This morning the conference part of JavaPolis kicked off. As expected the pavilion floor was now really crowded with people. Attendence broke the 2800 mark with people from over 50 different countries.
Wifi at the conference is still messed up, so I'm once again typing this from the hotel bar. How much of an excuse do you need in order to be forced to sit in a bar ? Beer please! 😉
I was thrilled to see some excerpt of my blog posts making it to the front page of the daily JavaPolis newspaper that was handed out to all of the attendees this morning!
Because we arrived early, we managed to get some front row seating to be sure not to miss out on any of the keynote action. The keynote session was broken up into three different parts. Stephan Janssen (president of BeJUG) started off with the official opening of the conference, or actually there was no real ceremony involved. His talk was mainly structured by announcements, of which the most important ones are listed below:
- SpringOne 2007 will take place at Metropolis Antwerp from 20 – 22 june 2007.
- Stephan opened up a new site called Parleys.com that essentially is a Web 2.0 vehicle for publishing all of the online talks and e-learning stuff form JavaPolis and related conferences. The site currently only works with FireFox because of some dojo issues.
- JavaPolis 2007 will take place from 10 – 14 december 2007. Most probably at Metropolis, but as the venue is rapidly becoming too small for the number of attendees this is currently under thought.
The second part of the keynote was by Oracle being one of the main sponsors of the event. Omar Tazi (Chief Open Source Evangelist) did a 45 minute non stop clicking show talking about Oracle Fusion Middleware world domination. Same thing as ever…it looks good at first sight, but when you think about it, or worse, when you start working with it you'll learn just to stay far from it. Got to admit the Business Activity Monitoring dashboard looked cool though. 😉
Keynote part three was the interesting part. It was the Sun Technical Evangelist team performing their geeky toy show. For me it was a re-run of the stuff they already did at the NL-JUG J-Fall conference we ran a few months back, but still the geeky toy stuff is entertaining to watch. There was even some new stuff in it as well. Angela crafted up a Sudoku application built in Java3D that ran on Project Looking Glass and could be controlled with the 'Minority Report'-style VR glove powered by Sun SPOTs. Cool!
When the team from Sun rounded up it was about time for a cup of coffee and a quick glance at the conference schedule to determine the next session to attend. I picked a session on 'Java EE Enhancements for Deployments' by Nagesh Susarla from BEA. He showed us some of the new deployment stuff present in Java EE and how BEA built some convenience tooling on top of that. The good thing is it should also work with other application servers so I will definitely check it out as Java EE administration and deployment is still an era that runs in the dark ages. The session ended up in some sort of a vendor pitch for WebLogic Server 10 and a quick run-through of some of it's more spectacular features. WLS 10 should be general available somewhere around februari/march 2007.
After lunch break I attended a session called 'Failure Happens' by Bill Venners. I really looked forward to this session as Bill is on my list of all time Java heroes. If you're not familiar with his work, you should definitely check out his series of articles on JVM internals at JavaWorld. Unfortunatelly, the session turned out to be a little bit of a deception. The talk wasn't really smooth and Bill barely scratched the surface of solid exception handling strategies as he intended to when showing some bad/best practices for exception handling. The session was more of a philisophical thing on dealing with failures in general. Too bad and quite a disappointment.
I spent most of the afternoon and early evening in sessions. In particular some of the new JSRs got my attention. JSR 277 Java Module System by Stanley Ho, who arrived fashionably late (30 minutes) but still managed to deliver his entire talk in the remaining time and JSR 296 Swing Application Framework by Hans Muller. I learned that the Swing Application Framework is mainly targeted at the novice Swing developer, just to get him up to speed with the overcomplicated set of APIs needed for proper rich client development. Although Hans stated there were few bells and whistles he presented some raunchy features like typed resource bundles and resource injection that made me jump my skin. In my opinion Swing definitely needs a lifecycle framework together with a means to run long running tasks off of the Event Dispatching Thread but that's about it. Maybe add a little state management. The promising thing is that the expert group has to be formed and all features are still under discussion. Hopefully the expert group will shoot the resource bundles stuff right in the head, because we already have horrible stuff like programming in XML (hello, Spring!) but programming in properties files is amongst the worst things that could happen to the Java language. As are XML literals (please cut them out of Java 7, please!) closures and some other funky stuff currently under debate for Dolphin.