Re: [PATCH 1/6] Revert "Revert "Merge branch 'ra/rebase-i-more-options'""

[Junio C Hamano <gitster@pobox.com>]

Junio C Hamano <gitster@pobox.com> writes:

> Elijah Newren <newren@gmail.com> writes:
>
>> On Tue, Apr 7, 2020 at 7:11 AM Phillip Wood <phillip.wood123@gmail.com> wrote:
>>>
>>> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>>>
>>> This reverts commit 4d924528d8bfe947abfc54ee9bd3892ab509c8cd.
>>>
>>> This is being reverted to enable some fixups for
>>> ra/rebase-i-more-options to be built on this commit.
>>
>> This makes sense to me, but it will be only the second 'Revert
>> "Revert..."' commit in all of git.git and I'm curious if Junio will be
>> unhappy with it.
>
> Nah, there isn't much to become unhappy about.  
>
> I however suspect that the alternative would certainly be much nicer
> and easier to understand, which is to rebuild the 7-patch series
> c58ae96fc4..d82dfa7f5b but bugs already fixed, instead of doing this
> patch to take us back to a known buggy state and then fix the result
> with 5 more patches.  Is that what you meant?

After looking at the conflict resolution while merging the result of
applying these patches on top of the older codebase, I would have to
say that the approach """I've opted to add some cleanup commits on
top of Rohit's work rather than reworking his patches.""" may not
have been particularly a brilliant idea, not because the conflicts
arising from an older codebase are unpleasant to resolve (they seem
to be reasonably contained), but because it resurrects other
unwanted cruft we have cleaned up since then, and worse yet, it does
so without triggering conflicts.  For example, we'll end up seeing
mentions of "'am' backend", which have all been updated to "'apply'
backend", in the documentation, and patches [2-6/6] do not fix them.

[5/6] is an example of one more "unwanted" thing the reversion
resurrects that needed to be fixed, I guess?

The result of applying all these patches and merging it to 'master'
and/or 'pu' may be more or less right, as far as the new features
added to the "rebase -i" by the series are concerned but there may
be many small "unwanted cruft" we may be resurrecting with [1/6],
so...