From: Junio C Hamano <gitster@pobox.com>
Date: Thu, 10 Sep 2015 17:45:04 GMT
Subject: Re: [PATCH v17 05/14] ref-filter: introduce match_atom_name()
Message-ID: <xmqqd1xq419b.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <CAOLa=ZQXP-U+dGYYFcvxVmuG2NOSiHYLHEwm2OtvhGQrBW9hdQ@mail.gmail.com>

Karthik Nayak <karthik.188@gmail.com> writes:

> It is one thing that the user can actually do the check themselves,
> but doesn't it make more sense that when we're using colon we expect a
> value after it, and something like %(color:) makes no sense when color
> specifically needs a value after the colon.

If you imagine the format being built by scripts (we are talking
about plumbing feature --format here), I think you will realize that
it perfectly makes sense to allow them to say "%(atom:$modifiation)"
without having to worry about a special case where $modification
happened to end up being empty.  So no, I do not agree with your
statement at all.