Re: git gc & deleted branches

[Junio C Hamano <gitster@pobox.com>]

Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:

> Well, I think that the setuid helper script would open a whole bunch of 
> other issues.
>
> I think that the shared repository problem is rather a semantic one, i.e. 
> it is only solvable between the owners of the repository by good-ole 
> talking, not something that can be solved by the tool (Git).

I very strongly agree with you that the suid helper would be the last
ditch thing we would want to avoid doing unless there is no other way.  I
also agree with you that the owners of the repository need to be talking.

But I think the tool _could_ help them do their talking.  It is
conceivable that just like you can explicitly allow selected others to
push into your own repository, you would want to explicitly allow some
others to ask you to keep objects their repositories borrow from you.

	Originally, I wrote "allow others to borrow from you", but that is
	very ill defined.  If they can read from your repository they can
	unilaterally borrow from you without having any write permission
	to your repository that is needed to install backpointers.

I am not fundamentally opposed to a backpointer that point at borrowers so
that the lender can protect the objects that are pointed by them.
However, there are two technical issues in the solution of pointing at the
borrower's .git/refs with a symlink from the repository that is borrowed
from, as I pointed out when this came up last time.

 - There are systems without symbolic links.

 - A symref (e.g. refs/remotes/origin/HEAD of borrower that points at
   refs/remotes/origin/master of borrower) is relative to borrower's
   repository.