Hi all, Just to clarify what is happening here, based on my reading of the discussion here thus far, and in #kde-devel earlier today.
I concur with the majority that a unified tool, similar to Gitlab would be an excellent improvement on our current infrastructure. Unfortunately we found that at one point Gitlab contained code to email debug data to a given address. This is an unacceptable security issue. While it has since been removed as far as we are aware, it makes it impossible to trust the software, particularly considering we are trusting it with the canonical copy of our Git repositories. It also has issues with groups (requiring us to keep them in sync manually), has undesirable behaviour when moving repositories (sends an email to every member of the group - so 2000+) and requires us to hack the software in order to integrate our own hooks. In regards to Phabricator - we have examined this in the past as a replacement for projects.kde.org. While generally impressed with it's performance, it still lacked the per-revision code review tools that people are after which Gerrit provides. It also relied on special capitalised names for each repository which are supposed to be a shortcut to them. Considering we have more than 600 repositories, we would need to use their full names here - which would lead to fairly ugly URLs. We have yet to have time to evaluate Gogs.io. In regards to why we are permitting Gerrit to be evaluated - it is primarily to allow for the community to come to a decision. The complexity of the user interface among other issues are still problems which sysadmin believes could block it's overall adoption. I had hoped that independent projects, rather than Frameworks or anything along those lines would be the test subjects in this case though. Thanks, Ben
