FULL DISCLOSURE: I’m an active member of the Fedora Project community, and I also happen to work at Red Hat, Inc. (not in a code development engineering role).
No. Not that F word. In an open source project, especially one with large corporate backing, there’s an even worse utterance out there…
Fork.
Is the Ubuntu project setting itself up for a fork down the road?
Obviously if I could tell the future I would be doing much more productive things than sorting through Canonical’s messy handling of their Linux distribution. Winning on the ponies at Aqueduct jumps to mind. At any rate, I have no idea if the SABDFL will hold his ship together or not, but there is evidence to suggest that it could easily be heading for a messy fork in the river.
What is Upstream? What is Downstream?
This has always confused me, but where does “derived” fit in the standard model of how an open source project lives. It’s derived from Debian, but what is the actual relationship? Is it upstream from Debian, Downstream? Were they just standing in line together at the DMV one Saturday? It’s a small issue, but it nags at me to no end.
The “Community” Organization
Back in the days of Dapper Drake, it was funny to refer to Mark Shuttleworth as the “Self-Appointed Benevolent Dictator For Life”. It was cute. It was quirky. It was sort of orangey-brown. Just like Ubuntu. But Ubuntu was about togetherness and community, and it was the coolest Linux distro out there at the time. Upstream drivers and a 6-month release cycle and holy crap, it supports my video card!
Ubuntu’s dictator also has the large task of staffing Ubuntu’s Community and Technology governing entities.
The Community Council’s charter is to:
The social structures and community processes of Ubuntu are supervised by the Ubuntu Community Council. It is the Community Council that approves the creation of a new team or project, along with team leader appointments. The council is also responsible for the Code of Conductand tasked with ensuring that community members follow its guidelines.
The Technology Board is responsible for:
The Ubuntu Technical Board is responsible for the technical direction that Ubuntu takes. It makes decisions on package selection, packaging policy, installation systems and processes, kernel, X server, library versions and dependencies. The board works with the relevant team to try to establish a consensus on the right direction to take.
Fast forward to 2012 and Canonical is tryin to monetize Ubuntu, squeezing it into anything that someone has the guts to ask them about. Sadly squeezing into places where Linux itself has been for ages is the most common use case I’ve been able to find (Ubuntu TV and Ubuntu in your car). You also have 8 years of Mark Shuttleworth picking the people on and direction of the two major governing bodies within Ubuntu itself. Fun examples of this attempt to monetize Ubuntu can be found in the latest releases Amazon “integration” (shame on you, Amazon) and also in this bug, talking about searches run from the Unity dash.
Unity
I’ve used Unity for a grand total of 8 minutes. But I know that one of two scenarios about it is true:
1. the minority of unsatisfied Unity users is exponentially more vocal than the satisfied majority
OR
2. there are a LOT of Ubuntu users out there that are NOT HAPPY WITH UNITY.
A search of Mark Shuttleworth’s blog for “unity” shows one early conceit that the original versions “sucked”, but that they are now “well positioned”. Whatever the community wants, it looks like Unity isn’t going anwwhere except into Ubuntu.
Pulling Bits out of the Community’s Hands
I’ve seen spin on this article and the blog post that caused it to be written, but I can only read it one way that makes sense to me.
Canonical (read Mark Shuttleworth) believes that a small group of people in a closed environment can do better work than a large community.
I’ve “done” FLOSS for a while now, and if there is a single immutable truth it is this: Open Source is noisy and often messy. But it’s the noise and mess where you find the genius that changes the world. The off-hand idea in a minor listserv. The idea floated as “impossible” or “impractical”, just like the GCC was back in the day.
I don’t care if he releases all of the code under the GPL after he makes it. When he decided to create skunkworks teams of his hand-selected people, Ubuntu stopped being a community project. And maybe it never was. I’m not saying that Ubuntu is invalidated as a product because it’s not community-driven. I’m just saying stop talking the talk. One thing I do know is that if happens within Fedora (a little response to this), it happens out in the light of day, on a mailing list or in IRC or in the web tools. If I were a contributor to the Ubuntu project I’d be seriously thinking about offering up my time and talents.
Lost In Open Source 
“Amazon “integration” (shame on you, Amazon)”
The Amazon ‘integration’ doesn’t have much to do with Amazon itself. Ubuntu’s just using the general Amazon affiliate program anyone can sign up for. If you click the links Unity gives you and buy something, Canonical gets the affiliate fee. As far as anyone outside Canonical knows, Amazon don’t have any direct involvement in this thing at all, Canonical are just using the affiliate program as thousands of others do.
“One thing I do know is that if happens within Fedora (a little response to this), it happens out in the light of day, on a mailing list or in IRC or in the web tools.”
I’d honestly be careful about saying this.
I don’t think it’s that unusual for F/OSS projects to be started in a skunkworks fashion: one or two folks just writing code and sharing it between them. There’s certainly stuff in Fedora now which started out with one person – sometimes, one person in a Red Hat office – writing code and not sharing it with anyone else. At some point they decide there’s benefit to showing the code to the public and they do so. But I don’t think there’s many F/OSS projects which have been developed in public right from the first time someone opened vi and typed ‘#include’. It’s not good for Red Hat, Canonical or anyone else to go too far in bashing this approach. Heck, Will Woods has been working on what will become Fedora’s sole official upgrade tool for several weeks without doing any public releases or really much public discussion: he’s been pushing the code to a public git repo for a while, but that’s all.
What I think we can say is that the Fedora feature process is pretty well followed by all Fedora contributors including those working for RH, and this means that most things that become major Fedora features are being worked on as proper public, community-based projects for at least several months before they show up in a Fedora release. And that’s a good thing. But I don’t think it’s a good idea for us (Fedora or RH) to over-claim in this area or to get into ‘we’re better than you’ bunfighting with Ubuntu or anyone else. Just accurately describe our positive attributes and trust people to figure things out for themselves.
I agree that most projects start “in somebody’s garage” and come out and evolve into a true project over time.
That, IMHO, is in now way what Shuttleworth is doing. He’s taking large parts of a huge, existing open-source project and intentionally closing portions of the ideation and development to a small group of hand-picked people.
I don’t read it as that. The big Ubuntu features that come out of Canonical have always been developed internally at Canonical in ‘skunkworks’ style for some time before they are given a big public splash announcement; check your history, this is how Canonical has done things for some time. The initiative Shuttleworth is announcing is to have some limited form of community involvement in this _existing_ process. Whether you think it’s a good process or not it’s true to say that it’s how things have been done in Ubuntuland for a while, and the specific move Mark recently announced is a (limited) move towards _more_ community involvement, not less.
“it’s how things have been done in Ubuntuland for a while, and the specific move Mark recently announced is a (limited) move towards _more_ community involvement, not less.”
And that may well be the case. It doesn’t mean that the community is or should be happy about it. That messiness and noise that he’s avoiding is where, IMHO, the true genius of a community-based project.
It amazes me that this is how that group of people function. This was a great and glaring example of that activity.
I don’t think Ubuntu has ever been considered ‘community-based’ to mean exactly what it means to RH / Fedora people, to be honest. I may be misrepresenting things, but to me it’s much more the model some more ‘enlightened’ large proprietary firms are using today of ‘active response to the concerns of a user community’. When you talk about ‘community’ in the Fedora model you’re talking more about people who are actively involved in the running of the project itself. They’re different models, and you can get into a bit of a mess if you just use the term ‘community’ uncritically and try to compare the two. You could certainly argue, for instance, that Ubuntu is more responsive to the needs of its *user* community, but Fedora is better at building a *contributor* community…
Ubuntu is heading for a lot worse than a fork. They seem to have progressively adopted an almost Apple-esq attitude of, “if you don’t like our products, F*You and go somewhere else”
It’s a shame because I like ubuntu, it was my first linux experience back on 8.04 but there’s been too many changes happening too soon, and without much community feedback. The project tends to announce something big, the community and wider media reel back at the drastic move, then it spends the next few weeks re-blogging and smoothing things over.
It’d be so much easier if they blogged something along the lines of, “Hey we’re trying this idea, do you like it, yes/no?” rather than blindly charging ahead.
Oh and Unity is better then Gnome 3 (not difficult) but it still lags and crashes on older hardware. Why on Earth remove the 2D option?
I wouldn’t exactly hold Red Hat / Fedora up as a shining example either, better yes, but not shining. Red Hat directs how their employees work on Linux, Fedora and Gnome for their own commercial benefits, I see very little done by Red Hat that benefits the wider community without first being a benefit for them, and plenty where the wider community is disadvantaged. I’ve also seen Red Hat ‘submarine’ key projects as well. Red Hat holds a powerful position where they can dictate large chunks of the core Linux infrastructure and everyone else has to follow their lead, yet Red Hat still closely controls these projects even though they are open. Sometimes that works to everyone’s benefit, sometimes it doesn’t and sections of the community that Red Hat has little financial interest in suffer as a result.
“Red Hat directs how their employees work on Linux, Fedora and Gnome for their own commercial benefits, I see very little done by Red Hat that benefits the wider community without first being a benefit for them, and plenty where the wider community is disadvantaged.”
Would you care to cite some?
How is having, oh, a good half dozen people working fulltime on developing a desktop environment we don’t control and don’t really derive much revenue from (GNOME) benefiting us? How is developing sound drivers benefiting us (remember our usual revenue-generating deployments are on servers)? How is developing 3D accelerated video drivers benefiting us? What exactly is it we’re ‘submarining’ into anything to our own benefit? And please, please don’t say PulseAudio or systemd, because you’ll be ripped to shreds.
FWIW, I too tried Unity. I’m not vocal about it, but I’m using Mint at home, Red Hat/CentOS at work, and happier with both of them.