All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge
@ 2020-02-29 21:22 Alex Henrie
  2020-02-29 23:30 ` Elijah Newren
  2020-03-02 22:37 ` Junio C Hamano
  0 siblings, 2 replies; 5+ messages in thread
From: Alex Henrie @ 2020-02-29 21:22 UTC (permalink / raw)
  To: git, rcdailey.lists, newren, rsbecker, gitster, annulen, tytso
  Cc: Alex Henrie

Often novice Git users forget to say "pull --rebase" and end up with an
unnecessary merge from upstream. What they usually want is either "pull
--rebase" in the simpler cases, or "pull --ff-only" to update the copy
of main integration branches, and rebase their work separately. The
pull.rebase configuration variable exists to help them in the simpler
cases, but there is no mechanism to make these users aware of it.

Issue a warning message when no --[no-]rebase option from the command
line and no pull.rebase configuration variable is given. This will
inconvenience those who never want to "pull --rebase", who haven't had
to do anything special, but the cost of the inconvenience is paid only
once per user, which should be a reasonable cost to help a number of new
users.

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
---
v3:
- Fix segfault if opt_ff is NULL
- Don't say that the behavior will change in a future version of Git
(that hasn't been decided yet)
- Mention that pull.ff=only is also an option
- Mention git config's --global option
- Make the warning message longer and more clear in general
---
 builtin/pull.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/builtin/pull.c b/builtin/pull.c
index 3e624d1e00..a93ee489f6 100644
--- a/builtin/pull.c
+++ b/builtin/pull.c
@@ -327,6 +327,20 @@ static enum rebase_type config_get_rebase(void)
 	if (!git_config_get_value("pull.rebase", &value))
 		return parse_config_rebase("pull.rebase", value, 1);
 
+	if (!opt_ff || strcmp(opt_ff, "--ff-only")) {
+		warning(_("Pulling without specifying how to reconcile divergent branches\n"
+			"is discouraged. You can squelch this message by running one of the\n"
+			"following commands sometime before your next pull:\n"
+			"\n"
+			"  git config pull.rebase false  # merge (the default strategy)\n"
+			"  git config pull.rebase true   # rebase\n"
+			"  git config pull.ff only       # fast-forward only\n"
+			"\n"
+			"You can replace \"git config\" with \"git config --global\" to set a default\n"
+			"preference for all repositories. Alternatively, you can pass --rebase,\n"
+			"--no-rebase, or --ff-only on the command line every time you pull.\n"));
+	}
+
 	return REBASE_FALSE;
 }
 
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge
  2020-02-29 21:22 [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge Alex Henrie
@ 2020-02-29 23:30 ` Elijah Newren
  2020-03-01  3:27   ` Alex Henrie
  2020-03-02 22:37 ` Junio C Hamano
  1 sibling, 1 reply; 5+ messages in thread
From: Elijah Newren @ 2020-02-29 23:30 UTC (permalink / raw)
  To: Alex Henrie
  Cc: Git Mailing List, Robert Dailey, rsbecker, Junio C Hamano,
	annulen, Theodore Tso

Hi Alex,

On Sat, Feb 29, 2020 at 1:23 PM Alex Henrie <alexhenrie24@gmail.com> wrote:
>
> Often novice Git users forget to say "pull --rebase" and end up with an
> unnecessary merge from upstream. What they usually want is either "pull
> --rebase" in the simpler cases, or "pull --ff-only" to update the copy
> of main integration branches, and rebase their work separately. The
> pull.rebase configuration variable exists to help them in the simpler
> cases, but there is no mechanism to make these users aware of it.
>
> Issue a warning message when no --[no-]rebase option from the command
> line and no pull.rebase configuration variable is given. This will
> inconvenience those who never want to "pull --rebase", who haven't had
> to do anything special, but the cost of the inconvenience is paid only
> once per user, which should be a reasonable cost to help a number of new
> users.

Thanks for working on this.

> Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
> ---
> v3:
> - Fix segfault if opt_ff is NULL
> - Don't say that the behavior will change in a future version of Git
> (that hasn't been decided yet)
> - Mention that pull.ff=only is also an option
> - Mention git config's --global option
> - Make the warning message longer and more clear in general
> ---
>  builtin/pull.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
>
> diff --git a/builtin/pull.c b/builtin/pull.c
> index 3e624d1e00..a93ee489f6 100644
> --- a/builtin/pull.c
> +++ b/builtin/pull.c
> @@ -327,6 +327,20 @@ static enum rebase_type config_get_rebase(void)
>         if (!git_config_get_value("pull.rebase", &value))
>                 return parse_config_rebase("pull.rebase", value, 1);
>
> +       if (!opt_ff || strcmp(opt_ff, "--ff-only")) {
> +               warning(_("Pulling without specifying how to reconcile divergent branches\n"
> +                       "is discouraged. You can squelch this message by running one of the\n"
> +                       "following commands sometime before your next pull:\n"
> +                       "\n"
> +                       "  git config pull.rebase false  # merge (the default strategy)\n"
> +                       "  git config pull.rebase true   # rebase\n"
> +                       "  git config pull.ff only       # fast-forward only\n"
> +                       "\n"
> +                       "You can replace \"git config\" with \"git config --global\" to set a default\n"
> +                       "preference for all repositories. Alternatively, you can pass --rebase,\n"
> +                       "--no-rebase, or --ff-only on the command line every time you pull.\n"));

Ick; we don't need the warning to regurgitate large sections of the
manual, we only need to provide pointers and then they can look up the
details in their favorite way (google, stackoverflow, looking at the
actual manpage, etc.)  How about something more like

> +               warning(_("Pulling without specifying how to reconcile divergent branches\n"
> +                       "is discouraged. You can squelch this message by setting pull.rebase,\n"
> +                       "or by passing any of --rebase, --no-rebase, or --ff-only\n"));

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge
  2020-02-29 23:30 ` Elijah Newren
@ 2020-03-01  3:27   ` Alex Henrie
  2020-03-02 21:51     ` Junio C Hamano
  0 siblings, 1 reply; 5+ messages in thread
From: Alex Henrie @ 2020-03-01  3:27 UTC (permalink / raw)
  To: Elijah Newren
  Cc: Git Mailing List, Robert Dailey, rsbecker, Junio C Hamano,
	Konstantin Tokarev, Theodore Tso

On Sat, Feb 29, 2020 at 4:31 PM Elijah Newren <newren@gmail.com> wrote:
>
> Ick; we don't need the warning to regurgitate large sections of the
> manual, we only need to provide pointers and then they can look up the
> details in their favorite way (google, stackoverflow, looking at the
> actual manpage, etc.)  How about something more like
>
> > +               warning(_("Pulling without specifying how to reconcile divergent branches\n"
> > +                       "is discouraged. You can squelch this message by setting pull.rebase,\n"
> > +                       "or by passing any of --rebase, --no-rebase, or --ff-only\n"));

I intentionally made the message similar to the message you get when
you try to commit without setting user.name or user.email. Providing
the user with clear information on the spot is much more effective at
getting them to do the right thing than expecting them to use Google
or reread the man pages. We also want the user to understand that they
only have to set pull.rebase or pull.ff once, whereas a command-line
flag applies only to the current pull. Nevertheless, any warning
message is better than no warning at all, so I would be happy to
resubmit with your proposed text if there's consensus that a shorter
message is better.

-Alex

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge
  2020-03-01  3:27   ` Alex Henrie
@ 2020-03-02 21:51     ` Junio C Hamano
  0 siblings, 0 replies; 5+ messages in thread
From: Junio C Hamano @ 2020-03-02 21:51 UTC (permalink / raw)
  To: Alex Henrie
  Cc: Elijah Newren, Git Mailing List, Robert Dailey, rsbecker,
	Konstantin Tokarev, Theodore Tso

Alex Henrie <alexhenrie24@gmail.com> writes:

> On Sat, Feb 29, 2020 at 4:31 PM Elijah Newren <newren@gmail.com> wrote:
>>
> ... We also want the user to understand that they
> only have to set pull.rebase or pull.ff once, whereas a command-line
> flag applies only to the current pull.

I do not think the phrasing you used has the intended effect.

    You can replace "git config" with "git config --global" to set a
    default preference for all repositories.  You can pass --rebase,
    --no-rebase, or --ff-only on the command line to override the
    configured default per invocation.

may be more appropriate.  "You can use it, but you need to do so
every time you pull" drives readers away by placing stress on how
cumbersome it *can* become, but in practice, nobody would use the
same command line option every time anyway.

In contrast, explaining the command line option as a useful
mechanism for one-shot override, and the configuration variable as a
useful mechanism for set-and-forget convenience feature, would teach
readers in what situation use of the configuration variables and the
command line options shine the best.

Thanks.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge
  2020-02-29 21:22 [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge Alex Henrie
  2020-02-29 23:30 ` Elijah Newren
@ 2020-03-02 22:37 ` Junio C Hamano
  1 sibling, 0 replies; 5+ messages in thread
From: Junio C Hamano @ 2020-03-02 22:37 UTC (permalink / raw)
  To: Alex Henrie; +Cc: git, rcdailey.lists, newren, rsbecker, annulen, tytso

Alex Henrie <alexhenrie24@gmail.com> writes:

> v3:
> - Fix segfault if opt_ff is NULL
> - Don't say that the behavior will change in a future version of Git
> (that hasn't been decided yet)
> - Mention that pull.ff=only is also an option
> - Mention git config's --global option
> - Make the warning message longer and more clear in general
> ---
>  builtin/pull.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)

Another thing tests noticed was that this broke #5521.2.

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2020-03-02 22:37 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-02-29 21:22 [PATCH v3] pull: warn if the user didn't say whether to rebase or to merge Alex Henrie
2020-02-29 23:30 ` Elijah Newren
2020-03-01  3:27   ` Alex Henrie
2020-03-02 21:51     ` Junio C Hamano
2020-03-02 22:37 ` Junio C Hamano

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.