From: Bert Wesarg <bert.wesarg@googlemail.com>
Date: Sat, 11 Apr 2009 19:50:35 GMT
Subject: Re: [PATCH&RFC] get_short_ref(): add strict mode
Message-ID: <36ca99e90904111250t7f4563e6x495937d8f3f45d40@mail.gmail.com>
In-Reply-To: <7vr5zzdmg8.fsf@gitster.siamese.dyndns.org>

On Sat, Apr 11, 2009 at 21:23, Junio C Hamano <gitster@pobox.com> wrote:
> Bert Wesarg <bert.wesarg@googlemail.com> writes:
>
>> I think of 3 alternatives to use this mode for the "refname" format (and
>> probably others):
>>
>>   a) Use core.warnAmbiguousRefs to control strict mode.
>>      This would change the current default behaviour, because this is true
>>      by default.
>>
>>   b) Introduce a new core config variable to control this, either for
>>      for-each-ref alone ore globally.
>>
>>   c) Introduce a "refname:short-strict" format to get the strict abbreviation.
>>
>> I'm currently slighty in favour for option b).
>
> Your earlier http://thread.gmane.org/gmane.comp.version-control.git/96464
> made a lot of sense to me.  The request "refname:short" cannot be for use
> by scripts (well, scripts may pass it to for-each-ref but that has to be
> for final consumption by humans wanting to view the names in a format not
> overly long, as opposed to scripts using for-each-ref to extract
> unambiguous names to be used for further processing, in which case they
> would be using "refname" without ":short"), so I do not see "change the
> current default behaviour" is a bad thing at all.  If anything, it is an
> improvement, isn't it?
Sure it is. So you're still with option a) and I'm ok with this. I
prepare a patch.

Bert