building hibernate or OSS management

problems:

• need for a maven version > 2.8 the current on is 2.7 (the silly maven enforcer plugin)
• the antlr maven plugin (under project mojo at codehaus) (this plugin is not available from codehaus repository)

Steps taken to resolve:

• the snapshot version in maven svn repository is 2.1 we could get it, build this version, install it then go on.
• get antlr sources from repository, build-it, oh no! it shocked on modello maven plugin ([WARNING] Failed to load plugin descriptor for: Plugin [org.codehaus.modello:modello-maven-plugin]. Assigning this plugin to be resolved again just prior to its execution. NOTE, This may affect assignment of the mojo: java if its default phase (given in the plugin descriptor) is used. Check debug output (-X) for more information.
• set codehaus repository in settings.xml

• rebuild the plugin despite the annoying errors displayed we have a jar and gonna install it manualy just to keep the things running...
• mvn install:install-file file=./target/antlr-maven-plugin-2.0-SNAPSHOT.jar pomFile=pom.xml now what?

1 required artifact is missing.

for artifact:
org.codehaus.modello:modello-maven-plugin:maven-plugin:1.0-alpha-8

I've been successfully (with some hacks) resolving this mess some nights later but it's Already 6:27 am and am giving up for tonight.

Really Hibernate it self has nothing to do with it the problem being the mojo project @ codehaus with all the stalled dependencies and missing artifacts/version, along with maven project with some minor responsibility (at least the snapshot builds) here for keeping the enforcer plugin (please make it optional and give some way to manually turn it off) despite the complains about it.
Folks @ mojo should realize that many other opensource projects are using their stuff. The source code availability the ability to hack into it and build it are the essence of opensource. Creating blocking problems wont help this movement grow.
The devs of antlr-maven-plugin maybe using his local repository with those missing versions deployed. Should they coordinate their releases with other projects (plexus, modello etc..)?
Should they better communicate with codehaust admins so the old stalled dependencies are kept on the server till no one uses them?
Should codehaust admins check their logs to see if some file is still in use and so keep them on the server?
We are all thankful for those opensource pioneer and admiring their work but opensource projects are a responsibility toward the community and require more manageability to keep the noble principles believable.
I still remember the sad and painful experience we've got with the Apache Avalon framework disappearing(or split in many other projects) and how we've got to re-architect the long term in-house developed framework to no longer relay on it.