Re: jc/fix-pull-ff-only-when-already-up-to-date, was Re: What's cooking in git.git (Nov 2021, #03; Tue, 9)

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Nov 10, 2021, 19:09 UTC
Message-ID: <nycvar.QRO.7.76.6.2111102000250.21127@tvgsbejvaqbjf.bet>
In-Reply-To: <xmqqpmr8x6zn.fsf@gitster.g>

Hi Junio,

On Wed, 10 Nov 2021, Junio C Hamano wrote:

Show 23 quoted lines

> Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:
>
> > On Tue, 9 Nov 2021, Junio C Hamano wrote:
> >
> >> * jc/fix-pull-ff-only-when-already-up-to-date (2021-10-29) 1 commit
> >>   (merged to 'next' on 2021-10-29 at ad4753e668)
> >>  + pull: --ff-only should make it a noop when already-up-to-date
> >>
> >>  "git pull --ff-only" and "git pull --rebase --ff-only" should make
> >>  it a no-op to attempt pulling from a remote that is behind us, but
> >>  instead the command errored out by saying it was impossible to
> >>  fast-forward, which may technically be true, but not a useful thing
> >>  to diagnose as an error.  This has been corrected.
> >>
> >>  Will cook in 'next'.
> >
> > I would like to register a vote for advancing this into v2.34.0,
> > still.
>
> I didn't recall I did it as a regression fix, but as a fix for
> relatively old change that is already in released version(s) in the
> field.  If the breakage is new between 2.33.0 and 'master', I think
> it is a good idea to merge it down.

I can say with conviction that this is a very recent change of behavior, as I needed this quick-fix on Oct 30: https://github.com/git-for-windows/build-extra/commit/fb311a97cf82243ea1fe4036f1b180f5a3e6bc7f

I don't recall the particular details how this fix came about, as it was the day of -rc0 (and I don't remember whether I needed this _before_ or _after_ Git for Windows -rc0 was released). All I remember is that an automated build broke because it created a local commit and then expected `git pull --ff-only` to succeed (because the remote was not ahead, but it necessarily was behind the local HEAD).

The change of behavior might have been introduced by v2.33.1, or by v2.34.0-rc0.

So yeah, I would argue that it is a regression introduced in this cycle.

Ciao, Dscho