> Now you install a version into your > repository that is not released yet. There's a reason why > Maven checks out the tagged sources again. What happens if > release:perform fails? Then you have polluted your local repo > with a version that is not existing. Even worse, even if > somebody is later on performing the release, you will not get > that version ever again unless you remove it manually from > your local repo.
How is this any different than doing an install of a SNAPSHOT and the build fails somewhere? You now a "polluted" local repo. Or even better, what happens if the release:perform fails. You now have partially released modules on your maven repo. I suppose the bottom line is, yes things can go wrong. I truly admire the goal of making a build system completely bullet proof. But is this goal realistic? I would love it if building projects with the release plugin were this way, but I just don't think that is going to happen. Getting back to the issue at hand though, what is the proper way in Maven to do a build with the release plugin when you have dependencies between modules? Are you supposed to do something with the dependency plugin to ensure that dependencies get properly resolved during the prepare? --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
