From: Jon Seymour <jon.seymour@gmail.com>
Date: Sat, 11 Jun 2005 07:20:46 GMT
Subject: Re: do people use the 'git' command?
Message-ID: <2cfc403205061100204cd43334@mail.gmail.com>
In-Reply-To: <7vpsutz88z.fsf@assigned-by-dhcp.cox.net>

On 6/11/05, Junio C Hamano <junkio@cox.net> wrote:
> >>>>> "JS" == Jon Seymour <jon.seymour@gmail.com> writes:
> 
> JS> Sebastian,
> JS> What is the justification for removing it?
> 
> JS> As it stands, git serves a useful function in that it provides a
> JS> single point through which access to all git-*-scripts is channeled,
> JS> yet still allows each script to be maintained as a separate entity,
> JS> coherent entity.
> 
> If I understand him correctly, Sebastian's point was that it
> channels only to git-*-scripts without channelling to any of the
> git-* binaries.
> 

What about generalizing it a little, so that:

     git word

calls:

    git-word-script if it exists and is executable in the path

or, if not,

    git-word if it exists and is executable in the path.

Then:

    git update-cache 

would work as expected.

jon.