Git Mailing List Archive on lore.kernel.org
 help / color / Atom feed
From: Elijah Newren <newren@gmail.com>
To: Jonathan Nieder <jrnieder@gmail.com>
Cc: "Johannes Schindelin" <Johannes.Schindelin@gmx.de>,
	"Phillip Wood" <phillip.wood@dunelm.org.uk>,
	"Elijah Newren via GitGitGadget" <gitgitgadget@gmail.com>,
	"Git Mailing List" <git@vger.kernel.org>,
	"Denton Liu" <liu.denton@gmail.com>,
	"Junio C Hamano" <gitster@pobox.com>,
	"Pavel Roskin" <plroskin@gmail.com>,
	"Alban Gruin" <alban.gruin@gmail.com>,
	"SZEDER Gábor" <szeder.dev@gmail.com>
Subject: Re: [PATCH v3 15/15] rebase: change the default backend from "am" to "merge"
Date: Thu, 16 Jan 2020 15:19:55 -0800
Message-ID: <CABPp-BE9XqOCG0VqA5XgDbXm4apw0-_mOqW-JSO+PDZhisO0Rw@mail.gmail.com> (raw)
In-Reply-To: <20200116223912.GB71299@google.com>

On Thu, Jan 16, 2020 at 2:39 PM Jonathan Nieder <jrnieder@gmail.com> wrote:
> [...]
> >>  B. stop running the post-commit hook in rebase
> >
> > AND sometimes in both cherry-pick and git-revert, depending on the
> > number of commit(s) picked/reverted (see below).
>
> Interesting.  Doesn't "git cherry-pick A..B" run post-commit after
> each commit today?  It would be possible to do (B) and retain that
> behavior.

Sure, it'd be possible to retain the post-commit calls for cherry-pick
but should we?  Isn't the primary reason the issue came up that having
batch-commit operations calling post-commit for each intermediate
commit is ugly/unexpected overhead?  If we're ripping it out of rebase
for that reason, it seems that cherry-pick should get the same
treatment.  I don't want to just move the inconsistencies around, I
want to fix them.

> > Also, even if we go this route, I think the post-commit hook should be
> > added to git-merge whenever it creates a merge commit.
>
> Agreed.

Yay!

> >>     Advantages:
> >>     * makes --am and --merge consistent with minimal user impact
> >
> > ...and also makes --interactive and --exec and other cases consistent too.
>
> Hm.  My initial thought with (B) was to stop running post-commit in
> non-interactive rebase but keep running it in rebase --interactive.
>
> I don't have a strong opinion about this, though.

I'm okay with treating them differently if there's a defining reason
why, but otherwise I think they should all behave the same.  The only
reasons I've seen for differences all fall in category C so far, which
gives me a strong opinion against treating them differently.

> >>     Disadvantages:
> >>     * meaning of the post-commit hook remains a muddle
> >
> > Why?  "Commands which create no more than one commit (git-commit,
> > git-merge, maybe single-commit git-revert or git-cherry-pick) call
> > post-commit, commands which create several commits derived from others
> > (git-am, git-rebase, git-cherry-pick, sometimes git-revert ) call
> > post-rewrite instead for performance reasons"
>
> Example: if I run "git revert A..B", following that rule we'd want to
> call post-rewrite instead of post-commit.  But post-rewrite gets input
>
>         <old-sha1> SP <new-sha1> [ SP <extra-info> ] LF
>
> so that it can be used to copy data such as notes.  What do I pass in as
> <old-sha1>?
>
> Relatedly, if I remember correctly, we don't call post-rewrite in "git
> cherry-pick" because cherry-pick is considered to represent a *new*
> commit instead of a modified version of the old one.  In C++ terms,
> it's a copy constructor instead of a move constructor. ;-)  See
> notes.rewrite.* in "git help notes" for more on that subject.

Ah, I see.  Yeah, there's some impedance mismatch between the
purpose/definition of the post-rewrite hook and what I think is needed
for B.  Perhaps create a new post-batch-commit hook
("batch-post-commit"?) and have git-rebase, git-am, git-cherry-pick,
git-revert, etc. call it while git-commit and git-merge call
post-commit.  (Or maybe even make git-commit and git-merge also call
post-batch-commit/batch-post-commit but do so with just one commit --
and just note that post-commit's design of not taking any parameters
was a mistake and any place that happens to call it beyond
"git-commit/git-merge" is solely by accident of past implementation
and is not guaranteed to continue.)

I think the realm of "copy constructor vs. move constructor" may make
sense for post-rewrite-hook, but I'm worried that trying to apply that
logic to notification of new commits may lead us into category C
craziness.

> [...]
> >>  C. stop running the post-commit hook in rebase --merge or commit --amend
> >>
> >>     Advantages:
> >>     * produces a consistent definition of post-commit ("it's about new
> >>       commits only")
> >
> > But cherry-pick never creates "new" commits if commit --amend doesn't,
> > so we'd also want to turn the post-commit hook off for cherry-pick,
> > right?
>
> With the copy/modify distinction described above, "git commit --amend"
> acts as a "modify" operation in a user's workflow, while "git
> cherry-pick" acts as a "copy" operation.
>
> That means that with this definition, post-commit is appropriate to
> run in cherry-pick but not in commit --amend.

What if there are conflicts?  Because with conflicts, aren't
cherry-picks more like the 'commit --amend' case since they behave as
a "modify"?  Conflict-dependent invoking of the post-commit hook?

> This matches the distinction used by note rewriting and the
> post-rewrite hook.  Looking back at the discussion in
> https://lore.kernel.org/git/cover.1266164150.git.trast@student.ethz.ch/
> I don't see this covered, so it's possible I'm remembering wrong (a
> pointer to the motivating discussion that inspired that patch series
> would be welcome).  In the context of note copying, this distinction
> ended up being a good decision in practice in most workflows I've
> seen.

Yeah, for note copying that distinction makes a lot of sense to me.
For "notification of new commits", it doesn't seem to jive at all for
me.

> [...]
> >>     Disadvantages:
> >>     * the change to historical "git commit --amend" behavior is likely
> >>       to be surprising
> >
> > That is certainly one issue, but I'd say this case is riddled with
> > problems and is the most muddled of all.
>
> I'm inclined to agree, based on the "git commit --amend" example.
>
> [...]
> > I think it was useful to phrase it this way and list the various
> > advantages, disadvantages, and mitigations; thanks.
> >
> > Is there anything I've missed in my additional details?  Anything
> > sound crazy (beyond what I said I think case C leads to)?
>
> To sum it up, I lean toward (B), but (A) is not so bad, either
> (especially with the "I need some time to adapt to the change" opt-out
> config described under mitigations).

Sounds like we might be starting to converge towards a similar
opinion.  I roughly feel the same with leaning towards something like
(B) _if_ we can address the inconsistencies rather than just shuffle
them around (meaning we also turn off post-commit for interactive
rebase and cherry-pick) and move all of them over to some kind of
batch-commit-notification mechanism.

> Thanks for patiently working it through with me.
>
> My two cents,
> Jonathan

Likewise.  :-)
Elijah

  reply index

Thread overview: 143+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-12-20 17:09 [PATCH 00/15] rebase: make the default backend configurable Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 01/15] rebase: extend the options for handling of empty commits Elijah Newren via GitGitGadget
2019-12-20 21:29   ` Junio C Hamano
2019-12-21  0:32     ` Elijah Newren
2019-12-21 18:52       ` Elijah Newren
2019-12-21 23:49       ` Junio C Hamano
2019-12-20 17:09 ` [PATCH 02/15] t3406: simplify an already simple test Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 03/15] rebase, sequencer: remove the broken GIT_QUIET handling Elijah Newren via GitGitGadget
2019-12-20 21:34   ` Junio C Hamano
2019-12-20 17:09 ` [PATCH 04/15] rebase: make sure to pass along the quiet flag to the sequencer Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 05/15] rebase: fix handling of restrict_revision Elijah Newren via GitGitGadget
2019-12-20 21:37   ` Junio C Hamano
2019-12-20 17:09 ` [PATCH 06/15] t3432: make these tests work with either am or merge backends Elijah Newren via GitGitGadget
2019-12-22  5:11   ` Denton Liu
2019-12-23 17:17     ` Elijah Newren
2019-12-20 17:09 ` [PATCH 07/15] rebase: allow more types of rebases to fast-forward Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 08/15] git-rebase.txt: add more details about behavioral differences of backends Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 09/15] rebase: move incompatibility checks between backend options a bit earlier Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 10/15] rebase: add an --am option Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 11/15] contrib: change the prompt for am-based rebases Elijah Newren via GitGitGadget
2019-12-20 23:07   ` SZEDER Gábor
2019-12-21  0:17     ` Elijah Newren
2019-12-20 17:09 ` [PATCH 12/15] rebase tests: mark tests specific to the am-backend with --am Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 13/15] rebase tests: repeat some tests using the merge backend instead of am Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 14/15] rebase: make the backend configurable via config setting Elijah Newren via GitGitGadget
2019-12-20 17:09 ` [PATCH 15/15] rebase: change the default backend from "am" to "merge" Elijah Newren via GitGitGadget
2019-12-20 18:51 ` [PATCH 00/15] rebase: make the default backend configurable Alban Gruin
2019-12-20 18:55   ` Elijah Newren
2019-12-23 18:49 ` [PATCH v2 " Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 01/15] rebase: extend the options for handling of empty commits Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 02/15] t3406: simplify an already simple test Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 03/15] rebase, sequencer: remove the broken GIT_QUIET handling Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 04/15] rebase: make sure to pass along the quiet flag to the sequencer Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 05/15] rebase: fix handling of restrict_revision Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 06/15] t3432: make these tests work with either am or merge backends Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 07/15] rebase: allow more types of rebases to fast-forward Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 08/15] git-rebase.txt: add more details about behavioral differences of backends Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 09/15] rebase: move incompatibility checks between backend options a bit earlier Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 10/15] rebase: add an --am option Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 11/15] contrib: change the prompt for interactive-based rebases Elijah Newren via GitGitGadget
2019-12-23 22:00     ` Denton Liu
2019-12-23 18:49   ` [PATCH v2 12/15] rebase tests: mark tests specific to the am-backend with --am Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 13/15] rebase tests: repeat some tests using the merge backend instead of am Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 14/15] rebase: make the backend configurable via config setting Elijah Newren via GitGitGadget
2019-12-23 18:49   ` [PATCH v2 15/15] rebase: change the default backend from "am" to "merge" Elijah Newren via GitGitGadget
2019-12-24 19:54   ` [PATCH v3 00/15] rebase: make the default backend configurable Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 01/15] rebase: extend the options for handling of empty commits Elijah Newren via GitGitGadget
2020-01-07 14:37       ` Phillip Wood
2020-01-07 19:15         ` Elijah Newren
2020-01-08 14:27           ` Phillip Wood
2020-01-09 21:32             ` Johannes Schindelin
2019-12-24 19:54     ` [PATCH v3 02/15] t3406: simplify an already simple test Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 03/15] rebase, sequencer: remove the broken GIT_QUIET handling Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 04/15] rebase: make sure to pass along the quiet flag to the sequencer Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 05/15] rebase: fix handling of restrict_revision Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 06/15] t3432: make these tests work with either am or merge backends Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 07/15] rebase: allow more types of rebases to fast-forward Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 08/15] git-rebase.txt: add more details about behavioral differences of backends Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 09/15] rebase: move incompatibility checks between backend options a bit earlier Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 10/15] rebase: add an --am option Elijah Newren via GitGitGadget
2020-01-07 14:43       ` Phillip Wood
2020-01-07 19:26         ` Elijah Newren
2020-01-07 20:11           ` Junio C Hamano
2020-01-08 14:32             ` Phillip Wood
2020-01-08 17:18               ` Junio C Hamano
2020-01-08 18:55                 ` Phillip Wood
2019-12-24 19:54     ` [PATCH v3 11/15] git-prompt: change the prompt for interactive-based rebases Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 12/15] rebase tests: mark tests specific to the am-backend with --am Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 13/15] rebase tests: repeat some tests using the merge backend instead of am Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 14/15] rebase: make the backend configurable via config setting Elijah Newren via GitGitGadget
2019-12-24 19:54     ` [PATCH v3 15/15] rebase: change the default backend from "am" to "merge" Elijah Newren via GitGitGadget
2020-01-10 23:14       ` Jonathan Nieder
2020-01-11  1:16         ` Elijah Newren
2020-01-11 14:41           ` Phillip Wood
2020-01-12 17:59             ` Johannes Schindelin
2020-01-16  6:32               ` Elijah Newren
2020-01-16  7:58                 ` Jonathan Nieder
2020-01-16  8:06                   ` Jonathan Nieder
2020-01-16 16:18                     ` Elijah Newren
2020-01-16 20:35                       ` Jonathan Nieder
2020-01-16 21:30                         ` Elijah Newren
2020-01-16 22:39                           ` Jonathan Nieder
2020-01-16 23:19                             ` Elijah Newren [this message]
2020-01-16 23:25                           ` Junio C Hamano
2020-01-17  0:51                             ` Elijah Newren
2020-01-16 15:35                   ` Elijah Newren
2020-01-16 20:05                   ` Junio C Hamano
2020-01-16 10:48                 ` Johannes Schindelin
2020-01-12 21:23             ` Junio C Hamano
2020-01-15 19:50             ` Jonathan Nieder
2020-01-15 21:59               ` Emily Shaffer
2020-01-16  6:14     ` [PATCH v4 00/19] rebase: make the default backend configurable Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 01/19] git-rebase.txt: update description of --allow-empty-message Elijah Newren via GitGitGadget
2020-02-09 15:59         ` Phillip Wood
2020-02-13 18:35           ` Elijah Newren
2020-01-16  6:14       ` [PATCH v4 02/19] t3404: directly test the behavior of interest Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 03/19] rebase (interactive-backend): make --keep-empty the default Elijah Newren via GitGitGadget
2020-02-09 15:59         ` Phillip Wood
2020-02-13 18:52           ` Elijah Newren
2020-01-16  6:14       ` [PATCH v4 04/19] rebase (interactive-backend): fix handling of commits that become empty Elijah Newren via GitGitGadget
2020-02-10 14:27         ` Phillip Wood
2020-02-13 18:54           ` Elijah Newren
2020-02-16 14:46             ` Phillip Wood
2020-01-16  6:14       ` [PATCH v4 05/19] t3406: simplify an already simple test Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 06/19] rebase, sequencer: remove the broken GIT_QUIET handling Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 07/19] rebase: make sure to pass along the quiet flag to the sequencer Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 08/19] rebase: fix handling of restrict_revision Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 09/19] t3432: make these tests work with either am or merge backends Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 10/19] rebase: allow more types of rebases to fast-forward Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 11/19] git-rebase.txt: add more details about behavioral differences of backends Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 12/19] rebase: move incompatibility checks between backend options a bit earlier Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 13/19] rebase: add an --am option Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 14/19] git-prompt: change the prompt for interactive-based rebases Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 15/19] rebase: drop '-i' from the reflog " Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 16/19] rebase tests: mark tests specific to the am-backend with --am Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 17/19] rebase tests: repeat some tests using the merge backend instead of am Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 18/19] rebase: make the backend configurable via config setting Elijah Newren via GitGitGadget
2020-01-16  6:14       ` [PATCH v4 19/19] rebase: change the default backend from "am" to "merge" Elijah Newren via GitGitGadget
2020-01-17 16:58       ` [PATCH v4 00/19] rebase: make the default backend configurable Phillip Wood
2020-02-05 21:06         ` Junio C Hamano
2020-02-05 22:38           ` Elijah Newren
2020-02-15 21:36       ` [PATCH v5 00/20] rebase: make the default backend configurable and change the default Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 01/20] git-rebase.txt: update description of --allow-empty-message Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 02/20] t3404: directly test the behavior of interest Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 03/20] rebase (interactive-backend): make --keep-empty the default Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 04/20] rebase (interactive-backend): fix handling of commits that become empty Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 05/20] t3406: simplify an already simple test Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 06/20] rebase, sequencer: remove the broken GIT_QUIET handling Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 07/20] rebase: make sure to pass along the quiet flag to the sequencer Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 08/20] rebase: fix handling of restrict_revision Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 09/20] t3432: make these tests work with either am or merge backends Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 10/20] rebase: allow more types of rebases to fast-forward Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 11/20] git-rebase.txt: add more details about behavioral differences of backends Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 12/20] rebase: move incompatibility checks between backend options a bit earlier Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 13/20] rebase: add an --am option Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 14/20] git-prompt: change the prompt for interactive-based rebases Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 15/20] rebase: drop '-i' from the reflog " Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 16/20] rebase tests: mark tests specific to the am-backend with --am Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 17/20] rebase tests: repeat some tests using the merge backend instead of am Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 18/20] rebase: make the backend configurable via config setting Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 19/20] rebase: change the default backend from "am" to "merge" Elijah Newren via GitGitGadget
2020-02-15 21:36         ` [PATCH v5 20/20] rebase: rename the two primary rebase backends Elijah Newren via GitGitGadget
2020-02-16 15:01         ` [PATCH v5 00/20] rebase: make the default backend configurable and change the default Phillip Wood

Reply instructions:

You may reply publically 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=CABPp-BE9XqOCG0VqA5XgDbXm4apw0-_mOqW-JSO+PDZhisO0Rw@mail.gmail.com \
    --to=newren@gmail.com \
    --cc=Johannes.Schindelin@gmx.de \
    --cc=alban.gruin@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitgitgadget@gmail.com \
    --cc=gitster@pobox.com \
    --cc=jrnieder@gmail.com \
    --cc=liu.denton@gmail.com \
    --cc=phillip.wood@dunelm.org.uk \
    --cc=plroskin@gmail.com \
    --cc=szeder.dev@gmail.com \
    /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

Git Mailing List Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/git/0 git/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 git git/ https://lore.kernel.org/git \
		git@vger.kernel.org
	public-inbox-index git

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.git


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git