It’s about accepting responsibility for your place in the world. For unto whomsoever much is given, of him shall be much required: and to whom men have committed much, of him they will ask the more.
With the dozens, or maybe even hundreds, of engineers in Canonical’s employ now, why do none of them do any of the heavy lifting in GNOME, or in any other upstream project, for that matter?
There’s a difference between Ubuntu and Canonical. The Ubuntu community has obviously done ridiculous amounts of good work in the open source world for multiple years, and will continue to do so. Ubuntu community members are great evangelists for open source. The Ubuntu brand machine is Canonical’s greatest strength, and a world-class model for others to follow. The existence of Ubuntu has grown the pie for open source in general.
But Canonical the company doesn’t get to hide behind the Ubuntu community forever — and you can wrap yourself in that flag all you like, but in doing so, you completely miss the point.
Canonical is supposed to be in a leadership position in the open source world, and leadership demands that you put your money where your mouth is.
It’s that simple.
You want to claim the high ground now, and paint me as a bitter old coot, and imply that my “tribalism” is subtly more dangerous than racism or sexism? I think that’s a bit of a stretch, but okay. I get it. Fine.
But maybe in your next blog post, you could attempt to answer the legitimate question at the center of this debate. Why is Canonical so relatively invisible in the upstream projects upon which Canonical completely relies?
And why is it that, whenever this question comes up, the answer is always some variation on the theme HATERS GONNA HATE instead of actual reflection?
Greg DeKoenigsberg Speaks 
There are two hallmarks of tribalism:
1. Categorical statements about “them”. Let’s see: “none of them do any of the heavy lifting in GNOME, or in any other upstream project”. Check.
2. Attempts to discredit evidence to the contrary of the above. Let’s see (and I have to paraphrase): “Ubuntu has done lots… but that’s not evidence of Canonical doing lots”. Check.
So yes, Greg, this started as tribalism and apparently will continue in that vein.
As for your question: “Why is Canonical so relatively invisible in the upstream projects upon which Canonical completely relies?” you may be familiar with the saying that “There are none so blind as those who will not see.”
Canonical has had a profound and positive effect across huge chunks of the open source landscape: codes of conduct are now common in open source projects. Cadence has spread to more projects, faster, including Fedora. Design is now a first class meme among open source developers that care about users. From the layout of the Fedora desktop to the recently published “future look of Gnome Shell” which is quite clearly inspired by Unity (despite the pooh-pooh’ing of Unity’s design by Shell developers), there is plenty of evidence that Canonical’s work is energising and positive for free software.
Canonical joined the free software ecosystem once many key projects had been established. We’ve set about bringing new ideas, new projects, and new ways of organising people into that ecosystem. If you insist on looking only through the filter of “what existed before” you will naturally miss all the new scenery.
Now here’s the part I feel most strongly: free software needs all the contributors it can get. It needs contributions of all forms, some of which I understand, many of which I don’t. Between you and I, neither of us can afford to make new potential contributors stop and thing “geez, do I really want to participate in this nasty environment”. So let’s both just focus on making the world a better place the way we best know how to. I trust that’s what you really care about, I hope you can trust that I share the same goal.
Mark
“Canonical joined the free software ecosystem once many key projects had been established. We’ve set about bringing new ideas, new projects, and new ways of organising people into that ecosystem. If you insist on looking only through the filter of “what existed before” you will naturally miss all the new scenery.”
What I care about is how much of that “scenery”, as you call it, makes its way back into the hands of all of the other landscapers.
So how long has Fluendo been around? How long has Collabora been around? And why is it that both of these companies, which are both a tiny fraction of the size of Canonical, are so effective at contributing back to the commons that everyone shares — and Canonical, despite its mammoth place in the heavens, isn’t?
There’s no denying that Ubuntu has been influential, and that Ubuntu’s influence has been a huge net good for the world. Continuing well-deserved kudos for that.
But this report is extremely damning, Mark, and I don’t believe for a second that you don’t know and understand that. Upstream projects are how open source gets done. Canonical has ZERO engineers in the top 25 contributors to GNOME after more than FIVE YEARS of reliance upon GNOME as the center of your user experience. Can you not admit, even for a second, that it’s a troubling statistic?
And more to the point: do you intend to do anything about it?
Canonical has no obligation to get any engineers into the top 25 contributors of GNOME (why is your theshold set at top 25 anyways?). If the existing developer base for GNOME has produced something for 5 years that has met Canonical’s needs so well that they never needed to get involved in it’s development, then I don’t see why there is any need to add another chef to the kitchen. Actually, if you get too much influence from downstream users of a technology, then the technology becomes less generalized and more specialized and risks losing value as it’s scope of applicability narrows.
Perhaps saying they should share the monetary rewards upstream is a valid point, but to support this point as being common practice you might consider giving some specific cited figures/facts of other companies that are similar to Canonicals position in the community and show that those other companies contribute resources and/or money upstream.
You actually forgot a third hallmark of tribalism:
3. Egocentrism, or categorical statements about “we” as opposed to “them” (“I”, “Canonical”, and any variation thereof). Let’s see:
“Canonical has had a profound and positive effect […]”; “Canonical’s work is energising […]; “We’ve set about bringing new ideas […];
What do you think, can we “check” that?
And just between me and you, the work you do downstream and that happens to be rejected upstream, it is done solely for your own purpose, and you couldn’t care less if GNOME actually rejects it on the basis that what you’re really going towards is something else completely. You’re basically not working to enrich GNOME from within, you’re merely working to beautify your own distribution from outside, and call the desktop environment something else (I guess these days you call it the Ubuntu Desktop Edition), so that you’re ready to complain when the make-up you created for Ubuntu’s face is not welcome in GNOME. Sooner or later you’ll completely fork it, and will stop pretending to contribute to it collaboratively.
I hope this will not sound too tribal to you.
How does the layout of Fedora’s desktop have anything to do with Canonical? If you’re citing the two panel layout as being it, I’m sorry but I think you must be mistaken. I recall when Seth Nickell and Bryan Clark discussing adding the extra panel when I was an undergraduate. I remember it distinctly because I was rather irritated by it, and preferred one panel! GNOME’s default layout was two-panel by at least January 2003 as evidenced by this usability@gnome.org thread:
http://thread.gmane.org/gmane.comp.gnome.usability/2074/focus=2083
Further in the thread you can see Seth Nickell discussing the history of the decision, apparently inspired by Ximian:
http://thread.gmane.org/gmane.comp.gnome.usability/2074/focus=2083
Seth also cites a 6-month survey he conducted when making the decision in favor of two panels:
http://thread.gmane.org/gmane.comp.gnome.usability/2074/focus=2083
Here is a screenshot of Ximian GNOME 1.4, released in 2001:
GNOME shell’s designs have been iterated publicly since at least the GNOME Boston Summit in October 2008 (http://live.gnome.org/Boston2008) although I remember working on mockups for it in Gimpnet earlier than that. Unity’s design was ‘unveiled’ at an event in Belgium that ran from May 10-14 2010 – please correct me if I am wrong but I found this cited at multiple news sites retrieved via basic search.
I don’t use OS X. I haven’t since 2005, and never as my main desktop – only to use Macromedia Fireworks for mockups before Inkscape was sufficient. So I definitely felt duped after seeing mockups for nautilus and other GNOME apps and being genuinely impressed by them, then recently visiting the Apple Store (a rarity for me) to see what all of the excitement about the iPad was about. I found myself jaw-dropped in front of a Powerbook with a new full realization of how much copying is going on. Myself and several of my FLOSS designer friends wouldn’t be capable of stealing ideas from OS X – because we don’t use it.
Did Microsoft steal ideas from Apple? Did Apple steal ideas from Xerox?
I know one of the designers working on GNOME Shell has left and we have gained a new full-time designer as of this past month, so you will likely be able to see visually in recent mockups evidence of a new set of hands shaking things up. Is imitation occurring? If you ask this question, then I have to ask in response, why are there two systems for there to be one to imitate the other in the GNOME community? I remember that GNOME Boston Summit 2008 when Jon was pleading for more designers to help out. I remember feeling so bad about it I skipped all the sessions after that and sat in a breakout room with Jon, Marina, Owen, and Dan and cranked out a bunch of mockups that are still up on the GNOME wiki.
But…. are such accusations really worthwhile? It’s better to just work together rather than puff up our chests and bemoan about how we are not working together?
I was referring to the order and naming of “Applications, Places, System”, the placement of the Trash and “Show desktop” applets, and the placement of the power off button. Ubuntu 4.10, the Warty Warthog.
It’s great that there is fresh perspective coming to Gnome Shell design, but the first mockups are very strongly inspired by Unity. I don’t take offense at the imitation – as I said, that’s flattery. But I do find it frustrating to be told that our design work is not good enough to help shape Gnome, only then to have it imitated with the “parallel evolution” apology.
As for calls for help in Gnome design, it’s well known that Canonical has helped. MPT puts a ton of effort into various aspects of Gnome design. Charline does lots of research on Gnome applications which is published and shared. We designed the indicators for Unity after conversations with Red Hat developers where it was agreed that they needed work, and there was no clear direction from Gnome Shell. Only after the work was DONE and proposed as a Gnome external dependency did we get told that there was a new vision for indicators and our work wasn’t good enough. Of course, Shell design is being redone now, so perhaps the final version of the next-gen Shell indicators will look just like the work we’ve already done, and offered to Gnome, who knows.
You can imagine the frustration of a team that is actively working on code that they genuinely believe will be valuable for GNOME, which they care about deeply, only to be told “sorry, we see your code but we have a new design vision and we’re going to write completely different code for that”. Each of the chunks of work we’ve taken on: notifications, indicators, the menus, and all the rest, could be valuable to GNOME and has been done with the intent that it be useful to GNOME. Our feeling is that base politics are playing a bigger role in the final decisions than they should, and we’re disgusted that that be the case.
Mark
Hi Mark,
Fedora’s desktop doesn’t have a ‘Show desktop’ applet – not by default.
I’m not sure exactly what you’re referring to by the placement of the power off button? Do you mean this menu? http://www.dedoimedo.com/images/computers/ubuntu-8.10-shutdown-2.jpg As far as I know Fedora hasn’t ever had a menu that looked like that. Rather ours looks like this: http://www.labtestproject.com/files/screenshot/fedora10/039_gnome_shutdown_b.png
I don’t understand how the “first mockups” of GNOME Shell could be inspired by Unity when Unity was announced in 2010 and the first GNOME Shell mockups I’m aware of are from 2008 so I’m assuming a typo!
I see a single Canonical design blog post from Charline made on June 11 referring to a single usability report on Empathy basically consisting of the same slides I reported on in my own blog post on March 2. (http://design.canonical.com/the-research/) Are there more that I am missing, as you seemed to speak in the plural?
I can’t doubt there’s useful work occurring from Canonical’s end – and I don’t think Greg or anyone else really is either. The problem is if a tree falls in the forest and no one hears it…. as a designer in the FLOSS community it took me a long time to realize this. I would have had no idea about that empathy usability study if I hadn’t happened to be at the UX hackfest this past year. It is very much about right-place-right-time and speaking up – working studiously and quietly in a boudoir did work for Emily Dickinson except that much of her work was discovered and published after her passing. This is why I dropped out of a PhD program to work for Red Hat.
Where are folks saying your team’s design work isn’t good enough? I don’t know where it is happening. I’m on the usability@gnome.org list and have been for some time now, and I’m also in #gnome-design and I definitely haven’t seen that kind of back-and-forth going on. If there is a list or some other forum that I should be following to see this happening, please let me know. I can absolutely sympathize with the frustration that you’re expressing – I’m just not 100% sure that it’s quite so political / malicious as you seem to be interpreting it. *Especially* considering it involves design decisions. My first work in GNOME of note started out with branding because I didn’t have the stomach to work on UX. Arguing on specific design points when each designer has a completely separate overall vision they’re using to inform the specific design point – yet neither has really shared that overall vision – is really an exercise in futility. And I faced that with Red Hat co-workers, so I know they weren’t really a ‘tribal’ argument. Once I understood Seth Nickell’s higher-level vision for GNOME, for example, I began to understand why he was so harsh in his critiques of my mockups when I was his lowly intern. With his vision in mind I finally understood why the mockups were wrong. And I had similar kinds of arguments with Jon McCann when I first looked at GNOME Shell until I saw his talk at the GNOME Summit in 2008 and sat down with him and basically heard him out on his vision. Then all those design decisions he made that I vehemently contested made so much more sense.
So perhaps a better approach is to talk at the vision level first rather than at the specific design decision level. If the vision is agreed upon it’s much easier to have productive design decision discussions.
Certainly talking about where which icon is on the GNOME panel and which menu is here or there, and who left their fingerprint here or there first isn’t nearly as useful as talking about our respective visions for the desktop, seeing where they intersect as to be mutually beneficial?
How much of the public discussion about GNOME Shell inspired Canonical to build the current set of in-house technologies? The design goals of gnome shell have been talked about at GUADEC publicly for 3 years now have they not? How much of that _public_ discussion about GNOME Project roadmapping influenced Canonical in-house design work on osd-notify and libindicate and friends?
Credit where Credit is due…. influence goes both ways… but influence its not the same thing as active collaboration. Apple influences a lot of thinking about design..but they aren’t contributors to the projects they influence. Microsoft influences a lot that they don’t contribute to…should we thank them for that influence?
The Gnome Shell developers have been discussing things openly with the rest of the Gnome Project contributors for a long time and Canonical has unilaterally chosen to make parallel implementations in-house instead of being an active part of that GNOME project roadmap to make sure what was being built was what GNOME as a project wanted to integrate..and that it integrates the _right_ way.
It’s one thing to build toy implementations as proof-of-concepts that you are prepared to have reviewed and to make adjustments to or to even throw away as part of the process of collaboration. Its quite another to build an implementation in-house and then expect it to be the _right_ implementation that integrates into the GNOME roadmap on review from other peers collaborators.
Yes Unity and Gnome Shell are looking similar..but the reality is they are working towards the same design goals..goals set forth in the public dicussions about Gnome 3 and Gnome Shell roadmapping. But the implementation specifics are very different and will continue to diverge.
-jef
Jef, your timing is all wrong.
Canonical decided to take on notification, the existing project was moribund. We did design work, blogged about it, and implemented it. Thereafter, Gnome Shell designers said they wanted to look into notification too.
Canonical decided to look into messaging as a first class part of the desktop experience. We designed, blogged and built the messaging menu. Thereafter, Gnome Shell designers said that messaging was important, and approached us about working together. We said we’d be delighted to, we’d like to see a common API. But it appears that the Gnome Shell folks don’t want to use an API designed by Canonical, so there may be a whole new API for their messaging interface.
Another technology challenge we took on was the cross-desktop menu system. We decided we wanted a panel menu for Unity, and worked with the Global Menu project as well as KDE folks, based on a specification we helped write for FreeDesktop.org. Now I’m told that folks from a competing company are objecting to that work, too.
Jef, you’re an insightful guy. I’ve seen you take an independent view before. Look at the data, REALLY look at the data, talk to some of the Canonical developers who’ve been frustrated by what are seen to be arbitrary, political decisions and unexpected changes “upstream” that make collaboration very hard.
There are more than 100 developers at Canonical ALL of whom come from deep free software roots. Why would they want to do the wrong thing? They all want to see their code reach it’s most productive audience. But they feel substantially blocked in that pursuit, and this latest “data based critique” is the real cherry on the cake.
Mark
Feel free to point me to a publicly archived discussion about notification in the context of gnome that supports your argument about timing. A GUADEC video archive would suffice. A gnome mailinglist discussion. An fdo mailinglist discussion.
I’m not keen on relying on personal recollections. Memories are faulty… public archives of discussion are very useful. I love publicly archived discussion references. Please provide one as a starting point.
Mark, the reason you find upstream changes ‘unexpected’ is that you’re Doing It Rong. Look at your own words:
“We did design work, blogged about it, and implemented it.”
…all of which you did outside the existing GNOME channels. If someone at Red Hat decides they want to implement an improvement to something in GNOME, they don’t write a blog post about it, then write the code and stick it in some SCM outside of the GNOME project and in a Fedora package.
That person is almost certainly already an active part of the relevant GNOME group, so they post a proposal and a mock up to the group, to make sure it’s in line with other plans in the area, everyone is happy with how it’s going to be implemented, and so on. Then they start coding it, in the GNOME SCM, and asking for feedback on it on the appropriate GNOME list. Then, when it’s done, they get it committed into the main tree of the component. *Then* they blog about how awesome it is and add it to Fedora.
This is what we’re talking about when we’re talking about being truly involved with upstream. Coming up with all of your own ideas, coding in private, then throwing a finished product out for acceptance suddenly with no prior discussion isn’t the way things work in a collaborative development environment, and this is probably why you’re experiencing what you perceive as frustrating friction trying to get your contributions accepted.
That’s right folks, Mark $hufflew0rth invented:
– Codes of conduct (what would we have done before he invented them?)
– Cadence (There was no open source “flow” before Mark $hufflew0rth!)
– Gnome Shell (Ok, he doesn’t admit it, but as you can see it was all him)
He is on the money, all the open source was established so there was no need for Mark $hufflew0rth to give back, it was already done! He just made it all better with his awesome.
Canonical doesn’t need to give back, free software is free, remember!
Man,You are lame.
This is not necessary.
Mark has made contributions and has helped open source. The questions are about the methods, team building, and vision.
I would like to thank Mark for his important contributions, but I also would like him to listen to the broader community.
I don’t think anyone wants to fight. He is just asking for Canonical to provide more upstream developers to Gnome. The first time I saw Ubuntu (warty) I heard your former employee Jeff Waugh boasting that you had more financial resources than Redhat and I know that the Gnome community could do with some extra manpower that your resources could provide.
Lots of people use, love and appreciate Ubuntu but still feel that Canonical is not really working with them upstream. If you could provide some nice, hard working Gnome devs to quietly work on improving existing functionality upstream, that would do much to improve your relationship with others.
GNOME Census gave facts, you give religion speech.
Could you please give us facts that Canonical is a decent player for upstream projects ?
Contribution to kernel, GNU tools, Xorg, GNOME or KDE are accepted.
By the way, can I actually thank a lot Red Hat, Novell, Nokia, Mandriva and other volunteers for the work they have done?
Who really cares about how much company X does or does not contribute to the “upstream.” All I care about is a product that works and works without customization. Not who adds more to a project… And you know what??? Ubuntu has always worked for me out of the box, other distros haven’t worked that way
I care, and so should everyone else who cares about free software, and here’s why:
http://en.wikipedia.org/wiki/Free_rider_problem
Excellent point made. If the Free Software Community really works like it is “advertised”, I can’t see how it can continue to do that with such unhealthy competition. But it sure does not.
Cannonical is probably making the most out of GNOME and other free software, with the least contribution, it seems, but still has managed to make Ubuntu the most popular distribution.
When it comes to Desktop users, the most important thing is not moral or ethical codes. Most end users DO NOT CARE if the software is FREE or not, they only care if their Video driver (or anything else) works out of the box. Cannonical also does not seem to care about free software in the cost of usability. The only thing Cannonical seems to care about is “PASS IT ON!” and they are quite clear about that.
If a typical Ubuntu user was really concerned about freedom, s/he’d probably go with gNewSense rather than Ubuntu in the first place. Most who are with Ubuntu even don’t seem to care about why choose FREE software than non-FREE one.
I think Ubuntu is going the Mac way, which has proved to be immensely successful among its fans. It can be clearly seen in how Ubuntu is being led. And I think it is great as long as it works.
To sum it up, if Linux users cared only about “freedom”, Ubuntu would have failed long ago. It is not the case and it is all about ease of use plus “PASS-IT-ON”. Linuxmint does not seem to care about PASS-IT-ON, yet it is quite successful as it only focuses on usability.. if it is was backed by something as big as Cannonical, it would have been more successful than Ubuntu, IMHO. The same explains why Mac OS X is much more popular and has larger Desktop OS usage share than Ubuntu (or even all Desktop Linux combined).
I think Red Hat values Free Software more than usability for Desktop users (for which it gets criticisms often from users who care about usability and better out of the box experience) not because they really care about “user’s freedom” but because it is good for its business in long term.
At the end of the day, the only thing that matters is that companies that provide Operating Systems will make decisions based on what is good for their particular company. Users should also do the same and stop ranting about all other individuals and companies who do not agree with them.
That is just my personal opinion.
Pardon my poor English.
“From the layout of the Fedora desktop to the recently published “future look of Gnome Shell” which is quite clearly inspired by Unity”
Mark, you are now trying to take credit for layout of Fedora desktop and GNOME Shell design but can you can actually back up your statements with evidence? GNOME Shell is publicly developed and was available before Unity in a open manner. Not developed behind closed doors like Unity was and presented in a keynote by you as some sort of finished product. Explain how Fedora desktop layout was influenced by Canonical?
@Neo
The very first version of Ubuntu changed the placement of Gnome panels and applets in a particular way. That become a Gnome default, then a Fedora default.
Gnome Shell’s design was this: http://lh6.ggpht.com/_1QSDkzYY2vc/S_18HGpQP7I/AAAAAAAABG0/FNaNgFwIkp0/Screenshot-2.png
Then Unity came along, and it’s design looks like http://www.omgubuntu.co.uk/2010/07/unity-as-of-now.html
And several months later the Gnome Shell designers started producing mockups like this:
http://git.gnome.org/browse/gnome-shell-design/plain/mockups/static/overview-application-picker.png?id=0aab5349a31b83b8d99442bd1b7376df00520d8c
You might notice the location of the launcher, the existence of an “Applications” place, the subcategories, are all rather reminiscent of the Applications place in the Dash of Unity.
Now, I am happy to raise the bar in FLOSS any way we can. And imitation is the sincerest form of flattery. But I’d prefer to be working in a friendly environment, and not be criticised for doing independent design work only to have that very work imitated :-). And in addition, to have the implementation of our designs rejected by the gnome shell team on the grounds that the design isn’t good, after which the designs themselves are adopted, is a sign of bad leadership in GNOME.
Mark
I’m sorry you perceive these criticisms as “unfriendly,” Mark, but I can assure you that they’re given in the spirit of the Dutch uncle. And really, outside of the insular world of Linux distro development, nobody much cares anyway.
Does Ubuntu have lots of stuff that’s worthy of imitation? Sure. That’s not the point. The point is the continued evidence that, rather than participating directly in a broader community, Canonical chooses to go its own way with a pseudo-fork rather than engaging with others to solve hard problems.
That’s certainly your choice, and it’s a choice that’s fundamental to the free software world. But it doesn’t make you a good GNOME community citizen, no matter how you may claim otherwise.
I read you post again, which is all about good leadership in Ubuntu, and flowers, and let’s make love not war. Then I come here to read about bad leadership in GNOME. You will agree with me that’s not exactly the best way not to scare away potential contributors from this nasty environment, or to make the world a better place, but just another form of tribalism that will fuel this flamewar for days on end. Well done, man! This is a sign of good leadership, indeed…
@Mah!
So what are you adding to this discussion then? 😉
Mark,
Duffy answered it in detail already at
I am pretty sure GNOME UI is where it is today more thanks to Red Hat than Canonical. Are you thanking them and giving them enough credit while claiming some for yourself?
Where can you point me to the discussions about Unity design? Why was it not designed in the open from the start like GNOME Shell? Why did you create Unity instead of working with GNOME Shell? Why is Unity not hosed in GNOME infrastructure?
Jono Bakon gives some good points: http://www.jonobacon.org/2010/07/30/red-hat-canonical-and-gnome-contributions/
Basically, it seems that lot’s of the stuff that Canonical builds on top of GNOME, because they feel that it is important to their users, is not what GNOME thinks their users care about, so GNOME doesn’t even bother accepting them upstream.
However, many Ubuntu users that never have tried another distro, will perhaps think that GNOME is exactly the way it is distributed in Ubuntu, and get surprised when they discover a lot of cool Ubuntu features are actually GNOME features. Being a veteran Linux user (and former Fedora user), one of the main reasons I use Ubuntu is because of all the nice extra features they offer on the desktop.
Also since you can pretty much argue that Ubuntu is the distro that single-handedly pushed Linux-on-the-desktop to a whole new audience you can even state:
A whole lot of people wouldn’t even be using GNOME if it weren’t for Ubuntu.
(But they only way they ever got these people to use GNOME is by adding a whole lot of extra tweaks and features.)
The same could be said for Litl… and litl is ahead of Canonical in contributing back. Litl’s whole operation depends on being able to differentiate the UI and is levaraging GNOME as a set of technologies to do it. But they are giving back even though they are a tiny company.
Differentiation is not the problem. A lot of companies who rely on Gnome differentiate at somewhere in the UI. No one expects Canonical to ship completely stock Gnome. Noone expects them to not have inhouse projects. But the core Gnome technologies need care and feeding. Companies who want to leverage Gnome as a set of core technologies as a basis of differentiation need to be a part of that care and feeding. Canonical isn’t doing their fair share of that. Not by a long shot. You look at how much they have invested in differentiated work..and you look at how much they have invested in the care and feeding of Gnome as a set of core technologies..and its way out of whack with other companies on that list who are ahead of them in the raw commit counts.
Entities who want to build _on_ GNOME still need to contribute _to_ GNOME.
-jef
Perhaps you missed my reply to this effect on LWN jef…
Mentioning Litl as having lots of engineering time available for GNOME is quite ironic since their OS platform is delivered by… Canonical 😉
Some Ubuntu users have also been surprised to find out that many cool Ubuntu features are actually Debian features. 😉
Same thing repeated over and over in blog after blog. So what?
And many Debian users have no idea that some of their features come from Red Hat. Do you know where every package on your system originated? Of course you do not.
Being a beginner is not a crime.
Now let us compare a new user with moderate computer knowledge trying Linux for the first time.
They try Debian. They try and create a share so their media player can see their music collection. They have no hope. (In Fedora it also will not work, you have to know to configure the firewall first)
They try Ubuntu. Right click share, the system says you need Samba I will fix that for you. A minute later their media player is streaming music.
They try Debian on their notebook. Wifi does not work (firmware not installed by default). They try a 3G USB dongle and do a bit better but must manually end APNs etc.
They try Ubuntu on their notebook. Wifi does not work, but a dialog pops up and says you needs these drivers and would you like me to install then for you? Yes please, job done. They plug in their 3g dongle and the system pops up a list of pre-configured providers for their county for them to choose from.
I could go on and on.
I have been a GNU/Linux user since the 90s and have tried many many distros, and for plug and play simplicity it is hard to beat 32bit Ubuntu.
Just because you can hand write a samba.conf (and I can) does not make you a guru, it just makes you pigheaded when all you need is “right click share”
Usability is a very important feature. To grow the desktop share you need usability, and at the moment there is one distro trying harder than the rest.
It’s not the fact that “they” don’t generally actively participate in discussions with the various upstreams (GNOME or otherwise) before “they” release whatever they have done in one enormous chunk. Only then do “they” propose it to upstream, who can’t really comment due to the fact that they feel nothing they say will be listened to.
http://video.golem.de/desktop-applikationen/1876/mark-shuttleworth-interview-%28part-two%29.html
Here’s the salient and most damning claim that Shuttleworth has made:
As percentage of employee headcount Canonical has more developers working on free software than any other corporate entity.
And yet that doesn’t show up anywhere in the shared technology stack that is common among these corporate entities. There is a disconnect. We keep expanding the effort to measure where corporate contribution looks like in the stack and every where we look Canonical isn’t a strong participant.
If every corporate entity worked the way Canonical worked..and focused myopically on the software it owns the copyrights to built and managed on infrastructure it controls.. we wouldn’t have an open ecosystem at all. That’s the problem. Corporate controlled walled-gardens are not the right collaborative environments to build an expansive ecosystem of shared technology.
I agree with this completely. I guess things haven’t changed much:
http://www.kroah.com/log/linux/lpc_2008_keynote.html
“There are relatively very few people who actually give back in source code _or_ in bug reports, so anybody who argues against free-loading in open source is a moron. ” – Linus Torvalds
Canonical is not a end user. They are a prominent distro. They can give back
No.. Canonical is not a distro. Canonical is a for-profit company which managed a distribution as part of a larger business strategy.
-jef
Well, a prominent company backing a prominent distro. It’s been noted already, I believe, that considerably more has been given back to GNOME and actually accepted upstream by Ubuntu than by Canonical (or something to that effect).
On a different note, I’m quite interested to notice that when someone points out the work Canonical has done on things that don’t exist in GNOME Shell, it’s discounted because it doesn’t work in GNOME Shell – and nobody seems to be taking into account that precisely because GNOME Shell is a very different way of doing things, not everyone is going to use it. If that results in a fork of GNOME, whoever it originates from, in order to keep a version alive that someone might prefer, then so be it.
I quite see the point about giving back, although I also support the idea that it’s not for anyone outside an organisation to say what that organisation should be developing. Getting what /is/ produced back upstream is not the same as doing work on the core of the project. I can certainly understand the point of view that says that what Canonical makes should at least be submitted for upstream inclusion. I also think that if anyone’s going to use data, it should be on what contributions are offered rather than which ones are accepted. What I don’t agree with is the idea that because Canonical back a distro that relies heavily on GNOME, they should be developing the core of GNOME.
I would like to close by pointing out that I’m not actively supporting one side or the other here – please don’t assume that I’m bashing any group that anyone sees as a “side”.
Mairin
It’s not the first Gnome Shell mockups I’m referring to, it’s the latest ones published in the past two weeks, *after* Unity’s dash landed and was widely covered.
Here’s Unity today:
http://www.omgubuntu.co.uk/2010/07/unity-as-of-now.html
Here’s the new Shell mockup:
http://git.gnome.org/browse/gnome-shell-design/plain/mockups/static/overview-application-picker.png?id=0aab5349a31b83b8d99442bd1b7376df00520d8c
Mark
Hi Mark,
To be honest, these look different enough that I’m really not sure what I’m meant to conclude here. The Activities item has been in the upper left corner of gnome shell all this time. The search box *was* in the upper left in the old mockups, it is now on the upper right; unity’s is in the upper left, where gnome shell’s used to be. The status icons are along the top panel area right-aligned as they have been in GNOME at least since 2004; this is also where OS X places those icons. You could maybe claim Ubuntu was the first to use single-color icons, but that would be silly since OS X has been doing this since 2001 at least. The application menu appears to be in the same place it has been in gnome shell for a long time; it’s just that the labels and arrow now come from the top of the application well rather than the left. The only similarity I can really see from a design POV is that there are application launchers along the left side of the screen. By the way, is the layout in the unity mockup you linked to persistent or is it available as an overlay? Because the mockup that you showed of GNOME shell I believe is the overlay view, *not* the desktop view. This is kind of a finer point that isn’t quite visible in flat / non-animated mockups.
This much ruckus over icons being along the left side of the screen seems a bit weak, Mark.
Also, fwiw when I first saw unity mockups/screenshots I thought (and I heard at least a couple of other people say this independently in IRC) the icons along the left looked the AfterStep window manager (which I used to use as a teenager): http://www.tuxradar.com/content/archives-best-window-managers-2000 They are aligned along the right there, though. I vaguely remember very similar tiles aligned along the left side of the screen on the SGI and Solaris workstations at my university, but I don’t know if that was a custom launcher bar written by the university or part of those window managers as well.
BTW one thing that really stands out to me in those mockups is that one is hosted at GNOME.org and the other is not……
Since when is Canonical an open source company?
If you want to compare Canonical with something, compare it to Google.
Since when is Canonical an open source company?
since it was born 😉
[…] DeKoenigsber: “It’s not about tribalism, Mark.” It’s about accepting responsibility for your place in the world. For unto whomsoever much is […]
The census is a report that by its own admission has a narrow focus, makes non-optimal choices for expediency, and is not even normalized to time.
If somebody is so indignant and upset over this data then that person is either ignorant of statistical inference or just happy that something can be used to justify that person’s pre-conceived decisions.
You can try to correct ignorance, but as Mark points out, when the person is unwilling to see what can one do. Any further discussion is futile.
The census is a report that by its own admission has a narrow focus, makes non-optimal choices for expediency, and is not even normalized to time.
If somebody is so indignant and upset over this data then that person is either ignorant of statistical inference or just happy that something can be used to justify that person’s pre-conceived decisions.
When the person is unwilling to see, any discussion is futile.
Tribalism?
I don’t care for this rather philosophical stuff, I only care for the facts.
1. Canonical has only 350 workers compared with Red Hat’s 2800. Since RedHat contributed 16,3% of Gnome’s code, Canonical would be on a level with RedHat’s efforts, if they “only” had contributed around 2,04% of Gnome’s upstream code. Canonical contributed 1,03% of Gnome’s code, at first sight this is too low.
2. If you look at the data of the Gnome Census, you see Eazel ahead of Canonical with 1,09% of code contributions. But Eazel went bankrupt in 2001, three years before Ubuntu was “born”. So it is obvious that the Gnome census covers the whole timeframe of 13 years of Gnome’s existence, but Canonical couldn’t contribute seven years of this timeframe, because it simply didn’t exist. In my opinion the Canonical amount of contributions would be significantly higher if the Gnome census would focus on the timeframe from 2004 to 2010, because Canoncial was founded in 2004.
3. You see also Litl in front of Canonical with 1.06% of code contributions. But Litl’s OS is Ubuntu-based. So you should notice that Ubuntu contributions and Canonical contributions are not the same, this fact is completely ignored in this report. (To be honest, I don’t know how deep Litl’s Gnome contributions are involved with Canonical or the Launchpad platform, respectively.)
4. Therefore, Canonical could have reached the amount of Gnome code contributions of around 2% which would bring them on the almost same contribution level with RedHat even now, if only the number of employees would count.
I can only say one thing to all of you…
In different fields, either Canonical and Red Hat have given too much value to FOSS in general. While Red Hat (not forgetting Novell aswell) have placed a lot on court fights against the heathens, Canonical increased the notoriety of linux quite a lot.
The Paladins of FOSS have always been blind… While some neglect the powers of Marketing (I mean real Marketing), I’m pretty sure Canonical will get some proffit in the future for their investments on end users.
While Microsoft screams in agony claiming that students are being recruited by FOSS projects giving them no man power (probably Google playing a great role here), some start to see a success on Canonical’s strategy and now they charge blindly against it.
I’m not a Ubuntu fanboy, neither I use Ubuntu, but all the mambo jambo I’ve learned on Marketing during my degree (which finally is over) can pretty much justify much of Canonical’s actions. Thank God at least one Corporation inside of Free Software realm dared to play real marketing. If others were too blind to neglect it, then it’s your own problem.
As for contributions… Come on… I’ve seen the GNOME Project advertising funding from Canonical to hire a sys admin for their infra-structure… I’ve seen on GNOME mailing lists many @canonical playing active roles…
As for gnome-shell…. It’s a nice motivation to get me back to WindowMaker or KDE if the gnome-panel is removed in the future. Maybe I’ll start writing my own stuff in PERL as an alternative in my free time…
As for the attack on Fedora Design… as I see it there have been efforts in the past and the Fedora Design Team is very committed. It’s not their fault that some cool initiatives from them have been tackled by Red Hat’s Desktop Team. I know some can’t stand up because they are employees of Red Hat also, but I know they know what I mean. So don’t attack Fedora Design… Thats cheap and vile from you.
As for the future… WE ALL are only bringing shame to the game… think on that.
For the rest, a small quote:
“It is more important to do what is strategically correct then what is immediately profitable” – Philip Kotler.
Think also on that quote, which I believe it does translate the view of Canonical (under my shy knowledge about it) and it’s something Red Hat failed to understand.
So if Canonical blows up Red Hat in notoriety in audiences that with a 5 to 10 year time frame can become major contributors, it’s only Red Hat’s fault that such a thing happened. The only question is “will canonical survive for that long to reap what they sowed?” – Lets hope so, despite of the number of contributions they give and in which way they provide them (being manpower, code or $STASH).
For me… I’ll stay with OSX since none of you open source people really cares about the segment of users in which I qualify for! So for us, non-IT guru’s, we always have commercial and proprietary which cares about us (even if it’s by our own money).
And… scene.
[…] apparently the recent GNOME survey aggravated some old wounds.1 Greg DeKoenigsberg of Red Hat fame lashed out at Canonical, who had contributed surprisingly little to GNOME, and Mark Shuttleworth (head of Canonical) […]