Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

gc: recommend git gc --prune=now instead of git prune #652

Closed

Conversation

johnlinp
Copy link

@johnlinp johnlinp commented Jun 8, 2020

git prune is a plumbing command and should not be run directly by
users. The corresponding porcelain command is git gc, which is
mentioned in the man page of git prune.

Signed-off-by: John Lin johnlinp@gmail.com

Fix according to #642.

@johnlinp
Copy link
Author

johnlinp commented Jun 8, 2020

/submit

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 8, 2020

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 8, 2020

On the Git mailing list, Denton Liu wrote (reply to this):

Hi John,

Thanks for working on this topic.

On Mon, Jun 08, 2020 at 02:02:18AM +0000, John Lin via GitGitGadget wrote:
> Subject: [PATCH] Recommend "git gc --prune=now" instead of "git prune"

The format for subjects is generally "<area>: <description>" so perhaps
something like "git gc: recommend use of `git gc --prune=now`"?

> From: John Lin <johnlinp@gmail.com>
> 
> Signed-off-by: John Lin <johnlinp@gmail.com>

This commit is missing some justification. In the commit message, we
should mention that `git prune` is a plumbing command and in the
documentation for it, it mentions that the porcelain `git gc` command
should be used instead.

`git prune` is a plumbing command and should not be run directly by
users. The corresponding porcelain command is `git gc`, which is
mentioned in the man page of `git prune`.

Signed-off-by: John Lin <johnlinp@gmail.com>
@johnlinp johnlinp force-pushed the fix-git-gc-warning-message branch from 83b7137 to 42aa638 Compare June 9, 2020 00:16
@johnlinp
Copy link
Author

johnlinp commented Jun 9, 2020 via email

@johnlinp
Copy link
Author

johnlinp commented Jun 9, 2020

/submit

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 9, 2020

@johnlinp johnlinp changed the title Recommend "git gc --prune=now" instead of "git prune" gc: recommend git gc --prune=now instead of git prune Jun 9, 2020
@Denton-L
Copy link
Member

Denton-L commented Jun 9, 2020

Thanks for the update, John. By the way, when responding to the emails from the mailing list, you should respond to the actual emails that you get CC'd on. If you respond to the GGG emails from GitHub, the response is only posted here and not everyone checks GitHub regularly.

@johnlinp
Copy link
Author

johnlinp commented Jun 9, 2020

Ouch. I didn't notice this. Thanks for telling me. I'll reply to the actual email next time.

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 9, 2020

On the Git mailing list, Junio C Hamano wrote (reply to this):

"John Lin via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: John Lin <johnlinp@gmail.com>
>
> `git prune` is a plumbing command and should not be run directly by
> users. The corresponding porcelain command is `git gc`, which is
> mentioned in the man page of `git prune`.

Much better than v1 that came without any justification ;-)

I however wouldn't say "should not"---that feels a bit too strong (I
personally use "prune" from the command line once or twice every
week and do not see why I should be forbidden from doing so [*1*]),
but the users who see this message would not need such a precise
control afforded by use of "git prune", so "gc --prune=now" is a
better recommendation.

Thanks.

[Footnote]

*1* The command does not even produce a machine readable output.
    The reason it is not recommended (but that is different from
    "should not") to casual users over "git gc" is because it is a
    very focused tool (it is only about removing loose objects and
    does not touch other things) and applicable to a very specific
    condition.  Those who can recognize that they are in that
    situation should not be forbidden from using it with a dogmatic
    "should not".

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 9, 2020

On the Git mailing list, Ævar Arnfjörð Bjarmason wrote (reply to this):


On Tue, Jun 09 2020, John Lin via GitGitGadget wrote:

> From: John Lin <johnlinp@gmail.com>
>
> `git prune` is a plumbing command and should not be run directly by
> users. The corresponding porcelain command is `git gc`, which is
> mentioned in the man page of `git prune`.

This change feels incomplete without a change to git-prune's
documentation, see 8d308b3540 ("Documentation: point git-prune users to
git-gc", 2008-04-29).

I.e. it still talks about "in most cases you shouldn't run this", but
here we are removing a case where it would otherwise make sense because
the user shouldn't use it directly.

I think instead the small change that makes the most sense here is to
just add "prune" to completions, i.e. I disagree with Denton Liu in
https://github.com/gitgitgadget/git/issues/642#issuecomment-640202110

The "right" change is much harder, there's numerous odd edge cases that
"gc" can get into with these loose objects (for the curious [1] and [2]
are a nice entry to that rabbit hole).

Fixing that is non-trivial, but in the meantime I don't think it's good
to:

 a) Recommend an option to "gc" which implies "repack -a", which as
    noted in gc's docs can produce corruption in a live repository.

    I.e. I don't think "let's not recommend this plumbing" should lead
    us to "let's recommend this thing that could corrupt the repository
    instead".

 b) Even if you don't get the corruption with "a" above "gc --prune=now"
    will produce a very differently packed repository. This might have
    unexpected performance etc. trade-offs the user wasn't expecting
    from a "run this for a quick fix" command.

 c) In the cases where "prune" here would help, e.g. on some large repo
    where something else added a lot of loose unreachable objects during
    the "gc" for some reason recommending a full "gc" might take a lot
    more time than just the "prune".

1. https://public-inbox.org/git/87eflvmovg.fsf@evledraar.gmail.com/
2. https://public-inbox.org/git/8736lnxlig.fsf@evledraar.gmail.com/

> Signed-off-by: John Lin <johnlinp@gmail.com>
> ---
>     Recommend "git gc --prune=now" instead of "git prune"
>
>     Signed-off-by: John Lin johnlinp@gmail.com [johnlinp@gmail.com]
>
>     Fix according to #642.
>
> Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-652%2Fjohnlinp%2Ffix-git-gc-warning-message-v2
> Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-652/johnlinp/fix-git-gc-warning-message-v2
> Pull-Request: https://github.com/gitgitgadget/git/pull/652
>
> Range-diff vs v1:
>
>  1:  83b7137abfd ! 1:  42aa6380067 Recommend "git gc --prune=now" instead of "git prune"
>      @@ Metadata
>       Author: John Lin <johnlinp@gmail.com>
>
>        ## Commit message ##
>      -    Recommend "git gc --prune=now" instead of "git prune"
>      +    gc: recommend `git gc --prune=now` instead of `git prune`
>      +
>      +    `git prune` is a plumbing command and should not be run directly by
>      +    users. The corresponding porcelain command is `git gc`, which is
>      +    mentioned in the man page of `git prune`.
>
>           Signed-off-by: John Lin <johnlinp@gmail.com>
>
>
>
>  builtin/gc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/builtin/gc.c b/builtin/gc.c
> index 8e0b9cf41b3..74185eac917 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -692,7 +692,7 @@ int cmd_gc(int argc, const char **argv, const char *prefix)
>
>  	if (auto_gc && too_many_loose_objects())
>  		warning(_("There are too many unreachable loose objects; "
> -			"run 'git prune' to remove them."));
> +			"run 'git gc --prune=now' to remove them."));
>
>  	if (!daemonized)
>  		unlink(git_path("gc.log"));
>
> base-commit: 20514004ddf1a3528de8933bc32f284e175e1012

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 9, 2020

On the Git mailing list, Junio C Hamano wrote (reply to this):

Ævar Arnfjörð Bjarmason <avarab@gmail.com> writes:

> On Tue, Jun 09 2020, John Lin via GitGitGadget wrote:
>
>> From: John Lin <johnlinp@gmail.com>
>>
>> `git prune` is a plumbing command and should not be run directly by
>> users. The corresponding porcelain command is `git gc`, which is
>> mentioned in the man page of `git prune`.
>
> This change feels incomplete without a change to git-prune's
> documentation, see 8d308b3540 ("Documentation: point git-prune users to
> git-gc", 2008-04-29).
>
> I.e. it still talks about "in most cases you shouldn't run this", but
> here we are removing a case where it would otherwise make sense because
> the user shouldn't use it directly.
>
> I think instead the small change that makes the most sense here is to
> just add "prune" to completions,...

That's perfectly reasonable stance to take.

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 10, 2020

On the Git mailing list, 林自均 wrote (reply to this):

Hi all,

Thank you Bjarmason & Junio. If Denton doesn't have other concerns,
I'll start working on adding bash completion for "git prune".

Best,
John Lin

Junio C Hamano <gitster@pobox.com> 於 2020年6月10日 週三 上午12:03寫道:
>
> Ævar Arnfjörð Bjarmason <avarab@gmail.com> writes:
>
> > On Tue, Jun 09 2020, John Lin via GitGitGadget wrote:
> >
> >> From: John Lin <johnlinp@gmail.com>
> >>
> >> `git prune` is a plumbing command and should not be run directly by
> >> users. The corresponding porcelain command is `git gc`, which is
> >> mentioned in the man page of `git prune`.
> >
> > This change feels incomplete without a change to git-prune's
> > documentation, see 8d308b3540 ("Documentation: point git-prune users to
> > git-gc", 2008-04-29).
> >
> > I.e. it still talks about "in most cases you shouldn't run this", but
> > here we are removing a case where it would otherwise make sense because
> > the user shouldn't use it directly.
> >
> > I think instead the small change that makes the most sense here is to
> > just add "prune" to completions,...
>
> That's perfectly reasonable stance to take.

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 10, 2020

On the Git mailing list, Denton Liu wrote (reply to this):

On Wed, Jun 10, 2020 at 11:43:09AM +0800, 林自均 wrote:
> Hi all,
> 
> Thank you Bjarmason & Junio. If Denton doesn't have other concerns,
> I'll start working on adding bash completion for "git prune".

No concerns. Thanks for contributing!

@gitgitgadget
Copy link

gitgitgadget bot commented Jun 10, 2020

On the Git mailing list, 林自均 wrote (reply to this):

Thanks Denton!

Best,
John Lin

Denton Liu <liu.denton@gmail.com> 於 2020年6月10日 週三 上午11:48寫道:
>
> On Wed, Jun 10, 2020 at 11:43:09AM +0800, 林自均 wrote:
> > Hi all,
> >
> > Thank you Bjarmason & Junio. If Denton doesn't have other concerns,
> > I'll start working on adding bash completion for "git prune".
>
> No concerns. Thanks for contributing!

@johnlinp johnlinp closed this Jun 10, 2020
@dscho
Copy link
Member

dscho commented Jun 10, 2020

Please note that closing PRs and opening new ones loses all history, and also does not add the range-diff that is so helpful for reviewers.

Next time, just force-push to the same PR and /submit again.

@johnlinp
Copy link
Author

@dscho Thank you for the advice. I would like to ask more about this issue.

I understand that I shouldn't close #651 and create #652 since the 2 PRs are closed related and the history of #651 is important for the reviewers in #652.

However, I think it's clearer to create a new PR for the bash completion of git prune since it's so different from #652. What do you think?

@johnlinp johnlinp deleted the fix-git-gc-warning-message branch June 13, 2020 03:21
@dscho
Copy link
Member

dscho commented Jun 13, 2020

Of course, you can do what you want, but if I had submitted these patches, I would have made them v1 and v2.

@johnlinp
Copy link
Author

I see. Thank you for the suggestions.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants