Second day at DebConf

Back online. Awake, I mean.

I had this article half-written and just saw that the DebConf team has published the proceedings. So, I’ll shorten the summaries and focus on my opinions about each of the events. This will apply for the remaining days of course, but I’m tired enough not to edit yesterday’s article (at least yet).

Breakfast

This breakfast was quite more interesting, since I sat with some other Debian guys. Talked about i18n and the Debian philosophy. Cool.

hp.com/go/debian

An interesting review of Hewlett Packard and Debian by Bdale Garbee. I particularly liked the part in which he emphasized that HP is pursuing profit. I very much appreciate sincere comments.

Managing 666 packages, or how to tame the beast

Not much to comment in this one. The speaker went way too fast, and as I’m not a DD it wasn’t that interesting to me anyways. As always, “yet”.

Lunch

Enjoyable. Quick, as I had to buy cheese for tonight’s party.

Quality Assurance in lenny+1

I had a big discrepancy with Lucas Nussbaum’s opinion on how to deal with orphaned packages. Anyways, we talked about it offline and we seem to have agreed to some degree.

The point is that he said orphaned packages with low popcon score should be removed from the archive. Meanwhile, on IRC we had a discussion about accessibility packages, which of course have low popcon score as it’s pretty little chance Orca will only be used by blind people, but are crucial for those few and I think they have to be available even if they were full of bugs and anyone is caring about them (finding someone to do care is topic for another discussion). Those packages are as important for those people as the kernel for us. Even more perhaps.

Locating bugs to kill with SOAP

Don Armstrong has a very funny and interesting way of presenting his projects. However, I’m thinking lately that there’s too much effort trying to enhance the way DDs deal with bugs. Effort that could perhaps be targeted somewhere else, such as actually dealing with those bugs.

Packaging with version control systems

Again, Madduck gave an excellent speech. This time about some sort of meta packaging tool, something that’ll help different distributions to interact with each other.

What I couldn’t shake off my mind is that such interaction must be given by upstream. And if upstream refuses to implement something all major distros are including by themselves, the problem shouldn’t be solved by such a tool but by making upstream come to their senses or understanding their reasons and maybe even forking.

My point is, I love interaction. I love sharing. I love working together. But I don’t think interaction between distros, dodging upstream is good in any way.

the Debian Videoteam — Behind the Scenes

The typical infrastructure presentation, with lots of networked devices in most of the slides. And, as in everyone of those presentations, the speaker was very funny.

Bugs in large packages

With little participation from us listeners, Don tried to enhance the way maintainers deal with packages that have way too many, or really old, bugs. He’s been working on Debian BTS’ frontend, and although it looked kind of a work in progress, it looked useful.

Dinner

Arrived late. There were no chairs left, anywhere.

Debconf cheese party

Just bought my little contribution to this event, some cheese made of sheep milk. But, there won’t be updates about this today. No drinking and blogging is a strong policy I respect quite a lot.