I really enjoy ASSP working in front of the SMTP server so that senders
*know* when their mail is not going to be delivered.
Unfortunately, my clients are really not capable of properly training
ASSP for Bayes, and so whenever I enable Bayes they inevitably make me
disable it again because of false-positives. Having had some experience
with SpamAssassin in the past, they believe that the Bayes filters in
SpamAssassin will work better for them (read: more reliably, less
false-positives) than those with ASSP. Based on what I have seen I tend
to agree... although I certainly am willing to admit otherwise.
So to get to the point, I would like to see an "extentible external
hook" option on ASSP that pipes the mail to the hook (external script or
executable, etc) and then looks for something like a return value or
hook output to identify whether or not to flag the message as spam or to
proceed with the rest of what ASSP would do. I am willing to pay for
I am primarly interested in this so that I can write a hook that will,
in turn, pipe the mail to spamc and then see what SpamAssassin has to
say about the mail to 1) keep some statistics on how SpamAssassin works
in comparison to ASSP; 2) to be able to utilize features found in
SpamAssassin from ASSP; and 3) to be able to flag a message as spam from
that hook program/script.
I do programming myself, but not much perl, and I did look at the code -
but it's not particularly readable to the untrained eye. So I figure it
would be best (read: quickest route to a solution) to ask one of you to
do it and offer compensation. I would like this done sooner than later
(or I'd just do it myself).
Please let me know if you are interested and able.