OSCON is over. It was fun, but it's a relief to get past it and back to life also…
Of the morning keynotes, Damian Conway's "Talk 2.0" was the most memorable for me. However, Dave Bradley's talk on the creation of the IBM PC was also quite nice, as was Bernard Malmud's talk on hacking government. (The trendspotting talk was fine, as usual; not clear what to make of it, I think.)
There was also the schwag drawings from the review form box; this turned into something of a farce as one person was given an award for reviewing Eben Moglen's talk, which hadn't happened yet, and it was clear from repeated draws that other folks had stuffed the ballot box. Apparently cheating is fair game; note for next year.
After that I visited a session on Djinni 2.1, an open source implementation of a clever local search algorithm for attacking instances of NP-hard problems. (Factoring, for example, is not known to be an NP-hard problem.) I wanted to next attend the session by Alex Martelli on Software Project Management, but by the time I got there the room was jammed so insanely full that I just left.
Finally, Eben gave his talk. (His first question was to the previously-mentioned reviewer, about "how did my talk go?") It was a masterful talk, as usual, and received a standing ovation.
Overall, OSCON was quite an interesting time. It will be fun to try to summarize it all for the article I'm supposed to write very, very soon. (B)