Looking forward to seeing you all again on Wed February 24, 2010.
My talk on Hadoop::Streaming had to cover too much breadth of background to spend much time getting into the code. I look forward to doing a part two where I display a non-trivial example. The talk and slides covered an overview of Hadoop, Hadoop streaming, writing streaming jobs manually, writing streaming jobs using Hadoop::Streaming from CPAN, testing jobs locally, and finally running jobs on the cluster.
The slides will be posted soon. There was a hardware failure on the *.pm.org hosting box on Tuesday into Wednesday, so I didn't push an update. You can see them in raw vroom format in the Git Repo.
Aran was kind enough to come down from Thousand Oaks and present a nice talk on "Coding Concisely in Perl." This was a fun, interactive presentation discussing various ways to make your code (or your inherited code) cleaner and more representative of its core ideas. From simple changes like using ternaries for setting variables, using map/grep to replace cut-and-paste code, and finally a look at the ValueClick method of mocking the production DB. Their DB Mock layer uses a shim to create a local sqlite3 db by autoloading the schemas from the production databases. It's quite clever.
The slides on DB testing have company code on them, so we are waiting on his CTO to approve releasing them for publishing. I expecte to have them up on la.pm.org soon.