I don’t understand why Groovy doesn’t steal JRuby’s native code for doing command line calls.

While im seeing how useful this could be(indeed), it worries me how this functionality could lead to a disorganized dependency spider web spread all over the place when having a medium sized project (with few developers) and not having an specific place to set all dependencies at once.

IMHO we may need to just educate the team where to place them and not leave the developers going wild declaring @Grapes all over the place.

So I take it that groovy Grapes is compatible with typical Maven-related artifact repositories, such as Sonatype Nexus or JFrog Artifactory?

Does that imply then that there is a one-time configuration of some setup file for the Groovy install so that remote artifact repositories can be referenced? Does it happen to also work with a local Maven repository as stored under the user’s .m2 directory subtree? (Such Maven friendliness for dependency management would be truly awesome.)

The simple answer is Grape will automatically download the dependencies for you and store them in your ~/.groovy/grapes directory. If you were to run the above code your grapes directory would consist of two directories, javax.activation and javax.mail, and the dependent jar’s would be stored there. Once the dependent jars are downloaded, Grape will automatically add them to the runtime classpath since it knows where the jar files are located.

The more complex answer is you can configure Grape to leverage existing dependency management tools such as Maven by updating your ~/.groovy/grapeConfig.xml to point to your existing repository. Details on how to configure Grape to work in this fashion can be found here.

I guess im not clear on how that works.

Thanks!