Leaving Maven

Up until now I have tried maven a few times and I am convinced, the best way to push maven to work is to drop it. The deceptive simplicity of it's initial usage leads into continuous maintenance issues fairly quick. I made some more comments on this in an earlier article. The web is full of statements supporting this opinion. Just try to google for "maven hate" for example. Or take this citation of Matthieu Riou on the buidr homepage:

"The biggest mistake afterward was to migrate to Maven2. I could write pages of rants explaining all the problems we ran into and we still ended up with thousands of lines of XML."

Another drastic example may be found here. Unfortunately I do not have an evaluated recommendation for a different tool. Nevertheless buildr seems to be an interesting concept. They extend the popular ruby build tool rake with several features specific to the java world including support for other upcoming languages with increasing popularity as for example groovy or scala. Most interesting for people completely annoyed by maven is builder's support for dependency management and it's access to maven repositories. And finally there is a tempting reason, why I will give it a try: Buildr seems to have some useful testing features. There is support for

• global test setup and teardown outside of the respective test/unit framework (useful for example to start a JEE server only once during a test run)
• checking the distributables (jars, wars, ears, ...) in order to not to forget license files, manuals or other artifacts a developer usually is minor aware of.

I then will share my experiences with you : -) .

(A hint for the scala people: If I got it right, you're switching from sbaz to maven. Is this really necessary???)