Re: [PATCH v2 02/18] Add a new builtin: branch-diff

From: Jeff King <peff@peff.net>
Date: May 21, 2018, 20:24 UTC
Message-ID: <20180521202414.GA14250@sigill.intra.peff.net>
In-Reply-To: <CAGZ79kYcWuVorfk7eYjUuLi2XeMS8sPrJYE0OQmgiQi2NkuDZA@mail.gmail.com>

On Mon, May 21, 2018 at 10:56:47AM -0700, Stefan Beller wrote:

Show 22 quoted lines

> > It is very much about
> > comparing two *ranges of* revisions, and not just any ranges, no. Those
> > ranges need to be so related as to contain mostly identical changes.
> 
> range-diff, eh?
> 
> > Most fellow German software engineers (who seem to have a knack for
> > idiotically long variable/function names) would now probably suggest:
> >
> >         git compare-patch-series-with-revised-patch-series
> 
> or short:
> 
>   revision-compare
>   compare-revs
>   com-revs
> 
>   revised-diff
>   revise-diff
>   revised-compare
> 
>   diff-revise

I still like "range diff", but I think something around "revise" is a good line of thought, too. Because it implies that we expect the two ranges to be composed of almost-the-same commits.

That leads to another use case where I think focusing on topic branches (or even branches at all) would be a misnomer. Imagine I cherry-pick a bunch of commits with:

  git cherry-pick -10 $old_commit

I might then want to see how the result differs with something like:

  git range-diff $old_commit~10..$old_commit HEAD~10..HEAD

I wouldn't think of this as a topic-branch operation, but just as comparing two sequences of commits. I guess "revise" isn't strictly accurate here either, as I'm not revising. But I do assume the two ranges share some kind of mapping of patches.

-Peff

PS I wish there were a nicer syntax to do that. Perhaps
   "git range-diff -10 $old_commit HEAD" could work, though occasionally
   the two ranges are not the same length (e.g., if you ended up
   skipping one of the cherry-picked commits). Anyway, those kind of
   niceties can easily come later on top. :)