I talked about Spring 3 at the Spring User Group (SUG) in London on Wednesday night.
The turnout was pretty good I reckon - thirty or so folks were there. The talk I gave was only... okay though. I really was making it up as I went along, so it has a very rough stream-of-consciousness feel to it. The audience seemed to like it though, and I tried to get some energy pumping around the room and to encourage some strong and contrary opinions to be voiced. One of the very fair comments I got back was "Stop saying cool all the time." Indeed... overuse of the words "cool", "dude", and "man" is a most tiresome peccadillo of mine. Duly noted.
Sadly, the SUG talk did suffer from fake-osity in marketing. The blurb for the talk - all done without my knowledge - said that:
"Since leaving SpringSource, Rick has continued to be one of the most
prolific contributors to the Spring Framework and is therefore
well-placed to deliver a fascinating insight into the next release of
Spring - Spring 3.0."
I haven't contributed anything to Spring beyond a single bug report on JIRA. I'll watch out for such fake-osity in the marketing blurb in future.
The compère for the evening - I use the word compère in the loosest sense - also introduced me as "the revered Rick Evans, who probably knows more about Spring 3 than anyone else". Juergen and Arjen - to name but two of the folks who wrote Spring 3 - know a heck of a lot more about the whys and wherefores of Spring 3 than me. I was quick to puncture this balloon of bollocks-osity at the talk by telling folks how it really was... hopefully they left that bit in in the edited video.
The SUG was the first Spring-related event I have done since leaving I21, and it felt odd talking about Spring from the other side of the fence. The compelling reason for having Andy Clement swan off and reinvent a wheel in the form of the Spring Expression Language still eludes me: I daresay SPEL will find much use in Spring Batch/Integration/WebFlow in the coming releases, but the immediate applicability of it is not apparent beyond some simple scenarios. (The test fixtures for the SPEL support are very pedestrian and contribute pretty much nothing to understanding the 'why' behind the implementation.)