TheServerSide Java Symposium was in Las Vegas on March 21-23. This year was my first to speak (and attend, for that matter). It was a well-organized conference with some top-notch content. The sessions that I enjoyed the most were:

• The Enterprise Service Bus: Do We Really Need It? By Mark Richards of IBM. Mark is an excellent and entertaining speaker. This presentation was similar (but not identical) to a presentation of his that is posted out on InfoQ. The answer to the question in the title of the presentation is, not surprisingly, "it depends" on your needs. One of the real values of the presentation is Mark's deciphering of the various vendors' descriptions of what an Enterprise Service Bus actually is.


• Measuring Up Performance by Kirk Pepperdine. This presentation is drawn from Kirk's consulting experience. He discussed performance tuning anti-patterns. In other words, things that people do wrong when they are attempting to solve performance problems. I recognized my behavior in one of the anti-patterns that he described. I discussed it the following day during my presentation: Memory Leaks in Java Applications: Different Tools for Different Types of Leaks.


• Glassbox: Open Source Monitoring and Troubleshooting by Ron Bodkin. I had heard of Glassbox but had not had a chance to study it. Ron included some very nicely done demos which helped highlight the value that Glassbox provides. I could see using it to help do an initial narrowing-down of performance problems on a web application.


• Effective Concurrent Java and Java Performance Myths: "How do JVMs Really Work?" by Brian Goetz. I had exchanged emails with Brian, but had never met him before. He's an excellent public speaker. Through a scheduling quirk, he ended up doing both his talks on the same day - and he did the performance myths talk twice. He was still going strong at the end of the day which was impressive. Anyway, of all the material in the performance myths talk the most important is this: advice on how to write Java code that performs well is short-lived because the JVM keeps improving all the time. So if someone gives you a tip, double-check to make sure it is still valid.


• JRuby: Ruby-Tinted Classes by Charles Nutter and Thomas Enebo. I don't know much about Ruby - although I have read some books about it. To me there were two things that were really interesting about this presentation. First, they started by doing a poll on usage of Ruby and a lot more hands went up than a year ago (according to the presenters). Second, they included a demo of the Ruby support available in the NetBeans IDE. For those of you who have an interest in that Ruby support, be sure to come to NetBeans Day on May 7 (it's free, but you do have to register, more info. here).

Next up for TheServerSide Java Symposium: Barcelona in June. Necssito diccionari anglès-català. :-)