Day two at Devoxx was another university sessions day.
I started off with a presentation about Seam by Dan Allen. It was a good presentation on a nice tool. Unfortunately it is in essence a tool about integrating JSF with EJB’s. It does a lot more too, but that was still the main topic of discussion. I really liked the part where you can ultimately just plug in any persistence tool or system and any state tool or system and Seam just taking care of it. But JSF and EJB’s are not all that interesting to me at this moment. Also I had arrived late, so I really had to go to the toilet and the break just got postponed and postponed, so in the end I just walked out and didn’t come back.
Lunch was a pasta salad with tuna. I really don’t know why they said it was a shrimp salad, there were about three shrimps in it. It tasted good, but I think a lot of people were hungry halfway through the afternoon. But then they had more salad and also some slight variations on the tuna-theme.
After lunch I went to a session about Glassfish. Or rather several short talks pushed into one long university session. They talked about Glassfish v3 Prelude for a while (Alexis Moussine-Pouchkine), and then talked about project Fuji (the next generation of OpenESB), OpenMQ, Jersey which is an implementation of JAX-RS (Paul Sandoz), Grizzly Comet (Jean-Francois Arcand) and WebSynergy which is a webportal tool (Satya Ranjan). This is no longer in the right order, cause I can’t quite remember what the order was. Andreas Egloff talked about Fuji. Fuji sounded nice and the graphical tool they had developed for putting together the applications looks a lot like what I have in mind for letting kids build applications on the SunSpot. You could just drag components onto a palet and then connect them. So maybe I could use Fuji as a starting place and simply define leds and sensors as Fuji components. Linda Schneider gave a short presentation about OpenMQ. Her laptop wouldn’t connect with the beamer, so the first half of her presentation was without slides. I was very impressed with how she handled that. Many people wouldn’t have been able to cope. For Jersey it took me most of the presentation to remember that JAX-RS was about web services. Of course _now_ I see that it was in the program, but it would have still been nice to have heard it in the first sentence. From the Grizzly Comet I only remember the explanation about how Comet has two other mechanisms beside the standard request/reply of HTTP. They had a delayed reply (letting the browser think that it’s a slow server) and a streaming reply (sending the answer in several partial replies). WebSynergy isn’t out in a final version yet, but it sounded very cool and I’m definitely going to look into it.
My third session was about Benerator by Volker Bergmann, a tool for creating performance test data. A truly awesome tool that could generate lots of test data from a database schema or even obfuscate production data.
And the final presentation was by Xavier Hanin about Apache Ivy, a dependency management tool. This tool looked very nice as well. You can plug it into your IDE and then create an IFL file where you tell it which projects your project depends on. In combination with a Maven repository it will then even download those projects for you as libraries, including all the projects those projects need. It can also checkthe inconsistencies in your dependencies, where you can specify which version you need and even that any version from a certain version up will do.
