The two-day conference is anchored with four keynote talks at the beginning and the end of each day, held in the main theater. The rest of the talks are split into five tracks in five theaters/conference rooms. It's a very fast paced conference, with ten minutes between talks and twenty minutes lunch break. Attendees would pick up the box lunch and go right back in to lighter lunch talks. The most often heard refrain is "out of the five talks, I want to go to three." Fortunately, the Strange Loop team has partnered with InfoQ to record all sessions and make them available over the next few months.
Unlike in past years, I did not take copious notes during the sessions :(. So the rest of this post won't be as long as in past years:
The four key note speakers this year are Michael Stonebreaker, Lars Bak, Jeff Hawkins, and Brendan Eich. Each talk is excellent in its own way. I was particularly impressed by Jeff Hawkins' talk Tuesday morning about the brain and how to do computing like the brain. Watch the video when it comes out if you missed this year's Strange Loop.
Michael Stonebreaker's talk about in-memory database is "controversial" at the conference because he has a different idea of how a new generation database should be designed and he think the NoSQL movement is misguided. I believe he's attack on "eventual consistency" is logically sound. And the counter argument, which I hear from other attendees after the talk, are: "We are using it, and it works very well for us," and "The world is inconsistent."
Lars Bak talked about his work on virtual machines old and new, including Self, Hotspot, V8, and Dart. As a distant observer, Dart strike me as a very reasonable goal to pursue. The performance advantage that Dart has over V8 will, I believe, play an important role in its eventual adoption. I feel my first reaction to Dart 331 days ago, If You Love Web Programming, Root For Dart, is more justified.
Of the break out sessions I went to, I enjoyed the session by Daniel Friedman and William Byrd demonstrating relational programming in miniKanren, Matthew Platt's session on the Racket Way, Jim Weirich's session Y-Not, Daniel Smith's session on Project Lambda, and Ola Bini's session on expressiveness and abstraction of programming languages.
Several of the talks are about databases and data in general. Rich Hickey's talk on Datomic presented a who new way of dealing with data. Michael Stonebreaker talked about "data coming at us with incredible speed and volume." Even Cliff Click is talking about a BigDatabase called H2O.
On the lighter side, I went to Gary Bernhardt's fantasy of writing his own editor and terminal and Matt Butcher's rumination on philosophy and how Plato is an object-oriented guy and Aristotle is a functional guy.
The #strangeloop Twitter stream contains a lot more snippets about the conference from the attendees as well as presenters. It's fun to participate in it.
This year's Strange Loop conference is another awesome event. My sincere thanks to Alex Miller and the Team for a job well done.