Devoxx 2010: Day 1
By Geertjan on Nov 15, 2010
There were not too many people attending the lab, but all of them were serious about their interest in the NetBeans Platform, one of them briefly showed me his NetBeans Platform application and I've been trying to find him ever since then... for a screenshot and interview, etc...
I then attended Guillaume's Groovy session, at least, the first two hours. It is great to see the Groovy ecosystem is quite large and diverse. I wonder, though, what proponents of scripting languages such as Groovy (the best of the lot, from a Java point of view) feel about "the fact" (quotes because it's not a fact but my opinion) that they're not pushing Java off the throne, which is what I thought they were aiming at, but that they're (especially Groovy) glue languages. If you look at Groovy as a glue language, then you're never going to be disappointed. But if you see it as a potential replacement for Java, you're in a whole different discussion. I simply don't see large business applications (except for those using Grails) moving to Groovy as their programming language, especially not large business applications (e.g., banks, defence, oil, etc). However, I do see the applicability of Groovy as a key glue language in all these sectors.
After that, I saw a session on Mylyn and one on JCite, which I had never heard of before, which is really interesting for synchronizing code in tutorials with JUnit tests (this combination, but without JCite, is what Jaroslav Tulach used too for his API design book, to ensure that the code samples compiled and worked as expected). Definitely something to investigate.
I attended the Groovy/Grails BOF in the evening and then a very interesting one by Heinz Kabutz on jpatterns.org, where he introduced an annotation for stating the intention to use a particular pattern. The discussion centered on what one could do with that annotation, e.g., create a tool to verify whether the pattern has been used, but how to identify that usage? Or maybe as a way for newbies to a project to understand the intention of the code, i.e., by looking at the annotations. Very interesting, anyway, regardless of what you think about patterns. (By the way, Heinz stated that he considers JDeveloper to have the best tooling for UML, i.e., better than any other UML tool, regardless of free or open source or whatever.)
The day ended with a long drinking session with CollectionSpy (sounds like a very interesting Java profiler tool), Heinz, Paris (Greek JUG leader), Sven, Florian, Martin, and others.