OSCON 2011: A Retrospective

I got home from Portland, OR terribly jet lagged on Monday morning, but buzzing from my trip to the US and my experience as both an attendee and speaker at O’Reilly OSCON. Now that I’ve caught up on some sleep, it’s probably about time I reflected on last week.

The first thing I noticed: OSCON is huge. Really. This year there were over 3,500 Open Source enthusiasts present. That’s a bit over 11x as the size of Australia’s OSDC and over four times the size of Australia’s biggest Open Source conference, Linux.conf.au. While they were spread over 15 streams at any one time, it was abundantly clear just how many people were present when we were all packed into the main hall for the various keynotes.

The second thing I noticed was the people. OSCON was packed full of leaders in the Open Source space. Josh Marinacci of Swing Hacks fame gave me a brief demo of Hewlett Packard’s webOS over beer. In another beer-fueled geekfest, I sat dumbfounded at a small table listening to Charles Nutter, Ola Bini, Daniel Spiewak, Carson Gross and Martin Verburg talk shop about programming language design and implementation on the Java platform. Later that night, Google’s Brad Fitzpatrick and Andrew Gerrand would cheekily suggest Go support for Rik Arends’ startup, Cloud9 IDE. (Was more beer involved? Perhaps. It’s a little hazy.)

As you’d probably expect, there were some great presentations, too.

One of Twitter’s engineers revealed in a keynote that they had to patch the garbage collector in the de facto Ruby standard interpreter, MRI, in order to keep it fast enough to serve up their pages. Meanwhile, they actively work to replace the underlying Ruby & Rails code base with JVM technologies (most notably Scala, but more interestingly Clojure is on the radar too).

Les Hazlewood of CloudDirectory.com gave a great presentation on Apache Shiro, a powerful Java security framework. Another presentation I missed but which I heard a lot about on Twitter was Daniel Spiewak’s proposed programming language that’s still waiting on an implementation. Daniel Hinojosa gave an interesting overview of Scala testing frameworks, and Joonas Lehtinen demonstrated an interesting Java web application framework called Vaadin.

The OSCON Data/Java streams were an interesting experiment — one that I’m not entirely sure was successful. I get the impression that many people chose to wait for the start of the main conference on the Wednesday, bypassing the first two days of OSCON Data/Java. For example, Charlie Nutter had perhaps 15-20 people in the audience listening to his presentation JRuby: Pushing the Java Platform Further on Monday. It was an awesome presentation and for a conference like OSCON, it was perhaps somewhat disappointing to see so few people in attendance.

I was fortunate enough to speak at OSCON Java too — Open Source Compiler Construction for the JVM. This was essentially a 40 minute live coding session in which I explained and demonstrated the various steps involved in building a simple expression compiler for the Java Virtual Machine using Scala parser combinators and Apache BCEL. You can see the code I wrote during the presentation on GitHub: http://github.com/thomaslee/oscon2011-awesome.

All in all, I had a brilliant time at OSCON. There were some great presentations and keynotes, but more importantly I was fortunate enough to meet of the most talented hackers in the business. OSCON was humbling and inspiring — I hope I’m lucky enough to have the opportunity to do it all over again next year.

About Tom Lee

Senior Consultant for Shine Technologies
This entry was posted in Opinion. Bookmark the permalink.

2 Responses to OSCON 2011: A Retrospective

  1. this sounds like it was a java/jvm conference

  2. Tom Lee says:

    @sulfidesmoldersulfide haha you’re absolutely right.

    In retrospect, I attended a lot of talks for the JVM, although that wasn’t necessarily my intention. There’s a lot happening in that space, especially on the programming language front. In case it’s not obvious, I have a bit of a passion for the language design & construction side of things and lately my focus there has turned to languages for the JVM (most notably Mirah of late: http://github.com/mirah/mirah)

    Probably the most interesting non-JVM presentation I saw was “Using jemalloc for Large-scale Memory Debugging” (http://www.oscon.com/oscon2011/public/schedule/detail/17939) on Friday. Had no idea jemalloc was so powerful, despite hearing a lot of talk about it a while back. There was also plenty of NoSQL stuff happening there too — and although I missed his presentation, Basho’s Rusty Klophaus (@rustyio) was great to talk to.

    In any case, apologies for the lack of non-JVM love in this post. Certainly not my intention to leave other tech out in the cold!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s