You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Jason Frisvold <xe...@gmail.com> on 2007/05/09 17:32:43 UTC

required_score not working?

Greetings,

Am I correct in saying that the "proper" way to set a default
required_score is in the /etc/mail/spamassassin/local.cf file?

I'm running SA 3.2.0 and I seem to be unable to change the default
required_score.  I'm using spamc/spamd for processing via simscan.
Per-user rules are enabled and seem to be working fine.  If the user
does not have any preferences set, however, I thought the default
required_score from the local.cf file was used.  Is that not the case?

spamd is run with the following options :

-d -q -x -m25 --min-children 10 --min-spare 5 --max-spare 10
--socketpath=/tmp/spamd.sock --syslog-socket=none


I'm attempting to lower the default required_score because I'm seeing
a ton of incoming spam that's flying just under the default score of
5.0.  Per-user scores are not working because simscan drops to
defaults if there is more than one rcpt_to, and most of the incoming
spam is BCC spam.

Speaking of which, is there any sort of BCC rule that pumps up the
score if the mail is BCCed?  I can see a problem with mailing lists,
though....  Would it be possible to trigger on the mailing list
headers as well?


Thanks,

-- 
Jason 'XenoPhage' Frisvold
XenoPhage0@gmail.com
http://blog.godshell.com

Re: required_score not working?

Posted by Matthias Leisi <ma...@leisi.net>.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'll just answer the second of your questions and leave the first part
to those that know better how to diagnose ;)

Jason Frisvold wrote:

> Speaking of which, is there any sort of BCC rule that pumps up the
> score if the mail is BCCed?  I can see a problem with mailing lists,
> though....  Would it be possible to trigger on the mailing list
> headers as well?

It seems to me that the risk of false positives for just scoring on Bcc:
is pretty high. However, I have some rules for *combinations* of
addresses -- if these recipients appear together in the same SMTP
transaction, I add considerable points.

And yes, I verified with those concerned that they do not have anything
in common business-wise ;)

- -- Matthias

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFGQfrzxbHw2nyi/okRAoGhAJ9Rq4i/LvHNlsQTK7YA9Rc+lmEaBQCdHj71
2btfoAVHTEqC9l6YhlDrb1w=
=37tY
-----END PGP SIGNATURE-----