Re: [RFC/PATCH] mergetool: use resolved conflicts in all the views

[Seth House <seth@eseth.com>]

On Sun, Dec 20, 2020 at 11:34:24PM -0600, Felipe Contreras wrote:
> I disagree. It's fine if you don't want to participate, but the fact
> remains that the position that some tools would want to turn this off
> hasn't been properly defended.

If you are _genuinely_ interested in the answer to this question, please
read the section in my post titled "Conflict Resolution" followed by the
sub-section "Custom Merge Algorithm", and finally "Merge algorithms" [1]
on Wikipedia. Then pretend you want to write your own conflict
resolution algorithm for a new mergetool you've been dreaming up and ask
yourself what versions of the conflicted file your tool will need.

[1] https://en.wikipedia.org/wiki/Merge_(version_control)#Merge_algorithms

Right now the algorithm Git uses is pretty best-in-class so it might
seem unlikely that someone would want to write one of those. However
a whopping *seven* of the tools surveyed do just that. Some of them even
do a pretty good job (I've tried to point those out in the reviews).
You're preoccupied with identifying a specific "adverse effect" but this
debate isn't about that -- it's about giving individual tools the option
to choose how they are used. If people out there want to try and write
a better algorithm than Git, I want to see them try.

That's the point I've been trying to drive home and that's the point
that David also made in his last reply to you.

On that note: you replied to David and said:

> [Y]ou spend your time implementing this on top of my patch. That way
> it's clear who made the mistake.

I plan to start work on exactly that tomorrow. You made the initial
patch so if you'd prefer to take it over the finish line yourself I'll
defer. But if you're not interested then I would be happy to credit you
and finish it.

> Thanks for doing this, but unfortunately one of the most popular isn't
> listed there: vimdiff2.

It's under the Vim section. Use the page search in your browser for
"vimdiff2".

> That's what I expected.

That's what we all expected. The purpose of this follow-up survey was to
validate that expectation.