Adventures in debugging

Making a quick entry in case I don’t get to later tonight; going out with a friend for his brother’s birthday.

Spent a lot of good time coding today. Trying to get Ant to play nice with JUnit, actually, and then trying to get it to pass my one single test that I’ve created thus far. I ended up turning to Log4J to do some logging within an application. I felt bad about this since I figured it was against someone’s design principles; as it turns out, it’s pretty much accepted that this is OK. Plus, I didn’t really want to learn how Java debugging goes with multiple threads. I read some on PPR about logging, which made me feel better about leaving the logging in, at least. I didn’t have to spend time creating my own logging framework, either — thank goodness. Lets see XP thumb its nose at reuse now.

CLASSPATH is still a bitch. For some reason I had to tell the JUnit task for Ant to fork off a new JVM; otherwise it didn’t find some of my classes, perhaps classes such as my RegularExpressionMessageFilter. BTW, I was shocked that this wasn’t in Log4J already. Especially since ORO is part of the Jakarta project now.

My test now passes, I’ve got some logging information, I can even build from Emacs (though testing is annoying: when there’s an error Emacs goes to build.xml because of the error message). I guess I should check this stuff in and then make some refactoring passes.