git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Felipe Contreras <felipe.contreras@gmail.com>
To: Junio C Hamano <gitster@pobox.com>,
	Felipe Contreras <felipe.contreras@gmail.com>
Cc: "Elijah Newren" <newren@gmail.com>,
	"Jacob Keller" <jacob.keller@gmail.com>,
	"Git Mailing List" <git@vger.kernel.org>,
	"Vít Ondruch" <vondruch@redhat.com>,
	"Alex Henrie" <alexhenrie24@gmail.com>,
	"Jeff King" <peff@peff.net>,
	"Philip Oakley" <philipoakley@iee.email>,
	"Theodore Y . Ts'o" <tytso@mit.edu>,
	"John Keeping" <john@keeping.me.uk>,
	"Richard Hansen" <rhansen@rhansen.org>,
	"Brian M. Carlson" <sandals@crustytoothpaste.net>,
	"W. Trevor King" <wking@tremily.us>
Subject: Re: [PATCH v2 02/14] pull: improve default warning
Date: Mon, 14 Dec 2020 20:36:37 -0600	[thread overview]
Message-ID: <5fd8213598c33_d7c4820837@natae.notmuch> (raw)
In-Reply-To: <xmqqzh2g9e0r.fsf@gitster.c.googlers.com>

Junio C Hamano wrote:
> Felipe Contreras <felipe.contreras@gmail.com> writes:
> 
> >> We do not have to or implement both, but either should give us the
> >> "when pull sees a non-ff history, it should stop without merging or
> >> rebasing, and the user won't be given the advice on how to choose
> >> between merge and rebase" behaviour, I would think.
> >
> > Right, so both should error out.
> 
> One of them, not both; the one we choose to implement would make it
> fail, I would think.

If only one of them fail, then it's impossible to tell the user:

  If you want to try the new behavior, type this command:

    git config $something

We have to go straight for the error, with no deprecation period.

> > And what should these do?
> >
> >   git -c pull.rebase=no -c pull.ff=only pull --merge
> 
> There is no --merge, so let's reread with s/--merge/--no-rebase/;

Back in 2013 everyone found it easier to speak in terms of --merge, even
you, BTW, who was the first one to send a patch with --merge [1], after
Linus suggested it [2].

> the command line would be saying "I only want to fast-forward, or I
> want the command to fail".  The advice code should not trigger
> (i.e. we gave an explicit preference to merge and not to rebase, so
> rebase_unspecified would be false), but the configured preference
> that says "we take only fast-forward merges" will still be in effect.
> If the history is fast-forward, the merge backend will happily
> advance our history.  If not, the merge backend will happily fail
> the pull.
> 
> If you want to countermand the configured preference from the
> command line and allow a merge to be done when non-ff history is
> given, what you'd need to override is not --merge/--no-rebase.  The
> configuration pull.rebase=no already says you do not want rebase and
> you want merge).  You want to override --ff-only, so I'd expect
> "pull --ff" (or "pull --no-ff") to be how you override your
> configured preference and merge in a non-ff history, either by fast
> forwarding or creating an extra merge commit.

Precisely, so you can't tell the user:

  If you want to try the new behavior, type this command:

    git config pull.ff only

Because, after the user gets the error:

  Not a fast-forward, either merge or rebase.

She cannot do:

  git pull --merge

Which is the best idiom to specify she wants a merge.

The command will still fail, even after the user has specified she wants
the new mode, and that she wants to merge in this perticular case.

> >   git -c advice.pullnonff=false pull --merge
> 
> Again with s/--merge/--no-rebase/; but that is showing the
> preference not to rebase and to merge from the command line, so
> shouldn't it just go ahead and merge without any advice?

Yes, but *before* we make this the default, what configuration get us
there today?

> > I'm going to answer because I think it's obvious what you would expect:
> > if you pass --merge, both should succeed.
> >
> > Except they won't, because "git pull --ff-only --merge" fails.
> >
> > Correct?
> 
> See above.

That's a "correct". There's no configuration today that gives us both:

 1. fail: git -c $config pull
 2. pass: git -c $config pull --merge

Like this does:

 1. fail: git -c pull.mode=fast-forward pull
 2. pass: git -c pull.mode=fast-forward pull --merge

Cheers.

[1] https://lore.kernel.org/git/7v4ncjs5az.fsf_-_@alter.siamese.dyndns.org/
[2] https://lore.kernel.org/git/CA+55aFz2Uvq4vmyjJPao5tS-uuVvKm6mbP7Uz8sdq1VMxMGJCw@mail.gmail.com/

-- 
Felipe Contreras

  reply	other threads:[~2020-12-15  2:37 UTC|newest]

Thread overview: 79+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-04  6:16 [PATCH v2 00/14] pull: default warning improvements Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 01/14] doc: pull: explain what is a fast-forward Felipe Contreras
2020-12-04 22:55   ` Elijah Newren
2020-12-05  1:21   ` Jacob Keller
2020-12-04  6:16 ` [PATCH v2 02/14] pull: improve default warning Felipe Contreras
2020-12-04 22:59   ` Elijah Newren
2020-12-05  0:12     ` Felipe Contreras
2020-12-05  0:56       ` Elijah Newren
2020-12-05  1:56         ` Felipe Contreras
2020-12-05  9:23           ` Chris Torek
2020-12-05 11:47             ` Felipe Contreras
2020-12-05 16:28           ` Elijah Newren
2020-12-05 21:27             ` Felipe Contreras
2020-12-06  1:01               ` Elijah Newren
2020-12-06 14:31                 ` Felipe Contreras
2020-12-07  7:26                 ` Junio C Hamano
2020-12-07  9:16                   ` Felipe Contreras
2020-12-07 18:16                     ` Junio C Hamano
2020-12-07 19:09                       ` Felipe Contreras
2020-12-07 19:53                         ` Junio C Hamano
2020-12-07 22:14                           ` Felipe Contreras
2020-12-07 23:30                           ` Jacob Keller
2020-12-08  2:23                             ` Junio C Hamano
2020-12-08  3:15                               ` Felipe Contreras
2020-12-08 20:16                                 ` Junio C Hamano
2020-12-09  9:52                                   ` Felipe Contreras
2020-12-09 19:05                                     ` Elijah Newren
2020-12-10  2:39                                       ` Felipe Contreras
2020-12-10  6:45                                       ` Junio C Hamano
2020-12-10  9:08                                         ` Felipe Contreras
2020-12-10 10:01                                           ` Felipe Contreras
2020-12-11  7:17                                           ` Junio C Hamano
2020-12-11 11:33                                             ` Felipe Contreras
2020-12-14 21:04                                               ` Junio C Hamano
2020-12-14 22:30                                                 ` Felipe Contreras
2020-12-14 23:14                                                   ` Junio C Hamano
2020-12-15  2:36                                                     ` Felipe Contreras [this message]
2020-12-15  2:31                                             ` Jeff King
2020-12-15  3:49                                               ` Felipe Contreras
2020-12-15 11:18                                               ` Junio C Hamano
2020-12-15 12:53                                                 ` Felipe Contreras
2020-12-07  9:23         ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 03/14] pull: refactor fast-forward check Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 04/14] pull: cleanup autostash check Felipe Contreras
2020-12-04 23:07   ` Elijah Newren
2020-12-05  0:47     ` Felipe Contreras
2020-12-05  0:57       ` Elijah Newren
2020-12-04  6:16 ` [PATCH v2 05/14] pull: trivial cleanup Felipe Contreras
2020-12-04 23:09   ` Elijah Newren
2020-12-05  0:48     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 06/14] pull: move default warning Felipe Contreras
2020-12-04 23:18   ` Elijah Newren
2020-12-04 23:36     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 07/14] pull: display default warning only when non-ff Felipe Contreras
2020-12-04 23:24   ` Elijah Newren
2020-12-05  1:03     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 08/14] pull: trivial whitespace style fix Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 09/14] pull: introduce --merge option Felipe Contreras
2020-12-04 23:27   ` Elijah Newren
2020-12-05  1:06     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 10/14] pull: add proper error with --ff-only Felipe Contreras
2020-12-04 23:34   ` Elijah Newren
2020-12-05  1:18     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 11/14] tentative: pull: change the semantics of --ff-only Felipe Contreras
2020-12-04 23:39   ` Elijah Newren
2020-12-05  4:01     ` Felipe Contreras
2020-12-05  4:06       ` [PATCH] experiment: pull: change --ff-only and default mode Felipe Contreras
2020-12-05 17:29         ` Elijah Newren
2020-12-05 18:16           ` Felipe Contreras
2020-12-07  7:34           ` Junio C Hamano
2020-12-05 11:37       ` [PATCH v2 11/14] tentative: pull: change the semantics of --ff-only Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 12/14] pull: show warning with --ff Felipe Contreras
2020-12-04 23:41   ` Elijah Newren
2020-12-05  1:25     ` Felipe Contreras
2020-12-04  6:16 ` [PATCH v2 13/14] test: merge-pull-config: trivial cleanup Felipe Contreras
2020-12-04 23:41   ` Elijah Newren
2020-12-04  6:16 ` [PATCH v2 14/14] test: pull-options: revert unnecessary changes Felipe Contreras
2020-12-04 23:49   ` Elijah Newren
2020-12-05  1:28     ` Felipe Contreras

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5fd8213598c33_d7c4820837@natae.notmuch \
    --to=felipe.contreras@gmail.com \
    --cc=alexhenrie24@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jacob.keller@gmail.com \
    --cc=john@keeping.me.uk \
    --cc=newren@gmail.com \
    --cc=peff@peff.net \
    --cc=philipoakley@iee.email \
    --cc=rhansen@rhansen.org \
    --cc=sandals@crustytoothpaste.net \
    --cc=tytso@mit.edu \
    --cc=vondruch@redhat.com \
    --cc=wking@tremily.us \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).