Did not sleep really well, but in contrast to Waldemar, I *did* sleep. Earplugs just rock… Got up at 8:15am, took a shower, and had breakfast with Scotty. Basically it was the same breakfast as yesterday, but this time we had Java waffles for dessert. The hostel was really strict with serving times and you could not even get a glass of orange juice after 9am.

We packed our stuff together, checked out and drove to the university campus again. (Following exactly the same route as yesterday, which was a really smart idea[tm]).

After checking mails, Waldemar and I went to the Scratchbox: Cross-compiling a linux distribution talk. The speaker was difficult to understand because he had a strong finnish accent but after a while I got used to it. Easing cross-compilation inside a sandbox on a faster host machine is probably a smart idea, though.

Next we headed to the Introduction to PHP SimpleTest talk. It was located in one of the smaller rooms and had a more familiar atmosphere. When we arrived the previous talk about eZ publish was still in progress and we followed the last slides about different localization methods for php application. The SimpleTest talk was disappointing though, the speaker was probably really nervous and his slides lacked a clear thread. But I’ve learnt that simpletest 1.0 had been released a couple of days ago. Someone asked the speaker how one could test the generation of dynamic webpages with simpletest. You can match the webpage output with some regular expressions, but that does not really work with dynamic stuff, like dates and so on.

We went for sandwiches and some beer for lunch which was a cheap and convenient choice before we went to the mysterious introduction of the OpenBSD track. The introduction was held by no one else than Theo de Raadt. Maybe he was still intoxicated but he made a nice impression.

I skipped the rest of the OpenBSD track and went to Alan Cox’s Kernel talk. He gave his opinions about the changed development process of 2.6, how Linus treats kernel security issues (rather bad in his opinion) and how he deals with kernel patches (he is using evolution for patch management). One thing that surprised me was how Alan was interacting with the audience in contrast to RMS. He also had trouble understanding questions, but unlike RMS he actually left the lectern and walked up to the questioners. There are lots of pictures with him standing in the middle of the audience and answering questions. One interesting question was, why he refuses to use bitkeeper for patch management. He revealed that while he can understand Larry McVoys position, he is unable to commit to the bitkeeper license part that forbids anyone to work on other source control systems within 2 years since he might have to fix other scs, like cvs, as part of his redhat contract.

The main reason why I wanted to go to FOSDEM was Alan Robertson’s talk about LinuxHA. And boy, he really delivered. (Well he works for IBM and has to sell stuff as well.) He had a great way of interacting with the audience and dropping lots of witty remarks. He presented upcoming features of the LinuxHA 2.0 release and gave an overview of the general framework. The first .0 release should appear in the upcoming weeks.

I stayed in the same conference room and listened to the OpenGFS talk. However, I was getting sleepy and due to a lack of oxygene I did not really follow.

FOSDEM came to an end and I rejoined Scotty and Waldemar to listen to the last talk, GPL enforcement held by Harald Welte. He spoke about how they managed to force LinkSys to open up their sourcecode because they were using gpl licensed software on their wlan routers.

Scotty was our geekladen “poster boy” for the day and we managed to sell Harald one of our pear shirts after his talk.

We left the campus shortly after that and headed back home (of course we got lost again, but that was only minor). We dropped Waldemar off in the middle of nowhere (= Odenthal) and fortunately Scotty also drove to my place and dropped me off.

Overall it was a really cool event. I really liked the idea of the smaller developer rooms and I wished that I could have attended more of those talks. A chillout zone would have been nice since the halls were always crowded but that was probably due to the really cold weather during the days. Oh, and we should have printed out some maps in advance but maybe next time we have a gps device on board.

Finally, if you are still reading, there are some pix from the weekend.