competitors.

    gnome-shell isnt here today, it isnt even usable outside of building it with jhbuild.
    Unity however is, and wherever we were with libappindicators, its here now, and it works swimmingly across the board thanks to individuals outside and inside canonical (kde, individual contributors)

    While Unity emerged as a skunkworks project, it isn’t today, worked upon in secret.

    There’s also the matter as to how individual gentoo developers — lets not forget the users — were met and handled by the GNOME project in 2003-4.

    @jeff, i remember the gnome-web issues, was that born out of mischief and malice by yourself or was that just a misunderstanding born out of bad communication as well?

    @Conzar, @homresiniestro: patterns of form, a design is meant to be good when it fits the world, not when it reaches a set of bullets in a list or meet a current trend, if the latter were true, we’d have iOS everywhere.

  49. Kang Says:

    I also feel that effective cooperation between multiple projects is critical for the success of FOSS community. There are plenty of enthusiasm and good number of skilled programmers. But what really lacks is the organization and cooperation between the projects working with maximum efficiency to deliver a uniform experience to the end users. Thanks Mark, I believe you are the leader that has vision and insight to achieve this objective.

  50. Jackie McGhee Says:

    Yes I see I failed to make that distinction at the time, good catch. Thinking on it further, then, how do you expect Canonical’s implementation to attract the best of developers when they have to give up so much? What I mean is, if GNOME took on the spec and various GNOME devs wanted to improve it there is an immediate and artificial wall to face.

    The point raised earlier about Mozilla, QT, MySQL not being community efforts still stands and let’s not forget that Emacs is a famous “Cathedral” and that the FSF seem even more remote.

    Now, I know Ubuntu isn’t a community-led distribution (I read the recent IRC log where you spoke about that), I know someone has to pay the bills and keep the lights on at Canonical but that only makes this debacle look even more stupid: one can’t credibly appeal to the community for understanding and then ignore it when it suits.

  51. Jose Says:

    Interesting… Mark

    When I see Apple I wonder how oppressed people will feel working for them, like “The Simpsons”, there is not credits in the programs documentation, there is no Mister X or Miss Y, not coding stars, only Apple. The Apple designers with Jobs are the gods that take all decisions.

    But this system works, it is the Cathedral model with their main architect. Today we say that Cathedral X was “built” by Gaudí, or that skyscrapper was “built” by Frank Owen Gehry. They must be very strong to built it all by their selves alone!

    It seems clear to me that the bazaar model in the open source world could be improved a lot empowering the individual, like a real bazaar market does. We could copy what we know that works from thousands of years of experience in the real life.

    This is affecting you personally, emotionally. This rant is not going to solve the problem. The Gnome guys are going to respond the same way because their ego is there too.

    This is like some guy fighting against another guy and both telling they have won.

    Acknowledging there is a problem you can solve it, my suggestions:

    Create a competition, like sports do.

    Make both parties accept the outcome.

    Create a neutral figure like a soccer referee that both opposing forces respect.

    Create a neutral testing population that both parties agree with.

    Make the process transparent so a lot of people could see it(The superbowl of Linux :-D).

    Let the neutral population decide what works and what does not.

    Accept the outcome and let everybody work in the same direction(and sense).

    BTW, This is the way Apple works too. They create multiple COMPLETE pixel perfect implementations and them only decide using one. The designers are the “referees” on Apple, people trained to judge.

  52. Ioannis Vranos Says:

    Isn’t it obvious that Ubuntu competing (against) the rest of GNOME and other communities is going to fail? Canonical will have to maintain and bugfix Unity and some subcomponents it depends on, like Compiz, which is going to be abandoned, by itself.

    One company can’t compete the entire open source world. Imitating Microsoft in open source world is not a good thing.

  53. Dave Neary Says:

    mark: A small factual correction. Firefox does not require copyright assignment. Mozilla has a committer agreement, where committers agree to ensure that code they commit may be committed to the project (is original work, and has the consent of the copyright holder). The people involved in the long Mozilla relicensing drive a few years ago might have liked it to be otherwise, but it is not.


  54. lelamal Says:

    I wonder why is it that lately you (Ubuntu, Canonical, etc.) are those fighting with anyone else. Flame-wars always have you in, one way or another. You’re at the center of that tribalism you so much despise.

    You know, when I was a little child, everyone used to quarrel with me. And when I tried to justify my reasons with my parents, I got always scolded. I was left with a feeling that I had lost anyway, no matter the outcome of the quarrel at hand. But you know what? It wasn’t until I grew up that I understood why: it wasn’t *everyone to quarrel with me*, but it was *me to quarrel with everyone*. Once I swapped the terms in the right position I finally understood *I* was the problem.

    Think about it: challenge your perspective, you might have a novel vision of events. Until you do, your words will remain only polished yadda-yadda, and your self-satisfied and self-defined mature leadership will only prove that you haven’t still grown up and still need to get scolded.

    Good luck with that!

  55. Sergof Says:

    As a user and application developer all I can say is that I find the whole “competition is awesome” attitude annoying and useless.

    It reminds me of a similar situation in the videogame industry back in the day when they weren’t as popular. Often when asked about their competition game developers mentioned that they don’t really see other games as competiton, the real competition are movies, when enough people sart playing games instaed of watching movies then one can start thinking about competing.

    What I’m seeing here is that instead of facing the real competition FOSS people just divide the few percent of marketshare they have. It’s really a shame since it’s one of the few places where collaboration is possible and still people continue to pull in different directions by either being stubborn or by trying to force their ideas onto others and instead of discussing the problem they argue about who threw the first stone.

  56. Luca Cappelletti Says:

    I’m very happy to hear you this words that I agree a lot. is not the future but the present and the future at the same time.It’s here and should be used to build all the networks of GNU objects that GNOME failed to build for all due to its own social architecture (it’s an historical problem).
    FD.o is the place where all the main common specification works a lot across GNOME KDE XFCE LXDE and much more…this mean that it works if you can play together.
    The 99% of the desktop services are really common services that you can find everywhere in Windows, OSX, GNOME, BeOS, KDE and so on… so this is the ultimate right place to develop and propose…
    I believe a lot in FD.o this is why times ago I provide there the initial SpatialBundle skeleton while I’m developing details like the meta-compiler/builder and interaction specifications (this far before the portableapps marketers) here you can check progresses:
    At the moment GNOME is the only real desktop that provide production stable releases of a really complete desktop environment, this is why Ubuntu use it as default, KDE is out there with tons of very good technologies but the “Windows” culture inside does not produce good community for the rest of us…so we are here to wait them to revolution all…maybe one day…I hope…
    What you have to concentrate it’s to not fork and be really very patient and be constructive against GNOME community, and due to your central role in terms of desktop diffusion, to try to glue communities together…like FD.o
    OSX has a very big gap with us and it’s going to accelerate because they take what we did and package better to spread to the world thinks that we use everyday like hackers but not as generic users…I cannot understand when Time Machine was sold to the world…for me it was a simple rsync but was impossible to let understand this to my girlfriend…today again with Versions, Full-screen Apps (really they are selling as a product what I can do from early stage of XFree86?? unbelievable) and what about Resume?? no comment really!!! you know what I mean…
    We are wasting a lot of times in terms of “divide et impera” social communities (this is good for Darwinian approach to solve things but we need to really deeply focus on FD.o), we need to find the mean way between Darwinian competition and focus narrow ways to run together.

  57. Steve Says:

    Great post, Mark!

    Is there any plan to switch to KDE and use Compiz (and Unity of course) by default instead of Kwin and Plasma Desktop?

  58. Dylan McCall Says:

    I’d like to point out, to be fair, that Gnome Shell simply does not NEED an indicator-like thing. It is all about the FD.O notification specification and persistent notifications. Where we have system indicators, they have menus built into the shell. (Which should actually improve the environment for indicators, come to think of it, since it demands that information be available off DBus).

    I know that doesn’t really touch your concerns, but I think going on to suggest that Gnome is inventing parallel stuff for the sake of it is incorrect here. (Not that it hasn’t happened; we have often reaped the benefits of improved implementation from that process).

    At any rate, Mark, I’m starting to dread seeing your blog posts on the Planet because I know they’ll ignite this argument again :(
    Personally, I think the best way to solve this is to let it sit and cool down for a while. Our first releases can do the speaking, and nothing can change in between that. (Frankly, I expect every major distro release in the next six months to be deemed a skipper, but the extent to which they are should be interesting).

  59. Melvin Says:

    I remain neutral on this subject. I see it as religion. There are many religions, all or most religions see god differently, or simply follow another god. All religions have different views of a similar subject. Like Gnome and Ayatana, they see the desktop from different points of view.

    I won’t criticize Gnome or Canonical, I like Shell and like Unity, none are perfect, both have issues and both have advantages over one and the other. The good thing is that you get to decide which one you use at the end. I’m sticking with Unity because it works better for me. I will probably install Unity on my family’s computer because they are used to how it works, but first I’m going to show them Shell and let them decide which one works for every individual.

    Canonical and Gnome have good intentions, they want to improve the user experience but both have a different vision. There’s no need for sensationalism or controversy, that’s not healthy for the free desktop.

    I believe competition is healthy because it drives innovation. Microsoft started to make Windows better when Mac really started to take off. That’s why Windows 7 a such a solid release and why they are working so fast on Windows 8. Back in the XP days, Microsoft had no real competition, that’s why Vista was released so many years after XP and was such a bad release. They didn’t have the need to innovate and drive the desktop forward. But now they have good competition with Mac and both platforms are pushing innovation.

    I hope in the end that this leads to great growth for Linux. Linux has had the stigma of being low quality for years, it’s good to see Gnome, KDE and Canonical trying to solve that problem with their takes on the desktop.

  60. Anyway, he tried a few switches at random and Says:

    “Oh, won’t somebody please think of the poor users!”
    I can’t tell you how livid I get whenever I hear this line of rhetoric used to rationalize beliefs contrary to mine. My area of study in college was completely different than computer science and I wear a uniform to work. Do you? Then I’d appreciate it if you not speak on my behalf as if I were a small child incapable of my own opinions. Competition is not what is best for my interests. My first introduction to GNU/Linux was from an avid fan of LUG radio who implied to me that the distro that is currently winning the Linux popularity contest stood for the freedom to choose how I access information/knowledge regardless of what corporations would like to feed me, the freedom to participate unbound, and accessibility to all citizens of the world without barriers. I ran Ubuntu for two and a half years, advocated, and supplied my local pub ‘community table’ with cds before I even knew that the four software freedoms existed. A user of a distro that wasn’t winning the Linux popularity contest informed me of them. I read free as in freedom and the cathedral and the bazaar. I learned what freedom really means. What I like best is gNewSense. Feeling like an multi-level marketing scheme burnout?, not so much. I like Ubuntus focus on ease of use, and I do believe Ubuntu can still become the balanced distro it set out to be, but replacing a social philosophy with capitalist dogma is not the way to do it. Y’know what I would also like? To watch any flash video I want or for my mouse to not randomly freak out from time to time. I can’t have those things at this time though. There are good people working on that, but people like them (and their skills) are also a resource that’s being competed for :(
    @Conzar: Judging from his recent behaviours/change in outlook, I think he’s referring to it being healthy for Canonical to be Duh!…Winning! Also, before anyone demands “Then why don’t you change things?”; Quietly, slowly, I am trying. Contrary to the beliefs held here, vulgar tourists, peasants, and members of the peanut gallery are not entirely ineffectual.

  61. bognarandras Says:


    Is there any plan to switch to KDE and use Compiz (and Unity of course) by default instead of Kwin and Plasma Desktop?

  62. varinski Says:

    And now a word from a self-appointed representative from the Noob Community…
    Dear Mark,
    we have this saying in French that says:
    “Les chiens aboient, la caravane passe.” Let the world say what it will.
    No matter what you do, there will always be someone to run you down. Can’t please everybody! Trailblazing is an extreme sport. Not for sissies.
    No doubt you have a thick skin. I just hope that the people around you are as well equipped as you are to put up with the venemous campaign that is currently taking place against Unity…
    Anyways, keep up the good work. Follow your instincts. Stick to your vision.
    Cheers from Montreal.

  63.   I hate to say I told you so… by Nathaniel McCallum Says:

    […] really want to believe Mark’s post as I have nothing but goodwill for Ubuntu (if Ubuntu succeeds, we all succeed) and Mark.  However, […]

  64. istok Says:

    what on earth do you know about strong and mature leadership. this is insulting.

  65. Has GNOME Rejected Canonical Help? Shuttleworth Responds | JetLib News Says:

    […] up the ‘they didn’t want it’ claim by Canonical and Ubuntu people. Today, though, Shuttleworth responds on his blog. ‘Competition is tough on the contestants, but it gets great results for everyone […]

  66. ethana2 Says:

    Religion is when you say “this is the Right Way to do this because I think so, let’s do it.”
    Science is when you say “Let’s do studies on how end-users interact with our software and document everything we can about their experiences and use them to improve our software.”

    Canonical is the only company I’ve observed to be practicing the *science* of UX, so saying that Ubuntu is a religion is something I find inaccurate.

  67. unwesen Says:

    @sect2k: When I read this post, I came across this bit:

    “This is because we wanted to be sure that whatever competitive dynamics arose were *internal* to Gnome, and thus contributing to a better result overall in Gnome in the long term.

    We’ve failed.”

    Did you not? What more of an admission that mistakes were made do you need?

  68. Robin Burchell Says:

    As per Dave’s comment on Mozilla, Qt hasn’t required copyright assignment since they opened their repository in 2009. They do require copyright *license* so that for instance once you contribute your code, they can still release under commercial license, but ownership of your work stays with you.

  69. RadekB Says:

    Two points here:
    1. Let the users vote. They are ultimate judges.
    2. GNOME and Canonical obviously don’t collaborate well (for whatever reason) with each other. The question is, does Canonical collaborate well with KDE? With Xfce? And does GNOME?

  70. txwikinger Says:

    I agree completely with the aim of this article. The strength of the open source desktops is the possibility to have some friendly competition that leads to innovation for everybody. For this to work and the walls between the desktops not to become insurmountable, good interoperability is very important.

    However, I have to also point out, that there are similar issues within the Canonical team. Ubuntu One is one of those examples. As far as I know (and I have given up about this since the last UDS) there is no proper UI for Ubuntu One running on Kubuntu. A lot of the problems have been as has become very clear in one session at UDS-N, the unwillingness of the Ubuntu One team to be open to a process that lets others to hook on and be able to use stable APIs in order to make things work.

    As a result of that session, I have personally given up on Ubuntu One, because with such attitudes I cannot see any positive outcome for Kubuntu and KDE users trying to use Ubuntu One efficiently. However, maybe your expression of frustration with the Gnome leadership can also bring understanding by yourself and the Ubuntu One team for the issues that Kubuntu and KDE users have with Ubuntu One and some positive changes can happen here too.

    In any case, thanks for picking up this topic and bringing it up for discussion. I think it is very important that the issues of collaboration are solved better than in the past.

  71. eet Says:

    @ethana2: Nahhhh, Novell did that way before Canonical and in way broader and more professional scope. The result was the SUSE Linux Enterprise Desktop. But they made pretty much the same mistakes as Canonical (why haven’t you learned from the mistakes of others?), so GNOME didn’t benefit from it.

  72. Temar Says:

    This link is also worth mentioning:

    It’s an attempt from Aaron to discuss the new specification with the GNOME devs.

  73. nmarques Says:

    “But perhaps if we had both Unity and KDE working well there, Gnome might take a different view.”

    The problem with GNOME is the corporate interests within it, only a blind man can’t see them…

  74. nixternal Says:

    Ahh, nothing like being called a muppet, thanks dude!

  75. Nick Wallin Says:

    What leaves a bad taste in my mouth is that it seems like Ubuntu said to Gnome “here guys, we have this awesome idea, you should use it… but if you don’t we’re still going to implement it without you.”

    Where is the collaboration in that? If you knew what you were going to do ahead of time, that isn’t collaboration. It is using the weight of Ubuntu’s popularity as a driver to force change in Gnome. At least that’s what it seems like to me.

    Ubuntu was my first Linux distro and I have been happily using it for 3 years. Yet I have never been so confused about what distro I will be using in 3 months. It’s always been “I can’t wait until the next Ubuntu.” Now I feel a need to choose between actual unity (i.e. distros working together to improve the user experience) and Ubuntu (one distro working unilaterally to do what they want to improve the user experience).

    Ubuntu is, in my opinion, the easiest Linux distro. Ubuntu has done a great deal in promoting Linux and open source. Even so, I think anyone with any sense of principle will have to take a serious look at the options available this Spring.

  76. mae Says:

    X-Monad forever! 😉 (and

  77. Bill Cox Says:

    I have also recently gotten fed up with how hard it is to get important patches into GTK+. I’ll just leave it at that, because I don’t believe this is just a problem with the GTK+ team. It’s part of a larger problem in open-source land where we have created so much red tape, that it’s very difficult for innovators to share their work. I believe the rapid expansion of Android shows just how serious this is. The key to Android is that any developer can get his app into the App Store with very little pain or delay. We’ve build the entire eco-system for innovation upside-down.

    The way it should work is that an innovator publishes his work (like he can in Android), gets a bunch of enthusiastic users, and then let’s his work filter upstream to whatever projects he branched his code from. Instead, we’ve allowed the oldest projects, like GTK+ in this case, to act as gate-keepers to innovation. Even when we succeed in getting past the gates, it takes years for average coders to get their work into user’s hands. We should reverse this trend. Instead of trickling innovations out from the root, we should be feeding them back from the branches.

    So, here’s the dumb idea to fix this. The hardest part is first admitting that there is a problem. This is where Unbuntu will most likely fail. While Ubuntu works to rhe seen in the Ubuntu Software Center. Clearly Mark wanted an answer to Apple and Android App store success, but what he got was a pretty, but less functional GUI for the exact same system we had before. The world’s response to it was a collective “Huh?”.

    So, in pie-in-the-sky land, Ubuntu, figures out that the real change is required to enable the kind of innovation we’d like to see, and are capable of. We have to fundamentally change how coders share their work with each other and with users. We need developers to be able to publish their work in Ubuntu with very little review, just like in Android. That means Ubuntu has to switch to a packaging system that runs apps in jails, just like Android, and where the exact binary version of each library used by the app is present for the app to use, and which never get upgraded unless either the app is upgraded, or there is a severe security issue which requires a forced library update. Such a system could be built and used in parallel with apt, and it could enable a massive increase in innovation.

    If GTK+ were already distributed through such a system, then the way things would work could be like this. Ubuntu clones GTK+ with a simple clone command, and adds support for Unity indicator APIs. Ubuntu then lobbies for GTK apps to support the API, and many GTK apps then switch to use Ubuntu’s branch rather than the root from the Gnome team. Any updates to Ubuntu’s version would be automatically e-mailed to Gnome, and at Gnome’s leisure, they would be free to incorporate any, all, or none of Ubuntu’s upgrades. It wouldn’t be Ubuntu’s responsibility to try to convince Gnome to accept Ubuntu’s work. Rather, it would be up to the wisdom of the Gnome team to adopt what works best, or face the possibility

  78. jospoortvliet Says:

    talking about strong competition, openSUSE 11.4 is out:


    Meanwhile, I’m quite a bit with you on this one… Not saying you guys did everything right but I’ve been harping on you for not contributing, then seeing contributions being rejected is quite sad. It’s true that there is no online proof that you tried to collaborate or did it properly, but seeing the threads on FD.o where Aaron Seigo tried to interact with the GNOME devs gives me the impression the fault might very well be on GNOME’s side :(

  79. ubesen Says:

    I think that the Canonical position would be much better if Canonical was funding a couple of core GNOME (let’s say GTK+) developers.

    If I were to guess the sentiment within GNOME, I would say that there are lots of manpower needed and little help from Canonical.
    It makes a big difference to see a couple of people with a address shooting patches for the core of GNOME.

  80. byeubuntu Says:

    “Fanen, you can still use the ‘classic’ gnome experience. In 11.10 you should be able to use gnome-shell (you can use it in 11.04 via a PPA.”

    Maybe you can’t see the pattern? Hint: recheck from 10.04 to 11.10 … in what repo you find the upstream gnome? A PPA, don’t make me laugh.

    It’s the same all over Ubuntu, admit it. If GNOME is ‘saying that Unity is not “in GNOME”’ to the same extent is Canonical (Mark) saying “Gnome-shell is not in Ubuntu”.

  81. Bill Cox Says:

    Grr… A good linux driver for HP touchpads would be really nice. As is, I get random clicks all the time, including the unintended click on Submit Comments on my above comment.

    Anyway, in summary, we should change how we release new code, because the current system puts upstream gatekeepers in the way of innovation. Any author should be able to effortlessly branch nearly any package in the system, and make their improved version available through Ubuntu with a very small review process. This would make it more like Android, where I can get my work out there right away. Unlike Android, Ubuntu could enable coders to share their work, as many packages are libraries rather than apps. To enable this, apps should be run securely in jails, using the exact libraries they were compile with. The source version system should deal with making patches available upstream, without any effort from the author. Let upstream pull changes, rather than trying to force them to change. Get innovation back where it belongs… with the new kids on the block, not maintainers of old projects.

  82. Ian Says:

    Seems like Jeff writes a lot but says nothing.

  83. terlmann Says:

    I’ve been a user of ubuntu and gnome since warty warthog, and I’m still not impressed or happy with the UI, or with ANY of the UI’s that I’ve dealt with so far.

    I think gnome-panel was rubbish, I think unity is rubbish, and I think gnome-shell is absolutely rubbish. It’s reinventing the wheel and destroying the advances developers like the compiz and cairo-dock teams have worked hard to deliver as an advanced, capable, Open GL accelerated UI. In unrelated news I dislike QT, because it contains plenty of bugs and has such a larger system resources demand that it makes gnome look like FVWM.

    Now, you think I’m trolling, but remember, I _Have_ been around since then, so take my bait with a little bit of salt and read it for what it is, one honest user’s opinion. I’ve used nearly every operating system on the planet, experimented with every single UI concept ever to hit the world, and I think windows 7 has the best operating system gui yet.

    Here are some UI’s that I’ve dealt with and that I’d like you to reflect on:
    Project Looking Glass
    Symphony OS
    Athene Os(Proprietary)
    Gnome 1-2, 3
    KDE 1-3, 4
    Windows DWM, 95-2000, XP, Vista/7
    Mac OS 9, 10.3,4,5,6
    Blackbox/many similar UI

    Ok so now that I’ve brought you through time and space to no particular point, let me reiterate:
    You are all rubbish that is full of bugs and you all need to deliver me, the consumer
    a consistent, smooth, customizable, reliable, crisp, accelerated/non accelerated UI.
    So far, the closest that anyone has ever gotten to that IMHO has been windows 7 and Linux Mint.

    Some notes I’d like you to take to heart:

    I totally despise people who put the window controls in the upper left corner of the screen,
    I use my mouse with my right hand.

    UI that takes up more than 45 pixels on more than one side of my display obfuscates my screen.
    I don’t like it and it doesn’t emulate real life, where all the buttons are BELOW the screen, not all around it on your monitor.

    UI that doesn’t allow me to make everything partially translucent WITH bumpmaps for glass effects,
    or even shaders to allow me to emboss/metallic my entire screen, are less than cool. They remind me of gnome 1. UI that gets in the way of letting me be super duper cool by implementing this is obfuscating as well.

    Now that I’ve left you some tips on UI design(I’m sure there are many more out there, none of which you follow), I’m going to leave some orders for Mr. Shuttleworth.

    Operating systems that emulate each other only emulate a superior offering. Mark, your current natty design is just emulating windows and OSX, and the worst part of OSX IMHO.For sure, Subliminate your Unity your unity right hand panel into suse’s slab menu, and put the entire thing back on the bottom of my screen. That’s where I want it and that is where it will work best. The top of the screen shouldn’t have anything up there but window controls and notifiers. When you pull your cell phone out of your pocket to check for text messages, you hold it up in a similar place. When you close your fridge doors, you tend to do it by grabbing them with the hand you prefer in an upper corner position. Personally , I am glad it just takes a simple tweak to put my window controls in the corner I like. I don’t like the “sensation” of how unity feels in acceleration terms, it feels like it completely takes over the screen and then renders other content through it. Same thing for gnome shell. This shouldn’t be. It makes the whole thing seem lagged.

    Finally, if any of you want to come work for me(I pay people in cheezits) I have an entire UI design mocked up that is, number one, revolutionary(it’s completely different in every way from current offerings), innovative, simple, usable, and instinctive. Please email me : terlmann at yahoo dot com for more details. I personally guarantee that the person who lets me lead their design team will become the next apple.

  84. Jef Spaleta Says:

    About internal competition,

    When was libappindicator formally introduced to the GNOME community? Was there a presentation on it at GUADEC in 2009 with the purpose of formally introducing the concept and getting feedback for the cross-section of GNOME dev community in attendance to lay the groundwork for an internal technology competition inside GNOME? Was there any public discourse at all on any publicly archived GNOME communication channels in 2008 or 2009 while libappindicator was under development to suggest at any point that Canonical wanted to see a vibrant technology competition inside the GNOME project development process.

    I simply cannot find evidence that Canonical made a reasonable good faith effort to drive a discussion about internal technology project competitions inside the boundaries of what is the GNOME project. The public record of that desire is not there.


  85. Alistair Withworth Says:

    @ Jef

    Planet Gnome is a public channel. Of Gnome.
    Jono Bacon’s blog is syndicated there. He mentioned AppIndicators there.
    That’s your public record. It’s right there.

    That’s how I heard about. And I immediately thought “wow, glad this is where Gnome is going!” People shared my enthusiasm in the comments.

    And you question Canonical’s good faith. You do realize that there’s two opinions on the current matter:
    1- that of people with an address
    2- that of EVERYONE ELSE.

    And people READ what you guys are writing. And it’s obvious that it’s trolling. Obvious. To everyone except people with an address, it seems. Most people are just trying to communicate this to you.


    Look them up. In a mirror.

  86. sect2k Says:

    @unwesen: Admitting failure is not the same as admitting a mistake, actually there is a big difference.

  87. Jef Spaleta Says:

    And as I have said in other places… can we compare and constrast what went right with CSD development and integration into gtk+ with what went wrong with libappindicators? CSD was Canonical staffer led and there’s little to no drama about the early engagement between dev from different companies which led to a successful integration of CSD. It stands in stark contrast to libappindictor’s development history. The truth as to what actually works well in terms of engagement is in the understanding of why CSD development went so smoothly.


  88. Allen Lowe Says:

    we, at the elementary team also know what it’s like to have our contributions to upstream rejected without a logical cause, and it is frustrating. However, Mark, you’ll be happy to know that in our own Desktop shell “Pantheon”, we are using appindicators on our new panel, along with notifyosd. and they are working out very well.
    Mark, keep up the good work. I really hope that you do not become overly discouraged because of the negativity of the peanut gallery. The fact of the matter is, you change the world for the better, and all I can do is thank you for that.

  89. Omnifarious Says:

    In the history of the Ubuntu project I’ve noticed that there is a distinct “not controlled by us” syndrome. Ubuntu has a strong bias towards things Canonical has control over. I do not like this, and I’ve stayed away from Ubuntu as a result.

    My own personal brush with this was when the Mercurial and bzr people got together for a summit. I had hopes that we could merge projects, but that all went out the window because Mark wanted control of bzr to use as the backend for Launchpad. Now we have three major DVCS systems instead of two.

    I do not like the way the Ubuntu project engages the rest of the community.

    That being said, I’m not all the happy with GNOME either. They do seem a bit hidebound and convinced of their own rightness to the exclusion of anybody else. I can believe Mark’s tale, even if I don’t really think he’s telling the whole story either.

  90. unwesen Says:

    @sect2k: You’re right, failures and mistakes are not the same.

    Failure is a consequence of one or more causes, some of which may be mistakes. Mistakes aside (whether mistakes in the initial plan, in the execution of the plan, etc.), the other possible causes can be summarized as “changes in circumstances”. A change of circumstances can missed or ignored, which could be termed a mistake. Or they could be adapted to, and this in turn offers great potential for making mistakes.

    The upshot is that if a failure occurred, you can be certain that mistakes were made. Admitting failure, then, is akin to admitting mistakes – even if, strictly speaking, they’re cause and effect.

    Do I get the nitpicking trophy back now?

  91. Nick Says:

    All of this is FAIL. Ubuntu is reinventing the wheel and dividing the Gnome community… Gnome is taking an arrogant approach with the Shell. Both Unity and Gnome Shell are doomed to be KDE 4.0 all over again. The real question is who will admit that first and move quickly to resolve the problems. The PR nightmare created in the meantime will leave a bad taste in everyone’s mouth and further discredit the Linux desktop.

    Good job guys.

  92. TheGh0st Says:

    Seems like this discussion never ends. People should accept that mistakes were made. It is right, things need to be said, even in the case of GNOME’s behaviour concerning external development. Canonical did mistakes too. To my mind this discussion is childish. Admit and stand to your faults and finally get over it, on all sides.

    I mean, it is obvious, that there are a lot of people putting stones in Canonical’s way. The reasons for this behaviour is mostly different. To my mind the success of Ubuntu, independant of this discussion, acknoledges that they are on the right way.

    My opinion as USER is, that I want a stable, innovative, beautiful and Out-of-Box working distro, without 10 extra repositories. Just follow these principles and you are on the right way. That’s what Canonical done (even if there were some bugs). Give the best and users will thank it. It might sound biased but Canonical has done a lot for the users, even if upstream is whining.

    I think there are a few people (all the same on different blogs) wasting their time bitching on every thing and step that is done by others. Just ignore them and listen instead to some great ideas of users, inside and outside the community.

    All parties should get over this (hopefully at Berlin) and get back to (collaborative) work, even if there are forces that don’t want to play in your team. :)

  93. Max Kanat-Alexander Says:

    This is the second time I’ve run into Jon McCann’s name as being a blocking force on positive change. I’ve also seen a few of his comments in the GNOME Bugzilla. Now, I don’t know enough about the GNOME community or the history of the GNOME Shell development process to make a judgment myself, but I have seen enough now to cause me to ask the question–is the problem that you’re running into mostly just Jon McCann, and not the GNOME community as a whole? I could be totally off-base, but it does seem strange that his name has come up a few times in situations like this, now.


  94. Vale Says:

    Every time you talk about Kubuntu you say always “some muppet calls Kubuntu the blue-headed-stepchild” or “the odd super-self-interested muppet who expects me to singlehandedly make his wet dreams of technology kfuturism come true” (

    Why you say that? Why you sponsor a product when you don’t have any interests on it, where the project manager (paid) has around contributors (not paid) which are much better than him, and where you call muppet who believe in this product or who develop without earn money this product? It’s just childish and ridiculous and it’s not good for Kubuntu and for who use it.

    Ah…for internal competition.

  95. The blame game | tante's blog Says:

    […] Mark Shuttleworth drops in and writes a really weirdly worded blog post basically claiming that Canonical oh so wants […]

  96. Jef Spaleta Says:

    Max Kanat-Alexander,

    You know what, everyone needs to be careful about how private conversations are construed several years after the fact. I can not find _any_ publicly archived record anywhere in the publicly archived GNOME channels of communication that would corroborate what Mark said was said in a private conversation that he personally was not a party to. No followup discussion in the context of a mailinglist. No bullet pointy blog entry just after the UX hackfest where anyone mentions a summarized version of that private conversation. No pingback on a mailinglist at all to give anyone else inside the working GNOME developer community aheads up about what was discussed in private.

    There is no public record. That is a huge problem. HUGE. Would Canonical engage with OEM partners on a handshake and a smile with no public summarized list of deliverables and deadlines? C’mon. GNOME is an open by rule meritocracy. At the very very least the fact that noone inside Canonical made any effort to come back into conversation with GNOME devs in the last part of 2008 and all of 2009 and touchbase should speak volumes about a deep misunderstanding of how distributed high collaborative projects like GNOME and like the kernel work. You have to engage in the public forums. Public record discourse matters… it matters a very great deal…because the public record holds us all accountable.


  97. Jef Spaleta Says:


    The sentence: “Would Canonical engage with OEM partners on a handshake and a smile with no public record”

    Should have read: “Would Canonical engage with OEM partners on a handshake and a smile with no effort to confirm agreed on deliverables in writing”

    It is of course absurd to write Canonical OEM dealings would be expected to be in public. But when interacting with GNOME which has an open development model I fully expect there be some evidence that Canonical came back to GNOME devs and confirmed the major points as to agreement within a few months of the conference conversation. Certaintly at the very latest a presentation at GUADEC 2009 on the technology to get feedback from GNOME devs as stakeholders in good faith.


  98. Fred Warren Says:

    I would be perfectly fine with Ubuntu calling Unity a new Desktop manager with some sort of plugin arctecture to that can work with a variety of backends. For example, the one current back-end that Unity works with is Gnome. Then if the Kubuntu folks want to work on better ingegrating KDE elements to work with Unity they could do that as well. Keep as much stuff as is usefull as specifications at Freedesktop.

    It would be easier to work outside of GNOME than work on the inside. Build it so the underlying libs are well documented and the api’s are easy to acess, keep it a freedsktop standard and that way it will be easier when GNOME decides they want to start integraing these things. Or if they keep on doing it their own way, it will be eaiser for someone to write software that that interfaces between the two.

    I think any work done on d-bus would be a benefit to the community as well. dbus should be able to handle modifying key mappings, screensavers and many other events. That way other apps that have an interest in watching or modify things will have a much easier (and well documented) time at it.

  99. João Pinto Says:

    Where can we find such a mature and strong leadership ?

    Most of the leadership parties involved in this GNOME vs KDE vs Canonical subject shamefully played a blame game at .

    When there is a genuine collaboration goal there are proper ways to communicate with people, preaching in blog posts is not one of them. If you have a genuine intention for collaboration then assume that there is a FOSS Desktop disruptive situation -which you partially responsible for. Stop arguing on the past, setup some to face to face meeting with the key people involved. Take the results from such a meeting to the public. If an agreement/cooperation plan is not possible at least it should be clear who/why/how is not available for collaboration.

    There are many real collaboration attempts observed between nations and/or organizations, they happen between walls and the outcome is clearly communicated to the public -at the end-. How many hours have the decision makers spent personally communicating in the attempt to setup an a collaboration plan ?

    Public exposure of personal interpretation, based on disputed facts, is not a path to collaboration.

  100. John Says:

    Ubuntu is supposed to be Linux for human beings, yet Mint is doing a better job.

    Unity? Not ready for the masses. Not modern looking, cool or very well thought out.

    Step back, re-evaluate, repackage it.

    The wrong decision at this juncture decides the fate of the company.

  101. andrewsomething Says:


    I know that this doesn’t directly address your concerns as it’s not from a GNOME dev list, but looking a Ted Gould’s blog history is interesting.

    11 Dec 2008 – Mockups of the initial idea for the messaging menu [1]
    05 Jan 2009 – Some early musings on indicators: “What I’d like to put forward is the idea of little flags that applications can hold up to say what they’re thinking or doing, which I’m going to call indicators.” Specifically mentions interacting with John Mccann at UX hackfest in Oct of 2008(though it sounds more like discussing general ideas rather than specific work). [2]
    07 Jul 2009 – Ted posts his slides from his talk about the messaging menu at Gran Canaria Desktop Summit: GUADEC + Akademy 2009! [3]
    16 Dec 2009 – First look at a functioning messaging menu. [4]

    I think that you need to revisit your timeline of events a bit. If nothing else, you should note that “certainty at the very latest a presentation at GUADEC 2009 on the technology to get feedback from GNOME devs as stakeholders in good faith” was given. [5]


  102. Jasper Nuyens Says:

    The real issue here – as far is I see it – is simply the following:
    Mark has another ‘vision’ of ‘The Desktop’ as the Gnome leadership.

    And of course, Free Software means that everybody has the right to fork a project. So Mark could decide for that to pursue his vision. Yet being able to get a vision included upstream means that other people will take care of your code and have to watch for future compatibility and/or code reuse.
    So of course Mark prefers his vision being included upstream as it is with KDE. It isn’t that hard on the other hand, for the Gnome people to let Mark have his visionary additions, even when they contradict a little bit the vision of their own leaders.

    I have the feeling that Mark wants to make clear with this post that his vision isn’t that fundamentally different from GNOMEs direction as it is with GNOME 3. Yet that however is the feeling of some of the GNOME developers. With this post, I guess Mark wants to change that.
    Having a few extra ./configure –extra-feature options, can’t be that bad, does it?

    Ultimately, upstream is upstream. That’s how it works, and if we don’t like our benevolent dictator upstream overlords, we’re always free to fork.
    Yet forking Gnome probably outgrows a little bit the ambitions and resources of Canonical, at least it’s more than what we expect from a ‘distribution’.

    Forking is certainly an option if there is a critical mass in developers who are tired with current leadership. And as such it can renew interest into a project. While forking can lead to fragmentation, it can also lead to re-merging later, as it did with the Samba project.

    Ubuntu has grown over the years to become – arguably – the most important Linux distribution (as we don’t care for overly outdated kernels on RHEL server installations or old packages in ‘debian stable’ as long as we don’t have to look at it, for our desktop, we prefer something shiny and new that ‘Just Works’).

    Marks visions have been a tremendous benefit for the Linux community at large, yet sometimes a bit contested. I am a Linux user since Linux kernel 0.98, I believe. And for Mark I hope some GNOME people will turn around about this, but psychological speaking, that’s rarely the case when discussions are elevated up to this level.

    Other then forking, is the only other option for Mark to go for a Kubuntu with Unity and a Gnome 3 without Unity?
    If that’s the way Gnome sees it. Why not? A good ‘distributor’ makes it easy for the users to choose and make up their mind about which project is best.

    Possibly meaning that the default choice of Ubuntu would become Kubuntu?

    I guess that we have not seen the end of this story yet. And as always it’s an exciting episode of great minds working together ;-D

  103. Jim Philips Says:

    Nice arguments and all. But after seeing the Natty alphas, I decided that Unity only gets in the way of the new Gnome experience. And, after six years of Ubuntu, I have abandoned you for Fedora. I don’t know what Ubuntu is becoming, but it isn’t the Linux I like. The “improvements” you keep adding–including Ubuntu One–are things I have no interest in using. I’ll side with the Gnome guys on this one. You should have collaborated with them on their terms.

  104. pt Says:

    I am glad Ubuntu is moving away from Gnome. They are so arrogant they ignore the regular user and run a closed shop. Just look at their Gnome-screensaver application, absolutely useless, no config button. Most complained feature and they refuse to address it.
    You can compare Launchpad and Bugzilla, you will see which one has the regular user in mind.
    Nautilus Elementary came into being because many of their patches were rejected.

  105. ebonz Says:

    I’m so saddened by this. But hey look at the bright side, aren’t we getting a real good shell in Unity and Gnome Shell(though I am not a fan of it)? Had Gnome collaborated with Canonical, we won’t have two shells competing with each other now. I don’t know maybe I missed the real issue here, but I see things right now as a positive outcome :-)

  106. Jesse Says:

    It’s hard to sympathize with you, Mark, when suddenly you’re lecturing the GNOME community on collaboration. Yes, I agree: you have failed. Repeatedly.

    First, you say that code talks and tell how profoundly disappointed you were when libappindicator was rejected. But, then we learn that, well, it wasn’t really a gift. GNOME developers would not be able to work on the library without copyright assignment to Canonical. And, in fact, there was no apparent time during the development of libappindicator when GNOME developers could have given their design or technical feedback — because, NO EFFORT was made on the part of Canonical to solicit such feedback. So, we have a code blob, that GNOME developers weren’t consulted on during development (I’m waiting to see a response from you to Jef Spaleta’s points on the public record), and after it’s done, cannot meaningfully contribute to without copyright assignment. You really don’t see the problem with that? Having observed kernel development for some time, I think it’s safe to say that if you took such an approach with them (a hypothetical — I know Canonical doesn’t do kernel development), you would get a not very polite FUCK OFF response rather than a quiet rejection.

    It’s fine to claim that you want healthy internal competition with GNOME and that you want a collaborative environment. The historical record doesn’t seem to bear that out, however, and it’s only supported further with this canard of a post.

  107. John Says:

    Your competing with those on your own side, when you should be competing with the likes of OS X / Windows, doing what you can to rally folks to your cause and get them involved.

    It’s behavior like this that is why Linux isn’t a serious desktop consideration.

    Donald Trump says, “If your going to bother to think, think big”

  108. Claudio Says:

    I understand your point of view, Mark. Doing good and then feel attacked and misunderstood in your intentions and words. But what is most important is that you have created a competitive environment to Gnome and this has accelerated the development of Linux desktop. The best environment in which to grow is something like competition. Without competition there is no development : I give you a product and if you do not like you have to be satisfied, because there is only one. Gnome has always been a highly customizable graphical environment but at the same time a bit too vague. You’re trying to make your alternative more defined and the time will give you right. Trampling feet at someone (in this case Gnome), is to provoke reactions often unjustified. Gnome-shell is a project that goes on for many years but was never completed. Why then this project is now taking its first release in conjunction with Unity? Thanks to you and the work done by Canonical! This seems absolutely clear and evident! So, once again, continue on your way without closing any doors.

  109. Ankit Tulsyan Says:

    “Power is the capacity to translate good intention into reality and then sustain it.”: Warren Bennis (pioneer of the contemporary field of Leadership studies)

  110. biju Says:

    You’ve spent months arguing with idiots who can’t design a decent file-open/save. GNOME was designed to filled – since CDE was proprietary, RHEL needed a similar desktop.

    Unity blows. This is no way to build a product.

  111. Manjeet Dahiya Says:

    Nice read.

    I have just launched my update manager to test unity-2d.

  112. Inge wallin Says:


    It’s obvious that you have a clear vision for the user experience that you want to provide to your users. Whether that vision is good or not, is not something that I will address here.

    But the question is what the best way is to implement that vision. I have long thought that you would have had a much easier time to implement your vision if you had used KDE technology instead of Gnome. I think the time has finally come to develop that idea.

    I think that the big difference between Gnome and KDE is that Gnome is a product while KDE is a community. KDE used to be a desktop, but since there are so many different user interfaces created by KDE (plasma desktop, plasma netbook, and more are in the works) and also a number of application, KDE was redefined — rebranded if you like — into being the community instead of the software.

    KDE produces a set of technologies: libraries, workspaces and applications. We have a lot of internal competition and are open to new ways of using our technology. Our pride is not tied to one user experience since we already have several. I’m almost positive that your vision would have been received positively within the KDE community.

    I will assert that if you used Qt and KDE software you would be able to implement your vision faster while in the process gain the following advantages:
    – Faster time to market due to better technology (esp. Qt)
    – More cooperative community, esp. more open to new ideas.
    – More flexibility to test other new ideas that you have.
    – An interest to work on cross desktop standards.

  113. Chauncellor Says:

    These arguments are getting out of hand. I’ve never seen so much infighting before. It’s really saddening to see GNU stumble all over itself. This isn’t the first time and this won’t be the last but it still doesn’t make me less depressed.

    Even though Gnome devs have a tendency to be terse and unreceptive (in my experience), they’re good people. They’re obviously not in it for wealth or power, they’re trying to make a great desktop.

    Mark could have easily retired for his entire life and spent his fortune on lounging around and hiring cheap hookers. Instead he’s dedicated much of his life to trying to help the entire ecosystem. He may leverage his position into important decision-making but he’s doing it because he feels that there needs to be a strong leadership to get important things done.

    There are things that seriously need looking at internally from both sides. Gnome needs to break down this exclusive treehouse club mindset and, like Mark said, allow competition to influence their designs to make their DE better, too. Canonical needs their communication to be less secretive and cryptic (remember that mysterious data-collector they had during testing a few releases back? They never fully told us what it collected!). It’s hard to keep a transparent communicative relationship with a huge community but it’s so absolutely essential to staying focused. Just look at the United States government; we’re shooting ourselves in the foot so hard right now (9 trillion more dollars in debt this week).

    What’s the point of GNU? We taste delicious food, discover the recipe, adapt it and improve it, and share it. Right now we’re trying to see whose is bigger. Who cares if Unity or Gnome shell is the default UI in Ubuntu? Canonical has a right to change the experience to how they see fit.

    Something that needs improvement without saying: communication. Something’s hitting the fan and we’re hurting people that don’t need to be hurt. We’re all supposed to be siblings, anyhow.

  114. yman Says:

    “That might be your problem right there: the messaging (at least from my uninvolved point of view) seems to have been ‘here’s a fully-formed piece of software, take it and call it GNOME’. For all gnome-shell’s failures in this regard, of which there have no doubt been more than a few, I don’t think you can really blame GNOME even primarily, let alone single-handedly, for not relying on a fully-formed, externally-developed[0] brand new UX and using that to execute a 180° turn in GNOME UX.”
    Canonical wasn’t asking GNOME to ditch GNOME Shell in favor of Unity, they were asking GNOME to acknowledge Unity as a GNOME app, just as Rhythmbox and Banshee are both regarded as GNOME apps. Instead it was regarded as an alternative not to one component in GNOME but to the entire project.

  115. mark Says:


    There is no need for copyright assignment for any app to use appindicators. For the core implementation, yes, we do require a contribution agreement, as do lots of other free software projects. Most recently, we were able to relicense libzeitgeist to lgpl v2.1 so that it could be included in Gnome because of that policy, rather than the usual sham of sending out a lot of email then illegally changing the license (which is what we’re having to do on the wiki, since there was never any assignment). However, I think it bears pointing out that (a) we were proposing it as an *external* dependency, so Gnome’s rules don’t apply (there are external dependencies on all sorts of stuff that would not fit on and (b) there are plenty of contributors to Canonical projects that do not work for Canonical, so I don’t think they are starved for input.

  116. mark Says:


    Whether one assigns and receives a wide license back, or gets a wide license, is a matter of taste for different lawyers in different jurisdictions. It has EXACTLY the same end impact, arguably. And the former has a much better result: someone still has the integral copyright on the code. Diluting that integrity is a waste, and if you accept wide licenses, achieves nothing.

  117. srinivas v Says:

    competition: is not what u see. U are just seeing one side of it. The side which appeals to you better. What u are doing to the free software world is what happened to the USSR. Now that is an example of internal competition gone awfully wrong. By telling that internal competition is good. U are supporting all internal competitors to come out and start talking individual goals instead of the collective whole. Gnome project has always worked towards simplicity rather than radical changes. It is upto u to implement radical apps/extensions over this simple and clean framework.

    U have not contributed anything radical to gnome development. All of a sudden u are acting godfather to the gnome project. I really have not found any problem with the gnome developers or how the project is being handled. U want more control over gnome development and also want to guide the gnome project according to ur whims. for the past 9 yrs I have loved the gnome dekstop environment and today also it is my favorite desktop. It is sturdy and stable. What is lacking is interesting apps and extensions which people like u are supposed to do.

    Mark, stop giving management tips to others. Concentrate ur money on radical ideas, apps and extensions to vanilla “free as in freedom” software and remember to keep them “free” so that the next Mark will have something to start on. Stop cribbing about the lack of support from gnome developers for Ur unity shell. unity is ur baby. make it grow. If u believe thats what gnome should have done, time is the answer. The whole world will see. what about ur support/commitment for GNU/BSD and GNU/Minix?

  118. Alexandre Says:

    One thing I find amazing is that Canonical actually has a manager for contacts with upstream projects. Isn’t *he* supposed to handle all this?

  119. Manish Sinha Says:

    I did not want to comment here, but going by same logic, you are choosing the side which appeals to you better. Nothing different from Mark.

    WRT USSR, I would say – wrong analogy, USSR’s case was different and not about internal competition. Point moot.

    >> It is upto u to implement radical apps/extensions over this simple and clean framework.
    Don’t call is simple and clean please. I have used it in GNOME 2.x. Some parts are great, some parts needs more work. If you used GNOME stack, you would know which components were a pain to use for development(still I never screamed at GNOME devs for those). I will look into GNOME 3.x technologies and make my opinion.

    >> I really have not found any problem with the gnome developers or how the project is being handled.
    Not questioning your capability – please tell your contributions to GNOME and which team you interact with more.

    >> which people like u are supposed to do.
    Supposed? So you can employing Canonical on contract and paying them to get it done. This is ordering.

    >> vanilla “free as in freedom” software and remember to keep them “free”
    Unless a software/library is not free, it cannot be included in Ubuntu nor in other distros.

    >> Stop cribbing about the lack of support from gnome developers for Ur unity shell.
    If you understand the topic or followed it, the bickering is on appindicators rather than unity. I use the word “bickering” since I am not finding this whole drama useful or productive.

    >> what about ur support/commitment for GNU/BSD and GNU/Minix?
    Mark is working on Ubuntu and Ayatana. Someone should take up the work of GNU/BSD and GNU/Minix. Ready to step in and help?

  120. Kang Says:

    Mark: there is one thing I do not understand: Why does Ubuntu put so much efforts cooperating with Gnome instead of KDE? In my view for each release the changes in GNOME is too small to be noticed, and on the country KDE is doing something amazing. In additional in KDE there are many top-quality applications, such as Amarok, Okular. KDE is also far more open and cooperative than GNOME. The progress of GNOME is too too slow and there is not hope in competing with the progress of other desktop such as Windows and Mac (In fact, the GNOME desktop always gives me a feeling of Windows 9x era.)

  121. Jef Spaleta Says:

    The irony is it appears that the Canonical kernel teams is actually doing a much better job. They use git internally day to day to interact with mainline kernels. Didn’t a Canonical employee pick up the ball on AppArmor and push it forward upstream in consultation with existing kernel devs?

    If we looked closely now I think we’d see improvement in terms of engagement with kernel devs in the last 2 or 3 years due to internal process improvements inside Canonical’s kernel team.


  122. Alexandre Says:


    > The irony is it appears that the Canonical kernel teams is actually doing a much better job.

    Yeah, the RT kernel in 10.10 is plain brilliant for example 😛

  123. Nathan Says:

    Gnome was doing fine for years before Ubuntu came along, and it will still be doing fine long after Linuxers have moved Ubuntu to the footnotes list.

    Canonical can dress this message up however it likes, but what I’m seeing (as an outsider to both parties, I’m just a Linux lover) is that Shuttleworth thinks he’s the boss of whatever he touches.

    Shuttleworth, Ubuntu does not own the Linux community. I’ll tell you what Linux users have been telling each other since time began: If you don’t like Gnome, don’t use it.

  124. srinivas v Says:

    Sri. Manish Sinha

    “I did not want to comment here, but going by same logic, you are choosing the side which appeals to you better. Nothing different from Mark.” –> Thats the whole point. People should be taking sides to foster “competition”. I have taken the side of the upstream. U have chosen its application. thats all.

    “Don’t call is simple and clean please. I have used it in GNOME 2.x. Some parts are great, some parts needs more work. If you used GNOME stack, you would know which components were a pain to use for development(still I never screamed at GNOME devs for those). I will look into GNOME 3.x technologies and make my opinion.” —> If it were pain, then Ubuntu would have gone with plenty of desktop environments out there. Ubuntu is still with gnome(primarily) and will continue(I believe) using it for its simple and clean design “only”.

    “USSR”, point is valid. —-> The complete ripping of it is owed to “gaining individual glory” instead of the collective whole. Every small state wanted to make a statement that he is the reason for the success of the country. But the real reason for it as we “all” except u know is that the “sum is bigger than the individual parts that make it up”

    “Not questioning your capability – please tell your contributions to GNOME and which team you interact with more”. ——-> You are welcome to question my capability, but only as a user. As a user I have found that the gnome project has met all my requirements to give a usable desktop with other plethora of “free as in freedom” software kit.

    “Supposed? So you can employing Canonical on contract and paying them to get it done. This is ordering.” —–> I would take it as ordering, If I really succeeded in getting Sri. Mark implemented my ideas.

    “Unless a software/library is not free, it cannot be included in Ubuntu nor in other distros.” —-> Out of context

    “Mark is working on Ubuntu and Ayatana. Someone should take up the work of GNU/BSD and GNU/Minix. Ready to step in and help?” —–> I am ready to step in at any time as a user. If Sri. Mark is making money out of GNU/Linux then it is high time he starts contributing it in supporting other “free as in freedom” operating systems to increase the periphery of the freedom environment. Now, That comment was a request to the richest man in the GNU/Linux land.

    Thank u for ur concern for Sri. Mark

  125. Jef Spaleta Says:


    Great references!

    Is the audio from the 2009 DesktopSummit presentation available? Perhaps the Q/A? Did you attend the session and perhaps the Q/A? I’d like to find a handful of people and ask them independently a series of questions concerning what was discussed in the Q/A.

    Is there any record at all in any gnome mailinglist or irclog concerning any discussion about integrating appindicators in the context of that desktopsummit presentation? I am not saying they don’t exist. I’m saying I’m having trouble finding them if they do exist. And I fully expect there to be some leakage into public gnome channels if there was a good faith effort to get this into GNOME.

    I would have expected to see some discussion about the potential of libappindicators listed here:

    ZG gets a shout-out there(contrary to Mark’s unsupported references to backstabbing and blocking by unnamed individuals inside GNOME)

    The existence of appindicators as a technology isn’t the question in April 2009. Nor clearly was the existence of ZG. The appindicator technology was certainly known to exist by the DesktopSummit. The question is what was communicated in terms of an expressed desire from anyone from Canonical in seeing this integrated into GNOME? So far the only references to direct conversations between Canonical and GNOME people in the timeframe are private discussions relayed to us by Mark, who was not a party to the discussions and is relaying them second hand. These recollections seem to be in dispute by those who may have been a party to them.

    If I had walked away from the 2008 face-to-face conferences with an understanding that I was being encouraged to work on a technology and expected to be able to hand it back over at some unspecified time sans any feedback on submission, I would have certainly made an effort to confirm that agreement with the wider GNOME but verify. The lack of a reference to my tech in that overview planning document would have raised a red flag for me and I would have attempted to ping GNOME in a public setting and get some clarity on the relationship. But that’s just me I guess.


  126. Ability vs Authority, Ubuntu vs Gnome « theNthDoctor's Better and Worse Ramblings Says:

    […] blogs like this with respect to your involvement in a community like this just makes you look […]

  127. Jef Spaleta Says:

    @Alistair Withworth,

    Can you provide a link to the specific blog posts in question?


  128. Manish Sinha Says:

    >> If it were pain, then Ubuntu would have gone with plenty of desktop environments out there.
    >> Ubuntu is still with gnome(primarily) and will continue(I believe) using it for its simple and clean design “only”.
    I am not sure if you have even heard about KUbuntu, XUbuntu, LUbuntu etc.

    >> Every small state wanted to make a statement that he is the reason for the success of the country.
    Are you sure you that this was the main reason for USSR split. BTW I don’t want to discuss those political issues here. Let’s be on track.

    >> As a user I have found that the gnome project has met all my requirements to give a usable desktop with other plethora of “free as in freedom” software kit.
    We were talking about “I really have not found any problem with the gnome developers or how the project is being handled.” which meant that how the whole project is managed from the developer perspective. These are the people who have to deal with the gnome ecosystem mostly.

    >> “Unless a software/library is not free, it cannot be included in Ubuntu nor in other distros.” —-> Out of context
    It is in context. You asked for it to be free. I ask why wont it be free. U1 server side was an exception, which doesn’t mean every software they make with be closed. If it is not open then it cant make its way in distros

    >> I am ready to step in at any time as a user.
    And who will develop the software and OS? I was asking for that role.

    >> If Sri. Mark is making money out of GNU/Linux then it is high time he starts contributing it in supporting other “free as in freedom” operating systems
    Managing one OS is not so easy. Talk to Gnome devs or any other distro devs. It is a pain. You want them to explore using other kernels too. Do it yourself if you want.

    >> Thank u for ur concern for Sri. Mark
    I am not paid by Mark to support him. I just watch all these things unfold, but I checked your blog – full of knee-jerk reactions against Ubuntu and conspiracy-theories of what Mark is going to do next with lots of factual errors about Ubuntu. It is these things which made me correct you which looks like I am taking side of Mark just because I use Ubuntu.

    By the way, lets take this discussion off Mark’s blog comment. Our discussion is going offtopic. Take it in your blog if you are willing. I will comment there.

  129. Andre Luiz Says:

    MarkShuttleworth my opinion the interface unity is ridiculous!
    I repeat
    is ridiculous!
    gnome is higher than unity at this interface!
    your company is a company that only thinks about money
    little help in the kernel

  130. Jesse Says:


    I think your observation is true, but it can be explained by a few things. For one, some in the kernel community is known for being terrifying in their criticism/rejection of contributions that are not acceptable for technical and/or procedural reasons, so even casual observers like myself can see what the correct protocol is for working with them. By not following it, you not only risk provoking their wrath, but you’re contribution is dead in the water. That’s clear. And, Canonical certainly must have had the motivation to get AppArmor into mainline, as the cost of maintaining external patches to the fast moving kernel is high. So they made the effort by following the known protocol with practices like you mentioned, and were ultimately successful.

    I don’t know what the key is for GNOME. More caustic and offensive (but technical and on point) feedback? I don’t know if that would work, as one key difference appears to be of motivation. Canonical doesn’t have the same kind of incentive to work with the GNOME community. libappindicator was something developed separately by them, and there’s no extra cost to them in keeping it that way — indeed, there are barriers (eg, copyright assignment), unacceptable to many, to even contributing to it!

  131. Jesse Says:


    It’s not about using appindicators. libappindicator is not something for end users, at all. It’s a library you expect open source software developers to adopt for their project, which Canonical 1) did not include them on during the development of, and 2) makes them sign over copyright to you if they do want to make changes now (this is open source, after all). Both of which are plainly unacceptable for a project like GNOME.

  132. pt Says:

    lol this reference to USSR, the failure of USSR is similar to how Gnome runs. A central bureau makes all the decisions, which leads to poor decisions being accepted and expanded on. The non capitalistic model is where each entity makes their own decisions then market (i.e. customers) decide which is best decision.

    Thats how it should be, how is it Ubuntu came out of nowhere and gobbled and expanded the Linux desktop market? They are far more in touch with public than this Gnome foundation.

  133. pt Says:

    typo above “The non capitalistic model is where ”

    should be “The capitalistic model is where “

  134. What does the user see? « fishsoup Says:

    […] Mark argues that GNOME should be a place where we have internal competition. But his idea of internal competition seems to be competition between different end-user experiences. His entrant into the competition is Unity, an environment with a user experience designed completely in isolation from GNOME. The other entrant would, I suppose, be the GNOME 3 desktop that GNOME has created. […]

  135. Jef Spaleta Says:


    I think Micheal Meek’s talk at FOSDEM concerning the approach libreoffice is going to take towards corporate sponsored contributions is very very illuminating. I think we should all watch very carefully how libreoffice as a project is able to thread the needle and be welcoming to contributions sent in in good faith while at the same time being actively “ungrateful” for anything which is contributed with copyright assignment strings attached. (Yes he did say _ungrateful_ in his talk) How they hold the line and keep a welcoming tone will be instructive. How Canonical chooses to interact with them will also be instructive.


  136. Nick O'Bact Says:

    Ok, everybody takes a deep breath. This is software, guys. No need to start hating each other on who rejected what, who runs what in a certain way, who is forking what, or whatever the hell is this flamewar about these days.
    Why not let everybody do its own thing? Unity vs Shell? Who cares. Both of them will be used, developed and have a large number of contributions from amazing people.
    Both have their vision (which are quite similar, as it turns out) and both have their backing.
    What sticks is that those are just shells. Not forks, not competitors. Those are just pieces of UX. Maybe it’s time to realize averybody’s been fighting on the side of lines of code.

    Great lines of code, certainly. Important lines of code, sure. But still. Text in a file.

  137. nathan w Says:

    a very interesting read, mark. i agree that there could have been more cooperation throughout the process, and that competition is healthy for open source projects. more variation, more room for choice by end users to decide the direction. for that reason i don’t understand at all the “sound and fury” as you put it (not very nice, btw…), that Ubuntu’s move to in house some design decisions (Unity) has caused.

    it’s interesting to read commentary stating that this discussion consitutes the opening shots in a “war” between Red Hat and Canonical. it seems a puzzling viewpoint, because from my perspective it seems the two companies are pursuing different markets in their primary efforts. RH being solidly focused on server environments (enterprise use case), and Canonical pursuing the usability and UI improvements needed to make desktop linux (consumer use case) a reality.

    looked at from this perspective it would seem to call for collaboration around interoperability and integration support, rather than wasting effort in these flamewars that roil the waters, but truly signify, and more importantly achieve, nothing.

  138. Nassir Says:


    I did enjoy the Ubuntu fonts and believe it was a wonderful thing when I used Ubuntu (now using Debian). The Ubuntu interface before Unity was also good design.
    But all this work on Unity, all the money and effort spent on reinventing something which is darn good as it is (Gnome’s UI) could have been better spent on solving the more important problems Linux faces.

    How about helping with the FOSS graphics driver development?

    How about a couple of references here and there to the Free Software Movement (for example, on front page) to increase people’s awareness of the philosophy behind the rock upon which Ubuntu is built (GNU)?

    How about working on finding a way to reconcile game development with the FOSS model and get game developers to consider making an Ubuntu port of their games (no, I’m not talking about Steam)?

    How about more employees in the testing and bug fixing phase during development? Many people see Ubuntu as a bit buggy.

    I’ll probably be flamed to hell by fanbois but this is my opinion.

  139. On cross-project collaboration | Maemo Nokia N900 Says:

    […] is currently quite stern discussion going on between GNOME, Canonical and KDE about collaboration on the free desktop. Angry words have been written, and I believe much […]

  140. Design Monkey » The truth about GNOME design Says:

    […] GNOME hackers. There is no cabal. All design decision happen in the open. We do not feel like competing with each other, but we want to collaborate and have fun on creating the future instead. Join us! […]

  141. Jim Campbell Says:

    Hi Mark – I want Ubuntu to succeed, and have even had a change of heart and have recently signed the Canonical contributor agreement. Here are a few (relatively) brief points, though (I do not mean any of this in an “attack-y” way. If you feel like I am attacking something, please skip to the last paragraph. I want to be constructive):

    – To me, competition within Gnome would likely occur with sources placed on, not from sources placed on Launchpad. I think the Canonical contributor agreement makes Unity an even more “outside of GNOME / competing with GNOME” effort, too. Why would someone who works for Red Hat or Novell contribute code that would then be owned by Canonical? (Yes, I know they could just fork Unity – it is still Free software. We saw this happen with OpenOffice-to-LibreOffice).
    – The FSF compyright assignment agreement and the one from Canonical don’t really provide the same protections for contributors, so it’s not really accurate to mention them in the same breath, as if there were no difference between them.
    – You commented on my blog post about Canonical’s Qt work and copyright assignment, so I’ve done some reading on other contributor agreements. You’re right that there are some similarities between what is in Canonical’s agreement and what is in other similar agreements. There are some differences between the Canonical agreement and other agreements, though, that I think may discourage community involvement–particularly community involvement that comes from people who work for other organizations. Maybe the Project Harmony can help with this somehow?

    Would there ever be an option for something like “Ubuntu Labs,” or “Canonical Labs,” for incubator projects that could be done more out in the open? I know that some work needs to be done in-house, perhaps in secret (even Meego has done this … there’s so much competition out there now), but doing some work more out in the open, perhaps even on w/o requiring the contributor agreement, could help smooth-out this rift a bit, and could benefit both GNOME and Canonical.

    In closing, I’ll thank you x1000 for putting your earnings, time, and leadership toward spreading free and open source software. I never see you or Jorge or Jono complain about putting all of your moolah and effort into this thing, and I know it’s all a big gamble. Thanks again to you and the Canonical team for doing what you are doing.

  142. Olivier Crête Says:

    As you and your employees have been saying for years, the goal of Ayatana is to differentiate Ubuntu from other Linux desktop distributions You’re trying to have your cake and eat it too. Either you’re differentiated or you’re part of the GNOME project, you can’t do both. Red Hat learned it the hard way with their horrible enterprise kernels of the past, Novell learned with gnome-main-menu, Ximian with Red Carpet, Google had Wakelocks (although Google seem to be slow learners). The list of companies that tried to differentiate by doing their own thing in their corner is long, the list of companies that succeeded is very short (I can’t think of a single one). The only real commercially successful Linux distro (Red Hat) is perhaps the most bland, they don’t differentiate by the content of their distribution. It’s plain GNOME, plain everything. They differentiate with top notch support, etc.

    Good luck to you.

  143. The libappindicator Story | Be the signal Says:

    […] participating in much of that discussion, Mark Shuttleworth (founder of Canonical and Ubuntu) responded in a strongly-worded blog post on Thursday […]

  144. Sekar Says:

    Oh my, grownups acting like 2 year olds.

    I don’t understand, why try so hard to collaborate if it’s not working out? Just fork GNOME and call it something else. Just move on. Sure, people will abuse you again for forking, but seriously people need to understand FOSS code allows exactly that. The fact is most people in FOSS don’t care too hoots about their code until someone tries to make $$ out of it. And when $$ is involved, they want the force the money maker to listen to their thoughts and ideas. Typical human behavior. I hope people see it now that people developing FOSS do not have some major ideals or something.

  145. enedene Says:

    Good post.
    You can’t force Gnome to do anything. It is good that you’re going apart. Ubuntu is made for ordinary people and and Gnome is not ready for non-geeks.
    For example, a women asked me if I could help her with her computer, she had viruses. I’ve suggested Ubuntu, she agreed. Although she was very satisfied, a year after it was windows back on her computer. And how did this happen? Since I don’t live near by, she had a problem, asked help from a local “hacker” and instead helping her, his solution was to install the Windows back, because “Linux is for geeks”.
    And what was the problem? By accident she deleted the upper panel. Since this happened to me when I first started using Linux couple years ago I know how difficult a problem for newbie that can be. The same thing happened to my sister, which fortunately asked me for help.
    Try changing resolution in gnome, and then change it back, see what happens to icons? They stay on a position of smaller resolution, so the clock is now instead right, on the middle of the screen.
    Some people want to try to modify and think it will be easy to bring it back as it was… nothing could be further from the truth…
    These are REAL WORLD usability issues and all this time gnome has no idea or intention seeing them. THAT is what turns people away from Linux once they try it and this is precisely why you have to go as far as possible from people making decisions in Gnome.

    You always need to look at things, not as they are in principle, but what they are in real world. For example, in principle one of biggest advantages of free/OSS software is it’s price. In real world, if ordinary people choose from Windows and Ubuntu, Gimp or Photoshop, they use one which they find better, the price is the same, pirated Win/Photoshop and Ubuntu/Gimp.
    Yeah, “it’s wrong and illegal”, so was alcohol during the time of prohibition, but did the problem go away? No. So will not the pirated software, which is the biggest ally of priced products since the biggest advantage of free/OSS disappears.
    But this is good, this makes your mission harder, you need to be best in quality, but in the end we all get a better product.

    “Regulators across the Western world failed to ensure that financial system participants had sufficient buffer, and thus the financial system failed to handle a shock. The same is true in many cases where there’s a market failure. Regulation, like leadership of internal competitive dynamics, is very hard. But the alternative (central planning) has all the difficulties and none of the benefits.”

    Not lack of regulations but central planning of US government led to the economic crises. Government vouched for peoples loans who would never be able to get them from the banks in the first place. This made these credits a great deal because who can you trust to return their obligations if not the government… it was all turned upside down and that’s what led to crises. For that matter every economic crises, when government interferes it ends up in disaster. Watch Peter Schiff on Youtube, a real world economist/broker, not some collage professor who can’t fun a successful local store, let alone world finances.

  146. Шаттлворт и Нири рассуждают о конкуренции и взаимодействии между проектами Ubuntu и GNOME | – Всероссийский портал о UNIX-системах Says:

    […] днях Марк Шаттлворт поделился в своём блоге очередными мыслями по поводу здоровой […]

  147. Fedora 15 & Gnome leadership « Alex's Blog Says:

    […] but it’s enlightening (I think) to read Mark Shuttleworth’s latest “Internal competition is healthy, but depends on strong and mature leadership” alongside Mark Wilcox’s “What happened to Nokia?” of a month ago. […]

  148. srinivas v Says:

    sri Manish, sri Mark,
    I would draw ur attention to this post.

  149. Shuttleworth Criticises Gnome Leadership Says:

    […] another scathing blog post, Ubuntu leader Mark Shuttleworth has criticised the leadership of the Gnome project for not […]

  150. Pasi Lallinaho: Internal competition and collaboration in Ubuntu | Christian eBuddy Blog Says:

    […] . /**/Mark Shuttleworth recently wrote about internal competition in Ubuntu in his blog. Mark says:In Ubuntu, we have a lot of internal competition. Ubuntu and Kubuntu and Xubuntu and […]

  151. jargon Says:

    Well, so much for Openrespect, hm?

  152. James Says:

    Jim Campbell: +1; internal competition means using GNOME infrastructure and norms, so git, no copyright assignment etc.

  153. Шаттлворт и Нири рассуждают о конкуренции и взаимодействии между проектами Ubuntu и GNOME Says:

    […] днях Марк Шаттлворт поделился в своём блоге очередными мыслями по поводу здоровой […]

  154. Pat Gunn Says:

    I think having different “shells” really is not a big deal. So long as GNOME apps behave properly across both shells, the platform remains whole; all there are is some different look/feel issues, none of which are unfamiliar to Unix. In times past (and currently among the most technically clued users), people use an incredible variety of window managers, even if they might run some GNOME and/or KDE apps.

    What I’m much more concerned about is Canonical’s intention to ditch X11. This will be incredibly harmful to the Unix platform, and IMO any Linux distro that does this without *much* better reasons than what Canonical has provided should be abandoned ASAP. As soon as alternatives to X11 become commonplace (provided this plan is not abandoned), we lose the common platform which has bound most varieties of Unix together for a very long time. We lose network transparency (and the managability that comes with it). We lose window managers. As software developers, we need to decide what we’re targeting for something much more critical than ALSA versus OSS. There’s no point squabbling over fragmentation of non-library parts of GNOME when Canonical is going to make a shot at destroying a much more longstanding, foundational, and important consensus.

  155. Ian Says:

    I think the main mistake Mark made was to use GNOME in the first place.
    GNOME were always a “me too” desktop group who produce an inferior product from a technical and design point of view, it was just a matter of principle, and not technical, to create something that didn’t use QT because QT was closed at the time. I bet they feel a little stupid now QT is open.
    GNOME, Canonical – learn to collaborate and stop working in your elitist silos, use fd.o and discuss with everyone when a spec doesn’t quite meet your needs because at the moment, GNOME especially, is behaving just like Microsoft by going off using their own specifications/standards. Perhaps we could rename GNOME to MSGNOME

  156. Canonical Vs Gnome: E agora o KDE “mete a colher” na história… | Cleiton Lima Says:

    […] Banshee a não doarem nem esses 25% para a Gnome Foundation. E, por último, o senhor Shuttleworth liberou um post em seu blog criticando, dessa vez diretamente, a liderança do Gnome enquanto Projeto, tendo o apoio de um dev […]

  157. mark Says:


    Canonical has never said the goal of Ayatana is to differentiate Ubuntu from other Linux distributions. The goal is to explore ideas there around the idea of “focus vs awareness”, and the balance between those two goals. That work is relevant to lots of distributions, code flows into lots of distributions, and ideas flow further. That meme, if you made it up, is broken, and if you’re just passing it on, then it’s toxic.

  158. mark Says:


    There’s lot of work to do, yes. But our focus has to be on solving the question “why should end-users and the ecosystem care about Linux?”. If we can do that, then lots of the problems you’re describing will solve themselves. We’ve seen most major manufacturers move to open source once they have a big enough market for their drivers, for example, because it’s simply the only way to make things work everywhere that Linux wants to be. But getting a big enough wave of adoption – THAT’s the challenge, and that’s where we’ve focused all our efforts.

  159. mark Says:

    @Nathan W

    You’re right, the markets of Canonical and Red Hat are different, with little cross-over. But the Red Hat *desktop* team, which is not specifically tied to Red Hat’s markets, would naturally feel differently.

  160. In fact, anyone can do it « Colin Walters Says:

    […] Mark says: I have little optimism that the internal code dynamics of Gnome can be fixed – I have seen too many cases where a patch which implements something needed by Unity is dissed, then reimplemented differently, or simply left to rot… […]

  161. Theo Stauffer Says:

    @Mark: Boet, I’m from SA like you are, so you’d tend to normally get my sympathies more than people from GNOME would but honestly, from where I’m standing, you’re on pretty shaky ground. Some points:

    1. You talk about wanting to improve things for the end user. I’m not too sure which end users you’re talking about, because fights like this generally only make things worse for those who, you know, just want a free, easy to use OS with lots of apps. Linux apps are still generally behind in both functionality and usability compared to Windows and Mac apps. You should be working here to improve this, not making drama about Desktop UXs.
    2. From my point of view as a mostly Mac OSX and sometimes iOS user it’s looking like you’ve traded the Linux desktop Windows catch-up story for a Mac OSX/iOS catch up story. Good luck with that because while Ubuntu definitely looks better than any other distro, it’s still miniscule in the real world. Tell me how many people have left Windows and Mac OSX for Ubuntu? NO, I didn’t think that many did either.
    3. Again, personally, U10.10 will be my last Ubuntu desktop for the foreseeable future. I had the mobile version of 10.04 and 10.10 on my netbook and I must say that the constant UX changes every release finally got irritating enough that I decided that I’ll go with something that I know works. In other words, your constant push for new and better is alienating some people. I don’t know how many, but enough to start pushing the more flexible distro, Mint closer to where Ubuntu is, closely followed by OpenSuSE and Fedora.
    4. Linux does not and will not supplant Windows or Mac OSX in enterprises, neither currently or for the foreseeable future. Canonical is STILL not generating any profit and the drama behind the Unity and Wayland decisions will not help your cause.
    5. Taking a step back from the Must-release-new-crap every six months schedule would allow you to take more time to work better with those people you’re having difficulty working with. The end result would benefit us all, not just you.
    6. You’re not Steve Jobs, and you really don’t want to be. Keep that in mind. Also keep in mind that he has a lot of excellent, working software on his platform. You don’t and until you do, all your desktop efforts will come to naught compared to the commercial software world.

  162. rodmra Says:

    @Fedora 15 & Gnome leadership « Alex’s Blog

    I agree. Internal competition and Nokia history:

  163. humble user Says:

    > Canonical has never said the goal of Ayatana is to differentiate Ubuntu from other Linux distributions

    And if so (which isn’t the case), what would be the problem with such a statement, is it not allowed to differertiate from others, is there a law or something?

    However, the meme (as often happens) might be spread out of context.
    Anonymus user said:

    and Canonical blog said:

    A little search finds argy-bargy goes on ever since.. and it gets somewhat boring now.
    Ubuntu: “We have no plans to fork GNOME”

    Discussion on gnome mailing-list about application indicators.

    I’m not pro Unity nor Gnome-Shell. I’d like to see ideas develop and innovate in products, I don’t care who implements what, just let each one it’s freedom of choice, users and developers. If you’re aren’t able to collaborate then make it on your own. Life is too short for wasting time in destructive matters.

    Hope that Apple won’t sue Canonical for imitating Overlay Scrollbars.

  164. pouet Says:

    you should buy Qt/KDE and leave gnome alone.
    The gnome guys are too smart… 😉

  165. Dread Knight Says:

    Unity is really awesome. Too bad for Gnome…
    Keep up the good work!

  166. srinivas v Says:

    humble user says “Hope that Apple won’t sue Canonical for imitating Overlay Scrollbars.” Oh my God, I just congratulated Mark in his previous post about the scroll bar thingy. Well Mark, 0 marks for that and 100 for copying the idea once again. btw have u applied for his post at apple?

  167. Ubuntu Linux and GNOME: The Disputes continue | ZDNet Says:

    […] went on to state that Mark {Shuttleworth] wants GNOME to have “strong, mature technical leadership.” Neary then stated that “My understanding of GNOME is this: GNOME does not have […]

  168. Timeline: The Greatest Show on Earth | Be the signal Says:

    […] least July 2010, Mark has made a number of claims about the User Experience Hackfest. They are all roughly of this form: We had described the work we wanted to do (cleaning up the panel, turning panel icons into menus) […]

  169. IS11S08 a IS11S10: pinceladas de 3 semanas con mucho software libre - No sólo software Says:

    […] El post que hace explotar el debate público, de Mark Shuttleworth, dueño de Canonical. […]

  170. John Says:

    These are commercial wars. The people who are fighting are having some Money (consulting opportunity) at stake. They are trying to shame Canonical because they are jealous of its success.

    All these guys have commercial company to back them up and these wars are not going to stop!

    Linux desktop community has a history on these flame wars. They are so used to sham Microsoft. Did it help Linux desktop?

    Canonical brought some light to Linux Desktop by bringing us Ubuntu. Later Canonical learnt that Gnome is not the best for end users. Canonical should have chosen KDE. They didn’t, because Mark Shuttleworth was a long time Gnome developer. Mark, you made a wrong decision by making Ubuntu default Gnome in the first place. Had you associated with KDE, we would have got a great linux desktop which Apple and Microsoft could not even dream.

  171. KDEer Says:

    I hope this is the first step towards Kubuntu becoming the primary flavor of Ubuntu. It’s been obvious to me for a long time that KDE is future-oriented, while GNOME is half stuck-in-the-past and half trying-to-reinvent-the-wheel. KDE is making real progress, while GNOME is half stagnant and half self-absorbed and deluded.

    This is my perception of the whole–I’m sure there are good folks in GNOME, but it seems like the inmates are running the asylum. Hopefully the good guys will abandon ship before the cookie crumbles, while the rest of them keep shooting themselves in the feet. I’m sure KDE would be glad to have some more good programmers.

  172. gefira blog » A kitty dies when someone says “experience” instead of the “user interface”. Aren’t you aware of it? Says:

    […] of “user interface”? I didn’t even know how wiedspread it was prior to reading the Mark Shuttleworth’s article and all the replies on various […]

  173. Nishant Says:

    The whole pain here is that Ubuntu GNOME users get nice patched apps with appindicators, and a smooth KDE system tray, while my gnome apps have fugly and out of place menus in KDE.

    Also can someone post a link to the ‘incompatable’ GNOME equivalent of the StatusNotifier Spec

  174. Ubuntu Linux and GNOME: The Disputes continue | Linux Desktop Says:

    […] went on to state that Mark [Shuttleworth] wants GNOME to have “strong, mature technical leadership.” Neary then stated that “My understanding of GNOME is this: GNOME does not have technical […]

  175. A. Peon Says:

    Wow, lots of noise here. I just feel like popping up with a “me too” re: GNOME Shell looking like it would actually have some major benefits to my workflow [in the legal industry]. My experience as a GNOME user for about 5+ years makes me suspect it’s a mess underneath, but if it’s “good enough”…

    Unity, in turn, looks good for grandma. Which the world also needs. [But I’m not grandma, so prerogatives and all that.] Unfortunately it does get you tagged for chasing Apple, because Apple at least has a product, while environments geared toward supporting a “workflow” beyond a bunch of terminal windows are a bit thin on the ground right now [those are So ’90s!].

    Any chance we can stop pretending the UI (“UX?”) is the OS and just accept that one distro of kernel and display server can support many choices of desktop environment? I get that K/X/Edu/… provide a final layer of spackle to resolve conflicts come release days, but how about we work on cutting it down rather than forking it out?

    UI should be more of a Browser Ballot situation than some sort of mark you must wear for life. Even if everything is doomed to have to deal with what it can scrape up over DBus / what can be agreed upon through the likes of Now that it’s *not* expensive in hardware terms to keep all of GNOME, GTK, KDE, Qt, and Java, and Flash, and Android, and … [and a couple VMs while you’re at it] resident at once … well, if people flopped around more, maybe they’d be more inspired to fill the gaps in each.

  176. A. Peon Says:

    Eh, one other undignified “waah” while I’m posting anonymously (though this might identify me!) – my personal preference for GNOME has little to do with the underpinnings and a lot to do with GTK/GNOMEy apps making relatively good use of language while KDE jumped straight into the “hieroglyphic” iconic strafe-for-tooltips approach.

    That’s been changing a bit lately – some of the most useful GNOME apps I’m using just now have no “Icons and Text” or “Text-Only” modes for their toolbars – but it might still be worth pointing out.

    [Of course Unity gets by without trying to provide the direct-access-to-‘data’-objects sugar that GNOME Shell is experimenting with, so there’s less call for text if each app has a ‘branding’. I suspect some FOSS devs would rather spend more time writing unique code and less time developing unique brandings, though, especially as the namespace gets used up and we startr appendingr “r” tor everythingr forr ther 2009r modelr yearr.]

    Tangentially: When is any annoying-notification-popup system going to give us something we can click on to scroll and filter the log of Notifications, anyway? And to ‘mute’ through one location rather than having to find where each app hides its options?

  177. nowardev Says:

    mark… you should consider to use kde.
    i guess the develop is faster. because an OS-company must be faster.

    kde lacks on direction , stability focusing, but the engine is good. you could be the man to give a direction.

  178. Andy Says:

    Maybe you should ask third parties to get a true and fair view. How does e.g. Pcman view the developments in the Linuxdesktop standard group?

  179. S04E02 – Stranger in a Strange Land – MP3 LOW | Ubuntu Podcast from the UK LoCo team Says:

    […] du-jour “CanoniGNOMEAppIndicatorGATE!” (snappy name needed, apply within!) Whereupon the past is woken up and given, a kick in […]

  180. NoahY Says:

    Just tried Unity for the first time tonight. Looks great, feels slick, even though I think like many Linux users, I may not leave my compiz cube + Docky for it. I would recommend it for a new Linux user though. Mark, you’re the man, go with it. I’m just grateful for what you’ve done for the Linux community; I don’t see where people get off attacking your work or that of the Canonical team. What happened to congratulating you all on your dedication and self-sacrifice? Even if I don’t use Unity myself (who knows, maybe I’ll warm up to it – it does have a cozy feeling to it after all), I heartily support the move and the intentions of the Ubuntu team. I think the Ayatana concept is great, the right track for sure.

    I don’t think I would be using Linux if it weren’t for Ubuntu and, of course, Canonical. But then I don’t think Linux would be at all what it is today without Ubuntu. Back in the day, I tried lots of different distros, none really worked for me out of the box until Ubuntu came along. I’m obviously not alone. Having made the switch from Windows for good sometime around Dapper, I can’t bear to use even Windows 7 anymore. Not a single Ubuntu release (not even Edgy) made me regret the move. Your team made my life better; or at least, allowed Linux to make my life better. Isn’t that what’s important? You all deserve pats on the back for what you’ve done to make Linux truly useable in a way that gnome alone never could have. Now you go the next step towards making people’s lives easier and all us nerds can do is complain? I can’t see room for anything but applause – I guess some people just like to complain.

  181. Sopha Mungkunkhunsisuwan Says:

    Sometimes what leader think is not what community can accept. I feel the best way is to make a direct discussion with the Ubuntu and Kubuntu users about their experience with GNOME or KDE. I believe some will give harsh critic. But all of us just wanted the best for Ubuntu or Kubuntu. When you do a poll of suggestion then we will know what is the best for development. The most important matter is to get a true fact and go towards it. Or else Ubuntu will go into disaster which is a sad sight to be seen and a victory for money making company.

  182. kikl Says:

    Well, competition is very good. I have been testing Ubuntu 11.04 beta for a couple of days. There’s lot’s of good stuff to right about, like the improvements in the software centre and even the global menu looks nice, but…. well but. Here are some critical remarks from an ubuntu supporter.

    Switching between applications with the Ubuntu-Dock is a nuisance. The dock disappears and does not reappear when you hover with the mouse in the area of the dock. You have to click the ubuntu-icon twice. I found out by coincidence that it appears when using the window (urg) key. But most people want to use their mouse and there is no convenient way for a mouse user.

    Adding new apps to the dock doesn’t work as expected by dragging and dropping. Unfortunately, neither gnome-do nor kupfer work with unity. But this may have been done on purpose. It took a while to find out and this time was wasted and frustrating.

    The application menu looks slick but coming from gnome I find it difficult to find the apps. I am used to using gnome-do for a quick search. After some frustrations I figured out that urg-a (application search) and urg (file search) does the job.

    There are quite a lot of things we users have to adapt to and learn. I don’t have a final opinion yet, but the process of adapting can be frustrating. A quick and easy tutorial for noobs would be much appreciated.

  183. kikl Says:

    Well, it took me three days to find out that you have to move the mouse beyond the ubuntu symbol way into the corner of the screen in order to display the side dock. These small changes can make a whole lot of difference in terms of user experience.

  184. kikl Says:

    I quit using unity because of two reasons:

    1. I find looking for applications and files easier using the gnome 2.x drop down menus. This is faster – for me – if I am looking for an application.

    2. Several applications like gnome do, docky don’t work with unity. These two applications can actually be replaced. But I use a clipboard manager and the force-quit application in the old gnome 2.x panel. The unity panel doesn’t feature these applications and doesn’t appear to be customizable. I couldn’t find replacements for these applications. Force-quitting an application is something I need to do occasionally. I frequently Copy texts between different applications, which is not possible by default using unity. I couldn’t find a replacement for my clip-board manager parcellite. That was the tipping point for me against unity.

  185. kikl Says:

    Well, I have been experimenting on and off with 11.04.

    Global menu: It has the same drawbacks as the apple implementation. It is not always clear which application occupies the global menu. Therefore, the global menu should start with the icon of the active application. Quit and minimize should pop up when you press the icon. Pressing the icon has already been learned when using the ubuntu dock. Next to the icon the application menu should appear. Hiding the menus under a name just adds to the initial confusion. A non-maximized application should never occupy the global menu. That’s non-intuitive and has no advantages. The only reason for the global menu is a better use of screen space when the application is maximized. Otherwise it is a prime source of confusion, which should be minimized.

    Ubuntu dock:

    1. The dock should appear when you hover over the ubuntu icon. This is a good solution because no conflicts with controls of particular applications may occur. This would be a real advantage over standard docks.

    2. The standard configuration shouldn’t have plural workspaces and the workspace switcher. Plural workspaces are a geek feature. My dad never got the hang of it, therefore I got rid of them on his computer. Geeks should have an easy way of customizing the behaviour of the ubuntu dock.

    3. The dock should not hide intelligently but stay on top all of the time in the standard configuration. This is just confusing for first time users or casual users. I like it, but I know it’s a source of confusion for lots of noobs. The geeks can easily configure the dock according to their needs.

    This is my last point: Both the ubuntu dock and the top panel should be easily customizable for non-pro but advanced users. I don’t like the detour to the compiz settings manager. I like docky and the old gnome 2.x panel a lot in this respect.

    A few words of praise: The search for applications and files has been implemented very well. It makes a lot of sense to use the whole screen space in order to display the most frequently and installed applications. It also looks really good. I like the way the keyboard short cuts are displayed in the ubuntu dock. The keyboard controls are very well thought out. Kudos to ubuntu!

  186. Un análisis objetivo de la disputa entre GNOME vs. Canonical. « gagotrust Says:

    […] añadir más leña al fuego, Mark Shuttleworth publicó en su bitácora personal acusaciones en contra del liderazgo de GNOME, y que el ambiente en GNOME era poco propicio para la […]

  187. Diego Viola Says:

    Please make Wayland a reality, Mark.

  188. A Rift Opens Between KDE and GNOME « UNIX Administratosphere Says:

    […] of not conforming to standards and not collaborating, and Mark Shuttleworth, founder of Canonical, expressed agreement with this […]

  189. Canonical Vs Gnome: E agora o KDE “mete a colher” na história… — Cleiton Lima Says:

    […] Banshee a não doarem nem esses 25% para a Gnome Foundation. E, por último, o senhor Shuttleworth liberou um post em seu blog criticando, dessa vez diretamente, a liderança do Gnome enquanto Projeto, tendo o apoio de um dev […]