[Debichem-devel] Fwd: Post-Receive Hooks

[James Clarke]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi,
I got the wrong email address the first time round for this. Please make
sure to copy in the relevant addresses from below when replying if
appropriate.

Regards,
James

- ---------- Forwarded message ----------
Date: Wed, 27 Apr 2016 23:20:02
From: James Clarke <jrtc27 at jrtc27.com>
To: debian-science at lists.debian.org, debian-astro at lists.debian.org,
     debichem-devel at lists.debian.org, debian-med at lists.debian.org
Subject: Post-Receive Hooks

Hi,
I noticed the other day that some debian-astro repositories were using
the debian-science post-receive hooks, rather than their own, leading to
things like deb.li links posted by KB in #debian-science pointing to
non-existent repositories. This prompted me to look through all the
repositories in debian-astro, debichem, debian-med and debian-science,
tabulating the results to see how consistent they are. The results vary
between the different teams, and are attached file, but here is a
summary of my findings:

1. A lot of debian-astro packages are calling debian-science's
   post-receive. This seems like a bad thing to do, especially given
   debian-astro has its own post-receive that seems to have been forked
   from debian-science (though may need updating, along with the KGB
   config).

2. Many packages lack any post-receive hooks. This is easy to fix, and I
   can't think of any reasons why this would be preferable for public
   repositories.

3. Many packages just call /usr/local/bin/git-commit-notice. Given each
   team has its own post-receive hook which does this, I again see no
   good reason for these not to use the correct hook.

   i. A few call git-post-receive-tag-pending as well (via pee(1)). I
     doubt maintainers would be very happy if this was done for every
     package, even though I am strongly in favour of the script (or forks
     thereof).

4. As you might expect, there are a few special snowflakes, including a
   few which I lack permission to read.

5. The official Debian Science policy is quite outdated when it comes to
   the section on repositories[1]. It has no mention of the
   setup-repository script, and encourages people to use
   /usr/local/bin/git-commit-notice, rather than
   debian-science-post-receive. If people are ok with it, I would be
   willing to update this.

Regards,
James

[1] http://debian-science.alioth.debian.org/debian-science-policy.html#idm46065384004448
-----BEGIN PGP SIGNATURE-----
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=+Kw1
-----END PGP SIGNATURE-----
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: hooks-audit.txt
URL: <http://lists.alioth.debian.org/pipermail/debichem-devel/attachments/20160427/cff58ae8/attachment-0001.txt>