You are here: Home > Participate > Join a Discussion > Mailman Archives
<<< Chronological >>> Author Index    Subject Index <<< Threads >>>

Re: [anti-spam-wg@localhost] I wrote a spam filter in Perl

  • From: Walter Ian Kaye < >
  • Date: Thu, 28 Aug 2003 13:32:15 -0700
At 10:19a +0200 08/23/2003, Gert Doering didst inscribe upon an electronic papyrus:

Hi,

On Fri, Aug 22, 2003 at 10:35:59PM -0700, Walter Ian Kaye wrote:
 > Should I sell it to ISPs? To individuals?
 > Since it's Perl code, not compiled binary, how can I protect it?

Just make it open-source and sell consulting :-)
Cool idea... I could even appoint affiliate consultants in other geographic regions and list them on my Web site. I like win-win solutions. :D


At 12:32p +0200 08/23/2003, pna.lists didst inscribe upon an electronic papyrus:

1) English-like syntax rules file is cool, but is it flexible enough? Does
it still filter spam when the spammers write V1arga instead of Viagra?

It supports regular expressions, so it can catch eye/one/ell substitutions.
It won't predict *every* misspelling, because that would near infinity.


2) You may want to start doing pilot projects with (fre)email providers,
hosting companies and the ASP's

3) Have you measure the false positive and false negative results?
I've only had one false positive, because someone who always sent plain text unexpectedly sent multipart/alternative; now they're in my whitelist.
As for false negative, I have yet to figure out how to catch the Nigerian-style messages. Most of the incest spam goes to a retired address so I catch 'em that way, but I don't know how else to identify those either.

4) There *are* some Perl compilers available - it depends on the platform
you are using / targeting

5) You may want to try some Perl obfuscators

Thanks!


-Walter
 having a ton of fun watching 90% of spam automatically destroyed :D
  • Post To The List:
<<< Chronological >>> Author    Subject <<< Threads >>>