Marc's blog

Even though Adrian von Bidder suggests that I'm starting a bit early with the DPL campaigning, I thought I could add a few more thoughts to the blagosphere.

Adrian suggested to make the discussion of our of "vision of Debian" a regular event on debconf - I think it's a nice idea, but that doesn't mean we should keep silent about that topic the rest of the year. I think one of the problems we see in Debian is that we are missing an aim to strive for. We are currently just moving from release to release, while more and more developers have actually lost interest in stable releases. The impression that a common ground like Debian 4.0 is irrelevant seems to prevail, because the people most vocal in our user base usually want to see the newest and shiniest software on their desktop. We have been spending more and more effort to cater to their wishes and have ignored the fact that a majority of Debian-based installations are not run by such people, but by system administrators who do mass roll-outs and neither want nor need the newest X.org with more *bling*.

Ubuntu, the biggest Debian-based distribution, has found its users by packing up new software behind a nice installer, adding some branding elements and then shipping it. Stability is an issue for Ubuntu, but they don't spend as much time as we do on eliminating the last seemingly important bug in some esoteric software package before actually releasing. That observation is the ground for one of my ideas: Our current metric to determine how far testing is away from a release is counting the number of rc bugs - it doesn't matter if the rc bug is X.org crashing for all people or if it is a policy violation in a package with 12 users. Perhaps we should start thinking about a new metric, keeping in mind that for a majority of users, some bugs matter a lot and some matter not at all. With the release of etch, the number of installations of popcon has risen a lot, so for the first time, we have actual usage data that could help us with such a metric. There are a few problems with that approach, though:

• Popcon data is inherently wrong, because single computer installations tend to have popcon, while large installations (like the one in Extremadura) don't.
• A majority of popcon users uses the last stable release, not the one currently in development.
• 3 bugs in packages seldomly used can be more annoying for our users than one central bug - you might be able to find a work around somewhere for something that happens to a lot of people, but a bug only hitting a few users will stay unfixed for a long time.

I don't have more time right now, so the rest of my ideas will follow in the next few days, so stay tuned or add me to your killfile.

Yoe seems to believe I see a cabal in Debian. Let me assure you: I don't - and if there would be a cabal, I would be part of it. I've seen most of Debian, played around with NM, the release team, have seen how our buildd network works (or, in the specific case of today: mostly works) and have close contact to people in all other core teams of the project. I still haven't seen a sign of a Cabal.

I believe that there are "clusters" of people who work better with each other, simply because they have known each other for years and developed trust, but that's nothing I would actually mark as problem. I do the same. Many of our core people have too much work to do, mostly because getting a job well done is usually reason to give you more tasks until your performance becomes abysmal. That doesn't mean that they are bad people or do bad things intentionally, but often, they don't allow people to help them - the usual reasoning is "new people need training, training needs time, time is what I don't have => no new people". Obviously, that argumentation is flawed in the long run.

Anyway, on to the more technical details:

• The real reason for not adding new architectures is that not one of the people able to do so cares enough to do it. That doesn't sound as nice as "I don't really know why we don't do it", does it?


• I haven't asked the security team to support backports, and I won't. volatile isn't under the supervision of the security team and works just fine.


• Most porting machines aren't "down", they are "locked down". See the list of Debian machines (which is, BTW, seriously out of date). Hardware failure is something completely different (though we might want to think about using Debian funds to replace broken machines, if needed), but restricting access is a sign of an overloaded team of system administrators.


• Dictating how a team should be changed is usually not a good way to do things - but there isn't another. A few of our core teams have failed to solve their problems on their own. Debian's job now isn't to stand by, watching and saying "He, that's bad. Maybe they'll fix the team in the next decade". Also, there are people experienced with the tasks of these core teams who are not part of them - you might want to use their experiences.


• With my release team and unofficial buildd network hat on I will tell you that adding one buildd maintainer for those architectures that only have one at the moment would improve the situation a lot. Please don't assume that I only talk about stuff I have no idea of. Sometimes I actually know about things!


• I don't think that we need to change our position regarding the GFDL, but I do believe that we should work on our relation to the FSF. Debian is the biggest free-as-in-free-as-in-free-speech distribution on the market, so the FSF should actually be interested in working together with us. Still, that doesn't happen, so we should perhaps start by having a discussion what the problems of each of these organisations is in the eyes of the other.

And, FWIW: I do hate conspiracy theories too. They come to life because people usually want to see the "bigger picture" - even if there is none and all of the weirdness in their environment can be explained by chance, human error and unknown social problems. And as I said - I don't believe there is a group of people making decisions that seem bad for Debian. I just believe a lot of people try to do their best for Debian - but they fail on a regular basis.

It's this time of the year again: The one where I look back at my involvement in the project and decide that I've been a member long enough, so I should probably get out. For Debian, there are two ways to do it: Either become MIA or become DPL. As I have too many real life connections to other DDs, I guess I can't really do the MIA thing, so I should probably get elected as DPL. Whenever I consider this way out, I think about all the issues that make me unhappy about Debian. This time, I've written them down. So, here's a list of things I would like to see and would work on if I had the power to do so:

• Enforce addition of new architectures to the main archive. This includes kfreebsd-* and hurd-* - after all, we claim to be the universal OS. I've sometimes heard that our ftp mirrors wouldn't be able to handle the increased archive size - I don't believe that. We had the mirror split to work around this, not everyone is forced to mirror all architectures anymore. If actual Debian hardware should be a problem, we should use the money we collect to get the hardware we need.


• Make backports.org official, just like volatile. This doesn't mean kicking out the current team and give the responsibility for bpo to the dysfunctional ftp-team, but simply move the service as it is into Debian. A start for that would be to add backports.debian.org as CNAME for backports.org and then add fitting hints to our release notes.


• Resurrect the practice of porting machines. In ancient times, Debian had all kinds of machines available to all developers so that they could log in and debug architecture-specific porting problems. Nowadays, Debian collects hardware, Debian's hardware collects dust and Debian's BTS collects porting bugs. At the moment, most of these machines are not available to the developer body and access is granted to only a handful of persons besides DSA.


• Fix the constitution, so that the ftp-team, the account managers, the buildd admins, the system administrators and all other teams that currently don't see themselves as delegates of the DPL can in some way be changed when they are not doing their job. There is currently a GR for that in preparation, see the excellent proposal from Josip Rodin for more information. When that is done, we can turn to actually fixing those teams, starting by giving full permissions to active people that currently have only restricted access (ie Joerg Jaspert should be able to modify the LDAP holding Debian's account data and should also get to act as full ftp-master, not only -assistant). There are also quite a few people experienced with the tasks of the other teams, so we should add some of these people soonish.


• [Not really depending on DPL super cow powers:] Work on half-releases, like etch + 1/2. Besides adding a newer kernel for hardware support and updating the installer, we could think about adding backports packages to update desktop stuff. Many people seem to be interested in that.

Besides those things that I really miss in Debian, there are also a few things I would simply like to see done in the near future:

• Add more buildd admins.


• Resolve our conflicts with the FSF, get them to accept that we don't like the GFDL and then go on with lobbying for free software.


• Improve our appearance on conferences by creating common Debian material. This means creating "official" ISO images for DVDs from time to time (just like they are done by people for a specific conference from time to time), creating official information material that can be translated by the l10n teams, discuss guidelines for the people behind the booth (like "Please don't wear a shirt which allows visitors to see what you've eaten in the last week") and give advice for presentations.


• Clean up the DD list, so that we stop telling people that there are ~1000 DDs if only ~600 actually work on Debian. Yes, work on that is already on the way.


• Create a place (a ML, wiki page, pseudo package) where Debian people can simply drop interesting ideas. Often, busy developers have great ideas for new projects, but lack the time to actually implement them. Usually, the inside knowledge is only needed to conceive the idea, not to implement it, so we should have a place where people with free time at their hands can find great ideas. It shouldn't be needed to actually describe every detail of your idea, it should be more like "Wouldn't it be cool to collect all left socks and then get to world domination by tying them together?". This could also be the basis for a bounty system (and of course for the GSoC people)


• dpkg development - symbol-based shlibs look almost ready, but what happened to Wig&Pen alias dpkg-source v2?


• At some point, we should discuss our vision of Debian. I guess we won't find a common one, but it would be interesting to see where people want to see Debian in 5 (or 10) years. Should we "just" be a supermarket so that CDDs and forks can be based on us? Should we work on reintegrating all those efforts into Debian proper? Is that even possible?


• We should create a team (or, preferably, a DPL delegate) whose sole task is to communicate with the rest of the Debian ecosystem, so that common problems can be noticed, discussed and solved. I have to admit that I have lost track and couldn't say how many distributions are using Debian as their basis nowadays, but there seem to be quite a lot. I wish for a place where I can find such information (or at least a person who I can ask about it).

So, what do you think? Are all of those ideas crap or should we perhaps try to implement them? Have I missed something important?