Ubuntu Down Under

I went to Ubuntu Down Under more to experience these mysterious events that Andrew flies off to every four months more than anything else. I wasn’t entirely sure what I expected from the day, but I wouldn’t recommend them to people who do not have a pretty developed interest in Ubuntu or something else Canonical is working on.

Yesterday, the ANZAC public holiday in Australia, was conceived as “Ubuntu Down Under Love Day” originally. I’m not quite sure whether I picked up the wrong message or whether I picked up the right message and the concept didn’t work out. The message I got was that it was going to be the most outsider-friendly day, and a chance to reach out to Sydney Linux users and Free Software developers.

What this amounted to, aside from SLUG’s small installfest, was that the Monday (like every other day of the conference) is open to drop-ins. It didn’t mean though that there were many sessions that had much need or interest for people not already, at the very least, following the Ubuntu development mailing lists. This conference is one week long, apparently because the two week conferences were burning people out too much to work during the second week, but the way that’s been achieved is to roughly double the output expected from the single week. The schedule, which is now sufficiently complicated that making it must be approaching the complexity level of the high school timetabling problem, was completely changed the night before the conference, leaving me with a pretty unfamiliar set of sessions I could attend.

And at a really basic level, the mindset just wasn’t about “let’s take outsiders and draw them in.” Very few BoFs started with an overview of what the session was set to accomplish and most were limited to conversations between two or three people while others listened in. All of them have a very formal aim, which is an edited write up including a set of goals for implementation.

I sat in on a couple of Launchpad sessions as a kind of observer with a professional interest in software engineering techniques, but it was pretty opaque to people not following development (the code is in-house and completely proprietary, so that essentially means Canonical employees).

Carlos was interested in talking to me as a representative of the documentation people, but unfortunately I don’t know the translation work-flow at all.

The Membership and Maintainership BoF Mako ran was the most interesting I went to, partly because the topic is community participation, which is easy to understand, and partly because the goal is moderately interesting. The idea is to separate the two things that Debian combines in the New Maintainer process: that is, power to help make decisions (by voting, usually), and the power to upload software into the distribution. In the Ubuntu community, the idea is that people shouldn’t have to jump through all the ‘can we trust your software skills?’ hoops if their contributions are, for example, helping out on the mailing list. However, they should be able to vote. Hence, the Member/Maintainer distinction.

I would have had a better day if a few more documentation team members had been there I think. At the moment I’m trying to do some facilitation for them (facilitation is relatively newly discovered talent of mine, and since I prefer editing technical documentation to writing it, I think it’s the main thing I’m going to offer those docs) and there are going to be some very pressing issues shortly surrounding two things: communication with the development team; and whether contributions should be tightened up and funnelled through the doc team (which would mean, among other things, bringing the wiki under tighter control). These could have used some nutting out with the developers in person.

As it was, I didn’t get a lot out of it for the most part. It’s not a marvellous ‘networking’ (read: meeting cool people) event, because the Canonical people are working 14–20 hour days, and the meals are in-house and expensive (breakfast and lunch were each $22 buffets, I don’t know how much dinner was). I doubt I would have met anyone much if it wasn’t for knowing several of the Canonical employees already. (It would be possible to, but you’d have to be a bit more outgoing than me, and also confident of making a good impression on very busy people.) I’m not familiar enough with distribution development to help write specifications for that; ditto Launchpad (and I gather they don’t want outside help anyway); and I’m way behind the eight-ball on distributed version control.

Assuming that the conference pattern follows this one, I wouldn’t recommend people without a really pressing need to get involved or existing involvement go along to these things in future. Assuming that this is a problem (it need not be, not all conferences are suitable for casual observers) I think the solution would have been to really pull SLUG into it and push a lot of the reach-out work onto SLUG. I would have enjoyed, say, a codefest with some token Canonical participation or some ‘donated’ talks from them, more than I enjoyed their development conference.