Re: [PATCH v3 0/4] git add --resolved
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Aug 1, 2026, 21:45 UTC
- Message-ID
- <xmqqjyq97bx8.fsf@gitster.g>
- In-Reply-To
- <xmqqv79t9101.fsf@gitster.g>
Junio C Hamano <gitster@pobox.com> writes:
Show 13 quoted lines
> Jeff King <peff@peff.net> writes: > >> That doesn't seem too bad to me. I was mostly worried about (1), but >> after digging into it, I think I've convinced myself that it's a >> non-issue. > > Yeah, our merges that began with "read-tree -m -u" have never merged > into a path with local modification in it, so I have blindly > assuming that (1) is a given. I think "ort" did not change it, but > quite honestly I do not know where the more recent invention that > renames paths that existed only on one side to follow adjacent paths > that were renamed from the same directory guarantees property (1), > so in these new code paths my assumption truly is a blind faith X-<.
So I got curious, and found out that t6423 has a test labelled as "Testcase 11b" that covers this exact case. So it is a bit better than blind faith now ;-).