Very quickly, just to complete my write up and impressions of this event, I had three panels on Sunday and then had to leave to catch my flight home.
I concur with my traveling companion, Chris Miller, that the first panel, the future of programming languages, was one of the best ones on which I participated. Peter Salus and Vernor Vinge were on this panel as well as Catherine Devlin and Jay Wren. The latter was moderating and a good thing as he is an averred .NET developer, an odd choice for an open source convention. I was able to keep my temper and speak with apparent humor when he let his own passion for Microsoft get the better of him.
Vernor and Peter provided a splendid historical view and the contrast to new and upcoming languages was very informative. I think Catherine and I managed to add some reasonable perspective on current efforts and from the view of practicing programmers. I was especially pleased that Vernor brought up the topic of many core and I was able to reference some of my recent reading on the state of research and practice in this area.
The next panel was actually more of a BoF session. It was the OLPC roundtable and consisted mostly of a gathering of XO1 owners and folks interested in the XO1. I mostly took notes, especially since Mako was present and others asked questions about the recent OLPC re-organization about which I myself am curious.
The last panel was another one where I felt out of my depth. I was late, which didn’t help, and sat for over half an hour without an introduction or saying anything. Eric Raymond, Randall Munrow and either Larry Kestenbaum or Garrett Kajmowicz.
The conversation during that time was mostly voting statistics and the various influences on it, including partisan politics. Towards the end, I finally got introduced and managed to mention e-voting. There was little time to discuss the topic, though, and probably the best though expressed was by Eric Raymond who expressed irritation at open source being suggested as a panacea for the security ills of e-voting systems. His point is that without a way to verify the loaded binaries, open source wasn’t any different from any other development model. Makes sense and doesn’t erode the value of open source in general, he was just suggesting that too many geeks were willing to overlook this significant problem.
Despite being ill for most of the weekend, I enjoyed the con immensely. Even on the panels where I felt overwhelmed or intimated, I was glad to be there and the other panelists were nothing but kind, smart and welcoming. I again agree with Chris Miller’s thoughts on the depth of the technical sessions, I would have liked to have seen something comparable to the typing of programming at WWDC or JavaOne, with source code and running applications on the projection screen. That isn’t enough to curb my desire to head out to this con next year, the travel cost and logistics will be the biggest consideration.