dnsbl'S + bogofilter = spam barbecue

[Tom Anderson]

From: "Chris Fortune" <cfortune at telus.net>
> Any single DNSBL can be mistaken, so mail should not be rejected if it is 
> listed on only one DNSBL.  To be careful query 4 of

That's a good idea as long as the ones being compared are all checking for 
the same standard.  For instance, you can't compare a list that tracks open 
relays to one that tracks improper MX records.  The fact that any mistakes 
are simply bounced with the cause listed makes me fairly secure in 
implementing just one list for each standard, as the sender can always try 
contacting in a different fashion if important, or else try to get 
themselves off the respective list.

> best and factor the results.  I like to roll my own solutions, and I don't 
> like most MTAs' implementations of RBL checking - too
> inflexible -, so I use a perl system call to a C program `rblcheck`, kill 
> the mail if there are 3 or more hits, and add points to
> the bogosity score otherwise.  Here is the algorithm that has worked very 
> well for me, written in perl pseudocode:

I'm interested in the MTA implementation in order to block spam at SMTP time 
and never have to receive or process it with bogofilter.  However, I'd love 
to use bogofilter to supply my own dnsbl list based on spam that gets 
through.  This way, if an address spams one account, it can be blocked 
before it spams another.  Unfortunately, I don't have time at the moment to 
research this.

Tom