spamass-milt-list
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Feature inquiry


From: Todd Lyons
Subject: Feature inquiry
Date: Thu, 4 May 2006 11:26:00 -0700
User-agent: Mutt/1.5.12

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On our system, we have about 40% of inbound email that gets delivered to
real mailboxes, and the remaining 60% are addresses that get forwarded
to customers regular mailboxes, such as @aol.com, @gmail.com, etc.  My
long term goal is to be able to do lax spam checks for email that will
be locally delivered and very VERY strict spam checks for forwarded
mail.

What does this achieve?  I mark as spam when the email has a score of 5.
Currently, I also block at the smtp level whent he email has a spam
score.  I want to change this behavior slightly.  I want to do rejection
at the smtp level with a spam score of somewhere around 5 for forwarders
and somewhere around 10 for locally delivered email.  This allows me to
accept marginal mail for local users and deliver to their Spam folder,
while still rejecting for forwarders.  (This keeps me from running afoul
of AOL's system, where users don't seem to understand that "Report as
Spam" is not the same as "Delete".)

I was thinking there were a few ways that I could do this:
1) Configure spamass-milter to do an LDAP lookup to see if the email
address exists as a posixAccount objectClass or account objectClass or
$objectClass.  This seems a bit too narrow though.
2) Configure spamass-milter to call a script which exits with a 0 for
success or any other value for failure.  The script would have to do
some sort of LDAP or PAM lookup to classify the email address.  This
seems relatively easy, but is laced with security issues in calling
anything external.
3) Build my own milter which can do some sort of LDAP or PAM lookup and
insert a custom header so that spamassassin can then adjust scoring.
This implies that it must occur before spamass-milter in the chain of
milters.
4) Build my own milter which can do some sort of LDAP or PAM lookup and
take action based on the spamassassin score in the headers.  This
implies that it must occur after spamass-milter in the chain of milters.
5) Other suggestions?

- -- 
Regards...              Todd
They that can give up essential liberty to obtain a little temporary 
safety deserve neither liberty nor safety.       --Benjamin Franklin
Linux kernel 2.6.12-18mdksmp   2 users,  load average: 0.05, 0.07, 0.07
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFEWkc4Y2VBGxIDMLwRAkl3AJ9PLQcWeqVFNGf8qeJFD45Id1Y5IACeN+t1
05zxBAMb0ao8vMQ74Cak9ig=
=0rqP
-----END PGP SIGNATURE-----




reply via email to

[Prev in Thread] Current Thread [Next in Thread]