Next up was Jim Gettys about the OLPC project, he talked about all the hurdles they had to go through to make such an efficient laptop, which is sturdy, usable for children and doesn't suck up much power.
A lot of development and unique features have gone into this piece of hardware. The screen, network controller, getting SUN to open up OpenFirmware, etc.
The morning ended with a presentation from Simon Phipps on the process of releasing the Java source code under GPL. Which is going pretty well, although there are some things that they cannot release, they are in the process of replacing all these bits by other GPL software. We can expect the first release somewhere later this year (that was rather unclear).
Lunch break! Time to eat my sandwiches and donate some money to the event (getting an O'Reilly pocket guide in return). Settle myself in the Chavenne room waiting for Keith Packards speech, this would be the second time I attended a session from him at FOSDEM.
He is a very entertaining speaker, easy to listen to. Not really a tech talk this time, but a chance to meet the whole X.org team. Surprisingly the group is not that large, for this rather gigantic piece of code.
Next topic was hard to pick, but I ended up going to the LinuxBIOS talk of Ronald Minnich. Who explained that no matter what you use as your bios, linux is always faster and as a bonus you have a real shell available to you when things go belly up. However, the project was going fine until kernel 2.6 came along. The kernel got to big and the flash on the MB was getting smaller. The LinuxBIOS was no longer really linux at that time, but luckily the tide turned and flash with sizes of 2MB is no longer an exception and has plenty room for a complete mini linux install.
Yesterday ended with Andrew Morton with a general talk on the linux kernel, where it is going and the different areas of attention. I must say that Andrew is not that good at talks, compared to Alan Cox for example. Ofcourse, that doesn't matter too much. There is a lot in store for the kernel the next couple of years, but one of the most interesting things that got my attention was that almost all kernel developers are now employees of some company, while most of the testing is done by the community. In some way this brings up the question if the linux kernel can even survive without commercial backing.
As a bonus this picture, if you look very hard you can spot the running fried potato slice.
