| [gmf-dev] About the unexpected flattening of GMF Runtime repo |
|
Hi Anthony, hi all, I can see that you decided to flatten the GMF Runtime repository without taking into account opinions from other contributors. Despite the -1 from contributors (me) and the only active committer in GMF Runtime (Aurelien), you forced this change. I think it was a bad move for the following reason: 1. it was not necessary to do this now. 2. Aurelien was working on moving to Tycho with interesting result. You refactoring simply broke everything, so the CI job has not run for 2 weeks now. 3. You'll probably see later that keeping a hierarchy in Tycho allows for example to put test-specific parameters to the tests/pom.xml and then simplify the parent pom. It has some technical advantages. Also you made this change on CVS rather than doing it on Git, so it seems impossible to revert it now. Such a big change has to be part of a discussion, and has to be planned with support from every one involved. Instead I would have suggested the following: 1. Move to Git 2. Suggest your refactoring in a Bugzilla or a Gerrit contribution, with build fixed 3. Merge it if nobody put a -1 on it. I really see this as a bad way to lead the project. You just added technical debt by breaking the job and Aurelien's work, and you imposed your opinion. That's not a collaborative approach. Cheers, |