From my perspective, what happened here was that we had a read-only mirror of the SVN repo and then somebody additionally performed a manual svn-to-git migration to a writable repo.
Looking at the oldest and recent history, one can see that the commits in the two repos mirror each other - but with different commit IDs. The reason for this is most likely that the author information in the repositories differs. One has proper email addresses and names while the other has only svn usernames. Unfortunately, it seems to be the "uima-as_old" that has the proper names... The histories run in parallel until commits c6fc777093d5b24e04f436c29785957375f267f6 15. January 2020 at 20:19:36 CET c6faa8b6fd8b4dd914e0fa6ded4ce330462f7236 15. January 2020 at 20:19:36 CET After those, commits continue only in the uima-async-scaleout repo. So much for my analysis. On the bottom line: [x] +1 - irrevocably delete https://github.com/apache/uima-as_old -- Richard
