I got up early this morning to check my mail (wasn’t able to yesterday due to the all the busy-ness) in the hotel. Fortunately I have quite an okay hotel with working internet and towel that actually allow me to dry myself after getting a shower (okay Jon, I’ll quit whining about your choice of hotel right now and won’t ever mention it again ). I hope I’m not being too much of my usual Dutch cynical self here, but hey, you been living in Holland right, so should know that Dutch are like that :).

Arghhh, shite! It’s as if somebody eveil is watching me type this! I just about finished that last sentence and then the internet dropped out (that’s no joke ). Anyway, I can keep on typing and post the entry later on anyway.

Yesterday was my first day of JavaPolis this year. In short, it was good! I had prepared a 3 hour talk for the university sessions that went quite well. I focused on three major areas in which we’ve improved Spring. Ease of configuration, simplifications in the web area and the decreasing the code-test cycle. It was great fun presenting together with Arjen (I hadn’t actually seen him for a while, we had both been on the road for quite some time), although I hope he though my jokes I put in the slides weren’t too cheesy.

Anyway, the good thing about JavaPolis is not just the sessions (the conference part of JavaPolis hasn’t even started yet). It’s also about meeting nice folks. Yesterday night we had dinner with a group of people at a Mexican restaurant near the conference center and I met Jan Newmarch from down under (presenting on Jini here) and several other nice folks I hadn’t met yet.

Jon (I mention him and his choice of… oh no, I promised I wouldn’t mention that anymore) and I also discussed the (still widespread) use of technologies such as EJB2. Jon thought it was amazing that these people aren’t yet using Spring. He perfectly summarizes my thoughts on that at his blog. For me one of the reasons for this is that for those companies it’s too costly (seemingly maybe) to migrate away from those technology because of the initial investment they’ve put in it. Because that initial investment has gone into code which isn’t (entirely) portable across different versions of application servers and more importantly, doesn’t allow for an easy migration path because of inter-twined concerns (more on that later) in their core code. In other words (as Jon puts it) they’ve mixed concerns related to the environment (the infrastructure, whatever you name it) with the actual business logic. Because of this, migration from environment version A to environment version B is quite costly.

This was also a major theme when Arjen and I were discussing the JavaConfig stuff during our session yesterday. Sorry for spoiling the fun Rod (I guess he wanted to talk about this during the conference later this week as well). To me separation of concerns is one of the most important concepts in software engineering that keep your maintenance costs down. With the JavaConfig stuff, we’re hoping to improve on this even further. Using specially designed and coded configuration class, that’s completely separate from your actual business logic (just as your XML files have been since Spring 1.0), we are (well, actually you will be) addressing the two different concerns in question here (configuration and business logic) in two different artifacts. If something needs to change to the configuration you don’t have to mess up your business logic. If your business logic changes slightly, you don’t have to change your configuration. So go on and check this JavaConfig stuff out (I would be writing more about it soon, if it weren’t for the fact that I have a holiday coming up and am still terribly busy with all kinds of other stuff).

In our session, we covered lots of things, but we basically stuck to the basics (Spring DI using XML and Java, Spring AOP, Spring MVC a bit, Groovy integration and bit of integration testing) but apart from that there’s sooo much more cool stuff in the Spring portfolio (I’m almost sounding like I’ve got something to sell here; but I just have this renewed feeling of excitement about what we have to offer ). Spring WebFlow for example is something I expect to be really big in 2007. It’s only been released for short period but already I’m seeing some major corporation here in Holland pick up on themselves. Some of those are of the less progressive kind and would’ve other stuck to Struts if it weren’t for the benefits they already got from Spring. There are some major banks and insurance companies thinking about putting WebFlow to use in their internal and public facing apps. Anyway, I’m digressing…

Back to the talk. I’m attaching the slides [i] I used yesterday as well as the sample application [ii]. I was about to upload the slides I used during the talk yesterday, but somehow can’t create a PDF document of them using PDFCreator so I’ll have to do that some other time maybe. I did upload the sample app already. It is quite small and really should be seen in combination with the talk, but it kind of shows some nice concepts (such as my first tiny attempt to use the JavaConfig stuff in anger–I have to say, it really worked well).

Oh and by the way, did you know Interface21 is hiring in the Netherlands? If you think you can add value to our team and have a decent share of experience with Spring and (modern and classic) J2EE technology and still don’t think you know enough, than you might want to send me your resume. My email is the usual, first name at company name dot com.

[i] the slides I used during my talk yesterday
[ii] the sample application