(Initial conversion from Docbook)
(Initial conversion from Docbook)
This appendix lists useful online resources for further information about spam, spam-filtering, SpamAssassin, each of the mail transport agents discussed in this book, and several other SpamAssassin-related software packages.
General Spam Resources
"Help! I've been spammed! What do I do?"—originally written by Chris Lewis and maintained by Greg Byshenk—is an helpful (if dated) guide to spam and spam-prevention for the beginner. Find it at http://www.byshenk.net/ive.been.spammed.html.
Internet Request For Comments (RFC) documents describe proposed standards for the Internet. You can get RFCs from http://www.rfc-editor.org. Some notable RFCs related to spam and spam filtering include
- RFC 2822: Internet Message Format
- The basic document that describes the formatting of email messages.
- RFC 2821: Simple Mail Transfer Protocol
- Explains SMTP, the protocol used to transfer email from system to system.
- RFC 2505: Anti-spam Recommendations for Internet MTAs
- Describes a set of best practices for mail servers.
The SPAM-L FAQ, maintained by Doug Muth at http://www.claws-and-paws.com/spam-l/, provides information about the SPAM-L mailing list, one of the oldest discussion forums for spam fighters.
http://spam.abuse.net is a long-standing site with information for antispam advocates and system adminstrators.
The Coalition Against Unsolicited Commercial Email (CAUCE) has a web site at http://www.cauce.org. CAUCE focuses primarily on advocacy and legislation.
The groups in the news.admin.net-abuse Usenet hierarchy are devoted to discussing and reporting Net abuse, including spam (see particularly news.admin.net-abuse.email).
http://www.openrbl.org provides a long list of DNSBLs that may be suitable for spam-filtering. It includes hit rates (but not false positive rates) against its own recently collected spam corpus for each DNSBL.
Vipul's Razor, Pyzor, and DCC are collaborative spam-filtering clearinghouses that can be consulted by SpamAssassin. Vipul's Razor is available at http://razor.sourceforge.net. Pyzor is at http://pyzor.sourceforge.net. DCC is at http://www.rhyolite.com/anti-spam/dcc/.
DSPAM is a spam-filtering system using statistical-algorithmic hybrid filtering—filters that are trained like SpamAssassin's Bayesian classifier. Find it at http://www.nuclearelephant.com/projects/dspam/.
CRM114, the Controllable Regex Multilator, is a spam-filtering system based on learning regular expressions. Download it at http://crm114.sourceforge.net/.
The home page for SpamAssassin itself is, of course, http://www.spamassassin.org. The site contains links to the spamassassin-users and spamassassin-dev mailing lists.
SAProxy Pro, a commercial client-side SpamAssassin proxy, is available from http://www.statalabs.com.
Mail Transport Agents
In this book, I describe how to use SpamAssassin in conjunction with several MTAs. The following sections point you to more information about each of those agents.
Sendmail's own antispam provisions are documented at http://www.sendmail.org/m4/anti_spam.html. If you're not a Sendmail guru, pick up Bryan Costales' book Sendmail (O'Reilly), the Sendmail bible. If you are a Sendmail guru, you probably already have a copy! Another good source is Frederick Avolio and Paul Vixie's Sendmail: Theory and Practice (Digital Press).
Sendmail's filtering architecture, milter, has led to the development of many filtering tools. The web site http://www.milter.org is the most comprehensive catalog of such filters.
The home page for the Postfix MTA is http://www.postfix.org.
Two useful manuals for Postfix are Postfix: The Definitive Guide by Kyle Dent (O'Reilly) and The Book of Postfix by Ralf Hildebrandt and Patrick Koetter (No Starch Press).
The home page for the qmail MTA is http://www.qmail.org. The netqmail distribution of qmail includes the QMAILQUEUE patch, which is used by most SpamAssassin-integration solutions.
The online book Life with qmail (http://www.lifewithqmail.org) provides excellent documentation for qmail.
The home page for the Exim MTA is http://www.exim.org. The site includes a link to the Exim specification, which serves as the manual for the MTA. Those preferring a real book should purchase Philip Hazel's The Exim SMTP Mail Server: Official Guide for Release 4 (UIT Cambridge).
Related Mail Tools
The following sections point to information about various other mail tools mentioned in this book.
procmail is a popular and powerful Unix filtering program that acts as a full-featured local email delivery agent. It's available at http://www.procmail.org.
MIMEDefang is a Sendmail milter application written in Perl that provides a framework for mail content scanning, including virus-scanning, spam-checking with SpamAssassin, and MIME validation. It's available at http://www.mimedefang.org. Versions after 2.42 support SpamAssassin 3.0.
amavisd-new is a high-performance daemonized content scanner, designed for use with Postfix, but it also supports Sendmail (including a milter version) and Exim. It's available at http://www.ijs.si/software/amavisd/. Don't confuse amavisd-new with amavis, amavis-perl, amavisd, or amavis-ng, all of which are other content scanners, most not actively in development, that now share little code in common with amavisd-new.
sa-exim is a local_scan.c replacement for Exim that provides SpamAssassin message-scanning during a SMTP transaction. It offers many options for how to handle spam that it detects, including teergrubing. sa-exim's home page is http://marc.merlins.org/linux/exim/sa.html.
exiscan-acl is a patch for Exim 4 that adds new content-scanning ACL directives to Exim. These directives can be used to invoke SpamAssassin on messages during a SMTP transaction. Some prepackaged Exim distributions already have this patch added. You can download it at http://duncanthrax.net/exiscan-acl/.
qmail-scanner is a content scanner for qmail that can run SpamAssassin on messages early in the delivery process. It contains its own implementation of spamc for faster checking. Find it at http://qmail-scanner.sourceforge.net.