JAOO Brisbane 2009

The JAOO 2009 Conference in Brisbane finished on Tuesday.

Day 1
I missed most of the days program, although I did get to hear Clemens Szyperski talk about MEF and Oslo at QUT in the afternoon.

Jonas S Karlsson from Google talked about Consistency, Storage, and Reliability in the Cloud. He was involved in the Google Megastore project which provides a scalable (in Google terms) storage layer with SQL style schema and secondary indices and consistency. I believe he made a comment about the importance of the dollar cost per user (of a hosted app) to be a single digit, especially when the app is available free of charge.

Mike Cannon-Brookes finished the day off with lessons learned at Atlassian.

Caught up with various people at drinks afterwards and went out to dinner with some of the presenters at SouthBank.

Day 2
Joshua Block gave the keynote on various technicalities in Java. It is not often that someone mentions covariance and contravariance in a Java talk! Seems to come up more in Scala discussions and then people are scared off by the terminology. :-) While probably not the intent, I think the talk was great in that it clearly demonstrated some of the complexity of Java.

Douglous Crockford, Writing Large Applications in JavaScript. Interesting talk, although I think it was a mistake for him to use Java as an example of strong, static typing in his argument for the benefits of loose typing in JavaScript.

Mahout by Jeff Eastman. This was given on a few hours notice to fill in for someone who was sick. I have done stuff with Hadoop before and am developing an interest in machine learning, so I found this talk interesting. There was someone in the audience who was a little confused though, as he thought Jeff meant java.util.Vector whenever he used the term vector. :-)

Unfortunately Joe Duffy was sick, so his talk on parallel programming was canceled.

The last talk I went to was Joel Pobar on F#. It was great to hear from a local who knows some stuff about functional programming (FP). He was suffering from a cold, but sill powered through some interesting algorithms. I don't think he helped the mainstream perception that FP is really technical (given that many are scared by this) when he mentioned vector spaces. :-) We talked to Joel afterwards about the Brisbane FP group starting up and he was really excited, so it will be fantastic to have him involved.