We got 593 attendees, a big jump from 410 last year: a 44% increase! When early-bird registration ended, I predicted around 530 attendees, but online registration was stronger than I expected. Even the at-the-door registrations were surprising: we had 10 last year, and a proportional increase would mean 14-15 this year, but we had around 20-25 instead. Unfortunately, conference swag was initially ordered based on the number of 530 ; the T-shirts and totes were later bumped to 575, but that didn't help the sponsors who had already shipped their advertising materials.
I suspect the improvement is because we did more publicity this year, thanks to Catherine Devlin sending out press releases and e-mails. The survey results are trickling in, and I think they confirm this hypothesis: about 50% of respondents say this was their first PyCon.
The keynotes were a great success and people kept complimenting me on them, but all credit for the selection should really go to my co-chair Jeff Rush -- I only saw portions of the keynotes. (I hope I'm not misrepresenting any of the speaker's arguments!) Ivan Krstić's talk about the One Laptop Per Child project created a lot of enthusiasm for their interface and OS, and people were eager to play with the machine. Dr. Adele Goldberg's discussion of the situation in US public schools was a sobering corrective, noting that technology is not a quick fix and requires instutitional and community support. Robert M. Lefkowitz's keynote was excellent, comparing the history of writing, literature and rhetoric to the still-early history of programming. Both Dr. Goldberg and Mr. Lefkowitz were surrounded by people for the rest of the day after their morning speeches; Mr. Krstić gave a well-attended hands-on session with the XO the evening after his talk.
The conference ran much more smoothly this year. Last year I was running around frantically for all of day 1, but this year, after the initial flood of people at registration before the keynote and during the first morning break, there were significant stretches where there wasn't anything critical going on. It actually is possible to chair PyCon and yet be able to attend talks, something that will reassure David Goodger, next year's chair.
I was worried about everyone fitting into the rooms. We figured that the venue could handle 600 people but such calculations could easily be wrong. In the end, the rooms and hallways were pretty full, but traffic kept flowing without any jams. The smallest room, Mesquite, got quite crowded a few times, but the larger ballrooms always seemed tolerable. (If very cold -- I kept asking building engineering to turn up the heat, and it helped a little, but the rooms were still on the cool side.)
Thanks to everyone who volunteered, and to everyone who came!