[Mailman-Developers] Usenet gating and archiving
Barry A. Warsaw
email@example.com (Barry A. Warsaw)
Fri, 2 Apr 1999 15:56:26 -0500 (EST)
Is anybody currently gating Usenet to a mailing list, /and/ doing
archiving with the default Pipermail? If so, you may not have noticed
that the archives are pretty broken; nothing that originates on Usenet
will show up in the archive.
I believe I understand why this is happening, and although I don't
have a fix yet, it shouldn't be too difficult. I should be able to
get it into 1.0b11. The problem occurs because Pipermail uses
Python's mailbox.UnixMailbox objects which expect "From " separators.
Message that originate on Usenet don't have these envelopes. The
patch should be to simply synthesize these on messages gated off the
However, there's a more severe breakage. You've actually lost
information that is hard to reproduce because the flat listname.mbox
files don't have the separators in them either. Running bin/arch over
the file will give you even more corrupt archives (the incremental
archiver just throws the Usenet messages away; bin/arch will tack them
onto the last email-originated message).
What you could do is to trek through the listname.mbox file looking
for Newsgroup: headers (which appear, but I don't think are
guaranteed, to be the first header in the message). When you find
one, you jam in a synthesized "From " envelope.
I would rather not add this hack to Mailman unless a lot of people
would benefit from it. For the test lists that I'm running, we're
willing to just trash the archives and start over; the lists aren't
live anyway. I'd like to know who would be really adversely affected
by this problem.