Re: fc/pull-merge-rebase, was Re: What's cooking in git.git (Dec 2020, #01; Tue, 8)

From: Felipe Contreras <felipe.contreras@gmail.com>
Date: Dec 11, 2020, 01:33 UTC
Message-ID: <CAMP44s2Po25VyHs-xQG0_oKhAOvPK_PNu5edQkYYsScZeBw1nw@mail.gmail.com>
In-Reply-To: <xmqqim99wm61.fsf@gitster.c.googlers.com>

On Thu, Dec 10, 2020 at 12:28 PM Junio C Hamano <gitster@pobox.com> wrote:

> It however doesn't give useful input to help us answer the questions
> Johannes raised: is it sensible to force users to tell "git pull" if
> they want to merge or to rebase explicitly, instead of defaulting to
> merge like we currently do?

Yes. As evidenced by the fact that this topic comes back again and again. There's the "Pull is Mostly Evil" thread, there's countless blog posts urging users to avoid "git pull" altogether, and many questions in Stack Overflow like "How to avoid merge commits from Git pull when pushing to remote" [1].

You didn't answer my previous question: "Do you want me to dig a decade of discussions and coalesce those conclusions into a summary so we can decide how to proceed?" [2]

I'll take your pondering about how sensible this solution is as a "yes".

However, we would not be doing that *today*.

> how much damage are we causing to
> existing users who expect the command to work the way it currently
> does?

Zero. Because my proposal does *not* make the pull fail, it merely prints a warning that it will change in the future.

After introducing "pull.mode=ff-only" we have all the time in the world to ponder if this is really the default we want.

In the meantime users can already start to move to this new behavior.

Zero people will be affected negatively, as the only thing that changes is the wording of the warning.

Cheers.

[1] https://stackoverflow.com/questions/30052104/how-to-avoid-merge-commits-from-git-pull-when-pushing-to-remote [2] https://lore.kernel.org/git/CAMP44s13YFZeOMz6V5sPdOnLXD-v3aQZiP7vvXXNfQLZP4Puwg@mail.gmail.com/

-- 
Felipe Contreras