As we know, sg was blacklisted at spamcop based on spam reportedly originating from the sg server.
I've looked at this a bit and I don't believe this is purposeful on anybody's part -- an example of the reported headers are:
From x Thu Oct 16 01:05:42 2003
Return-Path: <jqh1@gourmet.spamgourmet.com>
Received: from gourmet.spamgourmet.com ([216.218.230.146])
by mail1.aus1.texas.net (8.11.6p3/8.11.6p3) with ESMTP id h9GAqwi11511
for <x>; Thu, 16 Oct 2003 05:52:58 -0500 (CDT)
Received: from localhost.localdomain (localhost [127.0.0.1])
by localhost (8.12.10/8.12.9) with ESMTP id h9GB3wPK004142
for <x>; Thu, 16 Oct 2003 04:03:58 -0700
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 7bit
Mime-Version: 1.0
X-Mailer: MIME-tools 5.41 (Entity 5.404)
Date: Thu, 16 Oct 2003 03:52:55 -0700 (PDT)
From: x dan <alhaji@ratrunner.com>
To: x
Subject: urgent reply (satxrr: message 2 of 6)
Reply-To: x
X-Originating-Ip: [81.91.227.9]
Message-Id: <2003________________397A@sitemail.everyone.net>
Sender: ",,," <jqh1@gourmet.spamgourmet.com>
X-UIDL: I@H!!~#l"!BLY!!)"="!
I think this was automatically reported to spamcop via spamassassin or something similar, and this has likely been going on for sg's entire history. The difference here is that the headers *do* indicate that the message originated at the sg server, even though it's obvious to all of us that it didn't.
I *think* this caused by the Mail::Audit code discarding the old headers and "re-originating" the message when it sends. The old code preserved the previous headers and sg would correctly not appear as the originating server, and so never got blacklisted.
I have (with more than a little chagrin) once again reverted to the pre-Mail::Audit version of the code and I'll make some effort to prove up this theory in the next couple of days.
Syskoll, you reported the situation to Julian, no? I haven't contacted them yet, although I did join the mailing list. Do you think we should contact Julian again immediately or wait until we substantiate or disprove the theory above?
Josh