On Wed, Mar 3, 2021 at 11:32 AM Ruediger Pluem <rpl...@apache.org> wrote: > > All good. I guess it would have lowered my confusion if there would have been > a pointer in STATUS that this backports > stuff needed for 2.4.x from r1836237 as well.
Yeah, I'm a bit torn on this, removing the STATUS entry in the commit itself helps to figure out commit/vote relation, on the other hand it's likely to create conflicts for a cherry-pick. I guess distros/cherry-pickers can easily resolve such non-functional conflict though. As a matter of fact each backporter has their own logic/habits here, some of us concatenate the original commit messages while others use the STATUS entry as message, some remove the STATUS entry at the same time while others use a separate commit, plus a combination of all that of course. Maybe we should have a discussion on what's the best in this regard for the contributors/community, and share/evolve our scripts from time to time ;) Regards; Yann.