spamcan

Last Update 1997-07-10
From: timb@consult.ml.org (Tim Berger)
Newsgroups: comp.mail.sendmail
Subject: Announce: spamcan-0.87, anti-spam sendmail patches to capture spam by regular expression
Date: 9 Jul 1997 13:57:48 -0700
Organization: Tim Berger Consulting
Keywords: sendmail, anti-spam, regular expression

I am now using spamcan in production at a small company of
approximately 100 people and it is working very well.  It works on
Linux and Solaris and should easily port to other platforms.  A 
context diff is now available.

It is available from: ftp://consult.ml.org/pub/spamcan.tar.gz

From the readme:

spamcan is a simple yet extremely powerful junkmail capturing sofware
patch to Sendmail v8.

spamcan is a patch to sendmail designed to capture site-wide junkmail
to a file.  In my experience, bouncing spam is futile and wasts
bandwidth.  Most spam has forged addresses and invalid return
addresses.

spamcan was written with the philosophy that unsolicited email is an
evil and disruptive presence.  In my opinion it is unreasonable to
ask or require each and every user at a site with hundreds of users to
put up with unsolicited email or to manage their own anti-spam rules.
I beleve it is up to the system administration group to provide
excellent service to the user community and I belive spamcan is a
powerful tool that aids in providing excellent service.

Unlike most site-wide anti-spam packages, spamcan is capable of using
regular expressions to match header strings.  The envelope "from" user
is also examined.  Using regular expressions is a very powerful and
dangerous capability.  Extreme caution must be taken in formulating
your expressions.  Fortunately, spamcan does not discard or bounce
email.  Rather, spamcan 'cans' it to a 'spamcan'.  In the unfortunate
event that you've managed to capture non-spam, it may be forwarded to
the appropriate user.

spamcan was designed to stay out of the way of valid mail message
traffic.  It examines all non-message-id headers.  Outgoing mail is
not examined.  Incoming mail messages containing the in-reply-to
header are also passed through un-examined.  If spammers use this as a
back-door this feature may have to be removed.

spamcan adds a header to every message it identifies as spam 
before sending it to the spamcan:

'x-spamcan-reason' has the reason for identifying the mail as spam.

spamcan is known to work on Linux and Solaris but I'm sure it would
not take much to make it work on other platforms.

Please let me know if you find spamcan useful.  Send bug reports,
enhancements, or comments to spamcan-proj@consult.ml.org.

-- 
Tim Berger, UNIX Consultant
mailto:tim@consult.ml.org
http://consult.ml.org
Software Engineering & System Administration


[(links)] [Hints] [Avoiding Spam] [New]
Copyright © Claus Aßmann Please send comments to: <ca@informatik.uni-kiel.de>