Hitting a brick wall, aka receiving a minus one ;)

Well it didn't take too long after publishing a Release Candidate for Apache Karaf 2.1.0 that a community member discovered a flaw that we had believed to be resolved. A quick patch submission followed, so we are canceling the current RC and re-setting for another run as per our release guide.

So what does this entail?

1. Delete the current Apache Karaf 2.1.0 tag from svn.
2. Drop the RC from Nexus staging area.
3. Rollback the pom versions from 2.1.1-SNAPSHOT to their last version.
4. Reviewing and applying supplied patch for the reported issue.
5. Returning to the start of the release process.

This is the heart of the Apache release process, ensuring that only vetted releases are published to end users. I decided to post about this occurrence since many people have asked me before about how we handle situation like these. It's a lot of work but quality is among our top priorities.