Koala will be 'a definitive shift' for Ubuntu Linux

Ubuntu founder Mark Shuttleworth

Mark Shuttleworth

You may, by now, have heard of a little distribution called Ubuntu, unless you happen to have been in space for the past five years.

For now all you need to know is that Ubuntu is a popular offshoot of Debian, and the man behind it, Mark Shuttleworth, is considered to be a veritable Linux titan.

Mark is the dotcom millionaire founder of Canonical, first patron of KDE and the self-appointed benevolent dictator for life (BDFL) of Ubuntu. Linux Format magazine risked being crushed by his (rumoured) hordes of robotic minions to ask him a few questions.

Linux Format: Do you have big goals for the next 10 releases? You're dictator for life – is it really life, or like the British penal system where you get off after six years?

Mark Shuttleworth: Well, I hope it's for life in the sense that I love technology, and I love economics and I love what's going on in society. For me, Ubuntu brings those three things together in a unique way. I hope it's for life, because if we pull this off then whatever is particularly interesting at any one time, there'll be packages within Ubuntu.

Right now, I'm fascinated by how we figure out how to socialise concepts of design and user experience in the whole open source community. In five years time, in 10 releases time, the focus could be on something completely different.

LXF: The GNU Hurd?

MS: (Laughs) Its time may have come! But you know there will be packages of it in Ubuntu, and I think that's awesome.

LXF: We're looking forward to Koala. You've said that 'boot will be beautiful' – care to explain more?

MS: There's some extraordinary work that's been done [on Koala], mostly pioneered by the Intel/Moblin team, the X team, and the kernel team (kernel mode setting), so I think that's going to be a definitive shift for us. I'm really hopeful we get that in.

LXF: Are you thinking of Plymouth or something else entirely?

MS: Well, Plymouth is one option. So, kernel mode setting is the foundational layer, then on top of that you need something that's a framework for what you might draw before X comes up. Plymouth exists, that's great. We've used USplash in Ubuntu and that's great too, we're kinda the first to do that.

So, either USplash on a kernel mode setting frame buffer, or just using Plymouth. And Plymouth is great technology, so I've no objections to embracing that, although I don't think any decision has been taken at that level.

LXF: One of the most talked about things so far is the retention of 'no brown'. Are you thinking greens?

MS: No. What I'm thinking is that we've built a team inside Canonical that's really passionate about design and about how you make software more open, more usable, by focusing people's attention on the right things at the right time. We have the capacity to do a great job.

Brown has worked really well for us. It's distinctive, warm, human and friendly. But it's also challenging in a bunch of different ways. At this stage, there's no concrete decision as to what the next sort of theme will be. But we're starting to have the capacity where I'm confident to say, "Okay we can do something else now that might last five years." But I can't pre-empt that, I don't know what it is.

LXF: Presumably, you're not going back to the controversial semi-naked people approach?

MS: I'm still a fan of semi-naked people you know. It seems like a good first step!

LXF: We talked to Steve McIntyre recently (Debian's Project Leader) and he says he has a great relationship with the guys at Canonical. Do you think there's more work to do to streamline the flow from upstream?

MS: Working with Steve has made a big difference. Debian is a unique organisation and even though the role of DPL is one of many hats, [his] leadership makes a big difference. Steve has got a constructive view of how we can work together, and that's helped a great deal.

It's very important to me that not only do we have a good relationship with Debian, but that Debian feels great about what we're doing. In a real sense, Debian is the epitome of free software collaboration and community, and Ubuntu has never been an attempt to detract from that. Some of Ubuntu's detractors would like to make that case [against us], but that's certainly not true.

This is simply the most efficient way that we can move Debian forward. I believed in the beginning, and I still believe, that we couldn't do all of things that we're passionate about inside Debian, just because its nature is such that it considers every possible angle.

We're really focused and passionate about a few particular angles. But yeah, that relationship keeps getting better because increasingly the people who participate in Debian are people who are comfortable with Ubuntu's existence.

LXF: Elsewhere upstream, Greg Kroah-Hartman gave an interesting speech at the Kernel Summit 2008. He said there were 100 kernel patches in the last three years from Canonical, putting you in 79th position…

MS: Well, clearly, his stats are wrong and he's accepted that. But even the thrust of what he's saying is incorrect. Canonical has never believed that what the world needs is more Linux kernel developers. Greg goes on to say in that speech how amazing it is that there are tens of thousands of kernel developers and how we have more patches coming in and line changes than Microsoft possibly can. But I think there's a question about if the world really needs more of that.

What the free software community really needs is a supereffective conduit to get all of that goodness to end users. It's also a fallacy, I think, to say that measuring kernel contributions is the only way to measure contributions. Andrew Tridgell, from Samba, said it best when he said:

'Hold on a sec. I've been working as a developer for the last three years running Ubuntu. I've never contributed a patch to Ubuntu, but Ubuntu makes me hugely productive. I get thousands of security updates. They never break my system. This is everything I need. I get 10 hours a week extra because Ubuntu exists. That's an enormous contribution to me, and hence, an enormous contribution to Samba.'

And that's sort of up in the fluffy contribution side. If you look at the number of patches, the number of lines of code that get written within Canonical to make the free software stack better, I thought Greg's arguments were pretty thin.


Article continues below