On Wed, May 13, 2009 at 10:07 AM, Paul MERLIN <[email protected]> wrote:
> The plans for maven 3 are interesting. Some things I noticed and think can > really decrease all this cumbersomeness : > - use of a boolean solver for dependencies resolution > - pom mixins (!!) > - ruby, groovy, python DSLs for writing poms > - tagging in poms I am not sure I want 'programmability' in the poms. I think a the Qi4j approach of letting code generate a static model from which the system is derived is probably suitable for build systems as well. Currently, the scale is "fully declarative" or "on-the-fly" execution, which both have its ups and downs. And instead of some murky middle-way, I think the establishment of a model can be programmatic, but that the model itself is declarative and therefor can be 'visualized', 'added to docs', 'better tool integration' and what not. > If the qi4j builds have reproducibility issues I could give a hand. The problem with these kinds of problems is that the problem is not problematic until the problem occurs. Yes, we have not worried about putting versions on plugins, and I guess that is bad for later. Yet, we don't even know what plugins we are using, other than by doing rm -rf ~/.m2/repository mvn install find ~/.m2/repository/ -name *.jar | egrep "(apache|codehaus)" or something like that. Cheers -- Niclas Hedhman, Software Developer http://www.qi4j.org - New Energy for Java I live here; http://tinyurl.com/2qq9er I work here; http://tinyurl.com/2ymelc I relax here; http://tinyurl.com/2cgsug _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
