April 11, 2006

Interview with Debian Project Leader-elect Anthony Towns

Author: Joe 'Zonker' Brockmeier

The results are in, and Debian Developer Anthony Towns will take office as the Debian Project Leader (DPL) on April 17, 2006. Towns, also the secretary of Linux Australia, took the time to answer a few questions from NewsForge via email about his plans for the next year.

The DPL serves a one-year term, and has the power to appoint delegates to the Debian Technical Committee, propose general resolutions and amendments, make decisions with Software in the Public Interest (SPI) regarding property held for Debian, and a number of other powers and responsibilities. On with the questions.

NewsForge: What are your immediate plans as DPL? What do you plan to do over the next few months?

Anthony Towns: The immediate plan is to organise the various ideas I've had so that I can work out which ones are actually worth working on, and what order to do them in; and to make sure that all the people who volunteered to be DPL during the campaign, or offered their help don't go away without some good ideas about extra things they can do.

At the moment, the next few weeks are enough of a worry, let alone the next few months -- Andy Barth and Martin Zobel who recently took over from Joey Schulze are in the final stages of preparing the next point release of Debian 3.1 (sarge); AMD64 is almost ready to be included into etch (the codename for the forthcoming release, due in December); I've got a whole lot of things to become familiar with in the DPL role; and by this time in May I'll be on my way to Mexico for the annual Debian Conference.

NF: The Debian new maintainer process is often criticized for taking too long and discouraging potential developers. You didn't agree with Ted's proposal for requiring maintainers to re-up periodically -- do you have any ideas for speeding up the process or do you think it's okay as-is?

AT: I've been talking with a few people for most of the year about an idea that I think will help with this, which is essentially to introduce a restricted level of developership, that's both very easy and quick to achieve, but also has far more fine-grained restrictions on what can be done.

That's been suggested before, and in the past I've been against it -- it's really nice to say that every developer has pretty much the same abilities as everyone else; and it's also pretty nice to have any developer be able to dive right in and help out if they see a problem. But you can only do that if you have a high amount of trust in all the people you let be developers -- and since the process of establishing that trust is taking too long, I've come around to the view that we're better off finding ways of doing without it.

I've posted some comments to my blog about this today, and I expect that there'll be some interesting discussions following on from it over the next few weeks -- hopefully culminating in some real changes.

NF: For those who aren't familiar with Debian, what areas of the project are you active in or have been active in since becoming involved with Debian? Do you think being DPL will take time away from your current roles within Debian?

AT: I've been involved in a few areas; helping develop and maintain the bug tracking system we use, and the archive maintenance software. I wrote "debootstrap" which is the low level tool used to install Debian systems, and "ifupdown" which is used to configure network interfaces. I was also release manager for Debian 2.2 (potato) and 3.0 (woody).

NF: How "relevant" do you think Debian is these days? A number of people have expressed the opinion, for example, that Ubuntu might be making Debian obsolete -- I'm going to assume you don't share that opinion, but how important do you think Debian is now?

AT: You can't say how "relevant" Debian is without knowing who you're talking about. Someone who uses Ubuntu every day and loves it might think Debian has no relevance whatsoever, but most of the people who actually make Ubuntu are passionate about Debian as well -- e.g., Mark Shuttleworth thinks Debian is relevant enough to have reactivated his account and voted in this election.

Rather than making Debian obsolete, I personally think Ubuntu has done an excellent job of bringing Debian's technologies and principles to more people, and while the relationship between the two projects is... complicated, I think it's still a healthy one and will only become more so.

NF: What do you see as Debian's biggest challenges for the next year?

AT: One major challenge is ensuring that we find ways to allow all the people who want to contribute something to Debian to do so -- the time it takes to get through our new-maintainer process is one problem we have in that aspect, but it can also be hard just getting any idea where your help is actually wanted; and in a volunteer project like Debian, you need to make sure you harness all the help you possibly can.

NF: On the flip side, what are Debian's advantages as a free software project? What does Debian have going for it that other projects may not?

AT: By being a purely volunteer project, that traditionally doesn't rely on a strong leadership, we have a huge amount of diversity, whether you look at our developers and their skills and experience, or the software we distribute, or the systems it works on, or the uses its put to -- if someone sees a need that Debian can fill, and are willing to put in the work, they can simply join up and then they don't need to ask anyone's permission to have Debian fill that need.

NF: Do you think Etch will be released on your watch?

AT: Yes; the Release Managers, Andy Barth and Steve Langasek have a lot of skill and experience, as well as a good plan for getting etch out the door, and importantly also have support from both the installer team, who have already made a couple of beta releases of etch, and the security team which has a subgroup that continually focusses on the forthcoming release -- which should resolve the two main problem areas which have caused delays in previous releases. I believe the developer body has a lot of confidence in Steve and Andy, and I expect that in the next few months we'll start seeing some fairly significant activities to make sure their release plans hold up.

NF: I noticed that fewer than half of eligible Debian developers voted in this election -- any ideas why that is?

AT: One issue is that a number of eligible developers are simply not actively involved in the project; it's been some time since the last time we've sent emails out to developers who haven't been active in some time to see if they're still interested in being involved, so it's hard to say how many that is.

NF: You were asked about summing up your platform in three words -- you chose "vitality, recruiting, direction" -- could you expand on that for readers?

AT: Debian's a great project -- but we can get caught up in arguing about things, or overthinking them, or simply not finishing up some of the plans we make. By "vitality", I mean making sure we actually *do* things, rather than just talk about them.

By "recruiting" I mean making sure we have new people participating, whether that be entirely new people just starting to get involved in Debian, or people who've been involved in one part of Debian starting to get involved in other parts.

And by "direction", I mean making sure we're not just letting ourselves keep going in the same direction we have been forever, but thinking about new things we can do that benefit our users and make our software better.

And then there's always my entire platform.

If you think of Debian as a jogger, then vitality is our heartbeat, which we want strong and regular, recruiting our lungs, to ensure we've got a good flow of oxygen and we don't try to operate on stale air, and direction is our head, making sure we're constantly thinking about what we're doing so we don't trip on a pothole or run in front of a car.

NF: Thanks for taking the time to answer our questions, and good luck in the new role as Debian Project Leader!


  • News
Click Here!