From mboxrd@z Thu Jan  1 00:00:00 1970
From: Kyle Meyer <kyle@kyleam.com>
Subject: Re: git-annex: problematic shebangs in .git/hooks/pre-commit?
Date: Thu, 23 Aug 2018 13:05:09 -0400
Message-ID: <8736v5xbcq.fsf@kyleam.com>
References: <876001xe9c.fsf@kyleam.com> <87wosh2ftj.fsf@ngyro.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Return-path: <guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org>
Received: from eggs.gnu.org ([2001:4830:134:3::10]:41353)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kyle@kyleam.com>) id 1fst2u-0002KQ-Be
	for guix-devel@gnu.org; Thu, 23 Aug 2018 13:05:21 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kyle@kyleam.com>) id 1fst2q-0002tt-AY
	for guix-devel@gnu.org; Thu, 23 Aug 2018 13:05:20 -0400
Received: from pb-smtp1.pobox.com ([64.147.108.70]:58463)
	by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32)
	(Exim 4.71) (envelope-from <kyle@kyleam.com>) id 1fst2q-0002tA-1v
	for guix-devel@gnu.org; Thu, 23 Aug 2018 13:05:16 -0400
In-Reply-To: <87wosh2ftj.fsf@ngyro.com>
List-Id: "Development of GNU Guix and the GNU System distribution."
	<guix-devel.gnu.org>
List-Unsubscribe: <https://lists.gnu.org/mailman/options/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/archive/html/guix-devel/>
List-Post: <mailto:guix-devel@gnu.org>
List-Help: <mailto:guix-devel-request@gnu.org?subject=help>
List-Subscribe: <https://lists.gnu.org/mailman/listinfo/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=subscribe>
Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org
Sender: "Guix-devel" <guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org>
To: Timothy Sample <samplet@ngyro.com>
Cc: guix-devel <guix-devel@gnu.org>

Timothy Sample <samplet@ngyro.com> writes:

[...]

>> I'm wondering whether the shebang patching in .git/hooks/pre-commit will
>> cause a problem.  Using the patched shellPath_portable, 'git annex init'
>> generates a hook like this:
>>
>>     % cat .git/hooks/pre-commit
>>     #!/gnu/store/rbrandv7anzjxqkr40d7fkanzssslk4b-bash-minimal-4.4.19/bi=
n/sh
>>     # automatically configured by git-annex
>>     git annex pre-commit .
>>
>> But won't this break if that particular bash-mininimal is garbage collec=
ted?
>
> My understanding is that since git-annex contains a reference to
> bash-minimal, the garbage collector will leave it alone.  Indeed,
> running =E2=80=9Cguix gc --references /gnu/store/=E2=80=A6-git-annex-2018=
0626=E2=80=9D shows
> that Guix knows that git-annex needs bash-minimal.  As long as git-annex
> is available (a prerequisite to the pre-commit hook working)
> bash-minimal will be available.

Hmm, I'm not convinced that *that* bash-minimal will be available.

Perhaps I'm just missing something, but to spell out my concern a little
more: I have /gnu/store/A/git-annex and /gnu/store/X/bash-minimal.  I
set up a repo with 'git annex init', and
/gnu/store/X/bash-minimal/bin/sh is used in the shebang.

Some time later I am using /gnu/store/B/git-annex, which references
/gnu/store/Y/bash-minimal.  That repo's hook still uses the
X/bash-minimal/bin/sh.  But if /gnu/store/X/bash-minimal is no longer
referenced by any package, it could be garbage collected, leading to the
hook failing.