What's mentioned as drawbacks are powerful features Maven provides to handle the project/plugin dependencies better. It takes a bit of understanding and experience to appreciate these features and to use them effectively. I've used Maven the right way in one of the biggest enterprise environments as well as a couple of startups and the results have been nothing short of phenomenal. Am still baffled by how most of the expert Java/JEE programmers have very little knowledge about how Maven works.
I think there are two reasons Maven adoption is slower than it maybe ought to be:
1. It's actually pretty darned complicated to use 'correctly' for a non-trivial project.
2. It approaches impossible to integrate into an existing non-trivial project. Maven wants you to do thing it's way (that's the point).. but that makes moving to it on an existing project a challenge
I have had to work with maven and integrating it to an existing project is insanely hard if not impossible. On other projects I worked with Maven it spends more time updating itself than doing any useful work.
It takes a bit of understanding and experience to appreciate these features and to use them effectively.
That would be the reason why most expert JEE folks aren't getting Maven. It's not a short or easy learning curve. even on a good day with good folks to help you wade through it.
That said, Maven's dependency management engine alone is worth the price of admission. Bye-bye JAR hell. But the integrated build/release/tag/version features are the real treat once you get that far...
Yes. Thats why companies that can afford have separate build engineers. Maybe one shouldn't be called an expert if they have a poor understanding of how their build system works? My first couple of interview questions for Senior JEE engineers always revolves around their understanding of build and I've found it to be a useful parameter to validate their claims of being an expert.
