Ian Jackson writes ("Bug#1116600: git-debrebase: new-upstream failure"):
> Soemthing disturbing is going on. The behaviour seems stochastic.
> Sometimes it prints
>
> git-debrebase: Merge resolution failed: interchange merge-base
> (bb4d6750816e39569c84a2b52a950b16fe85f95a)'s breakwater
> (00845b43953b760a13d225a694c7589d145c0c11) != breakwaters' merge-base
> (26c3c9b853355117608ce06de33290246d9a8e64)
>
> But more often the message David quotes.
During its attempts to figure out what to do about the merge, it seems
to be pre-emptively rewriting the inputs to the merge. The difference
in timestamps is part of the explanation for the difference in
behaviour.
David, how much of a hurry are you in? I will continue to
investigate, but the algorithms are decidedly nontrivial. If you
like, I can probaly sort it out by hand somehow and present you with a
working branch. But if you can wait, that would save me that work.
Thanks,
Ian.
--
Ian Jackson <[email protected]> These opinions are my own.
Pronouns: they/he. If I emailed you from @fyvzl.net or @evade.org.uk,
that is a private address which bypasses my fierce spamfilter.
