Re: What's cooking in git.git (Dec 2018, #02; Fri, 28)

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Jan 3, 2019, 13:27 UTC
Message-ID: <nycvar.QRO.7.76.6.1901031424340.45@tvgsbejvaqbjf.bet>
In-Reply-To: <CABPp-BEd5-0Vcv8YApUxo0jK_ofxCORSG5H0wU=kiR2aOY1ztQ@mail.gmail.com>

Hi Elijah,

On Fri, 28 Dec 2018, Elijah Newren wrote:

Show 17 quoted lines

> On Fri, Dec 28, 2018 at 10:04 AM Junio C Hamano <gitster@pobox.com> wrote:
> 
> > * en/rebase-merge-on-sequencer (2018-11-08) 2 commits
> >  - rebase: implement --merge via git-rebase--interactive
> >  - git-rebase, sequencer: extend --quiet option for the interactive machinery
> >
> >  "git rebase --merge" as been reimplemented by reusing the internal
> >  machinery used for "git rebase -i".
> >
> >  Expecting a reroll.
> >  cf. <CABPp-BF8RupyfP69iqAVTXxEhBGyzVd-wUgp3y0pf+CbBFAQeg@mail.gmail.com>
> 
> Quick update: Two re-rolls have been sent in[1]; v3 on November 22 and
> v4 with only a minor error message tweak on Dec 11.  I think I've
> addressed all review comments from v2, but neither v3 nor v4 has
> received much review -- Dscho was also heavily busy during the run up
> to 2.20 and needed some recovery time afterward.

Yep. There have been quite a few problems in the -rc period, and at least one frantic bug fix of mine introduced another regression, and then there was the problem with cURL where it would try to use HTTP/2 with NTLM (which does not work, and probably never will) and as you probably suspect, NTLM/Kerberos authentication is *quite* common on Windows, so that would have been a total non-starter if we had shipped Git for Windows v2.20.0 with an unfixed cURL.

So yes, I was quite exhausted after those weeks.

> I was going to re-ping in early January.  Anyway, it may be worth at
> least updating your note to "reroll exists".

It is early January! ;-)

Ciao, Dscho