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

[Timothy Brownawell]

On 5/1/05, Nathaniel Smith <address@hidden> wrote:
[...]
> Here's another pathological case for 3-way merge:
>    A
>    |
>    B
>   / \
>  C   D
> [...]
> So, clearly, picking A as an ancestor in the above graph is terrible.
> 
> This leaves us in a bad situation, though.  The possibility of
> criss-cross merge means that we are sometimes (or often) forced to
> choose ancestors that are far away, ...

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...

Tim