[Monotone-devel] Re: 3-way merge considered harmful

[Bruce Stephens]

Timothy Brownawell <address@hidden> writes:

[...]

> Maybe there's a way to not have to choose distant ancestors after
> criss-cross? I'm wondering what happens if you merge using *each* of
> the least common ancestors, and then treat any differences in the
> result as conflicts...

That feels a bit hacky.  I'd guess the clean approach is to adopt a
darcs approach, where the idea is to take the union of the patches in
the ancestors.

Codeville is alleged to have a better solution to this than 3-way
merge, and the rather vague descriptions I've read seem to suggest
they've got something novel.  It's a shame the descriptions aren't
better.