Re: [PATCH v2 (resend)] ls-remote: fall-back to default remotes when no remote specified

From: Tay Ray Chuan <rctay89@gmail.com>
Date: Apr 8, 2010, 17:10 UTC
Message-ID: <t2mbe6fef0d1004081010sb6dfd0e1l2e73a36ef1af3870@mail.gmail.com>
In-Reply-To: <20100408071639.GJ30473@coredump.intra.peff.net>

Hi,

On Thu, Apr 8, 2010 at 3:16 PM, Jeff King <peff@peff.net> wrote:

Show 13 quoted lines

> On Thu, Apr 08, 2010 at 03:07:07PM +0800, Tay Ray Chuan wrote:
>
>>       remote = remote_get(dest);
>> +     if (!remote)
>> +             usage(ls_remote_usage);
>
> I don't see an update to ls_remote_usage, but shouldn't it now be:
>
>  git ls-remote [--heads] [--tags]  [-u <exec> | --upload-pack <exec>]
>    [repository] [<refs>...]
>
> or something now (while we're at it, maybe we can wrap it better as it's
> larger than 80 characters).

yes. I think

  git ls-remote [--heads] [--tags]  [-u <exec> | --upload-pack <exec>]
                [<repository> [<refs>...]]

would be more accurate.

> But once that is done, shouldn't the (!remote) case say something like
> "you don't have a default remote". The user didn't invoke ls-remote
> incorrectly (as the usage message shows), but rather there was a
> configuration problem.

Noted.

Show 22 quoted lines

>> +test_expect_success 'use branch.<name>.remote if possible' '
>> +
>> +     # Remove "origin" so that we know that ls-remote is not using it.
>> +     #
>> +     # Ideally, we should test that branch.<name>.remote takes precedence
>> +     # over "origin", but that is another matter altogether.
>> +     #
>> +     git remote rm origin &&
>> +     git config branch.master.remote self &&
>> +     git ls-remote >actual &&
>> +     test_cmp expected.all actual
>> +
>> +'
>
> Wouldn't your "ideally" just be:
>
>  git clone . other-remote &&
>  git push other-remote HEAD:unique-ref &&
>  git config branch.master.remote other-remote &&
>  ...
>
> and check for "unique-ref" in the output?

hmm, upon clone, tracking branches are already created, so we can just treat them as the "unique refs", without having to create a unique ref.

-- 
Cheers,
Ray Chuan