[Mailman-Users] SPAM Control

[Chris Barnes]

Mark Sapiro <msapiro at value.net> wrote:
>
> If you want mailman to deal with it, you can use header_filter_rules
> to recognize the headers and act accordingly, but why not just delete
> the spam before it gets to mailman?

That would be fine for MailMan, but not for everyone else (MM exists on 
the main mail server with all the other users).  The users get their 
messages tagged, but not deleted. The reason for this is that no matter 
how good a spam filter is (or in our case - a combination of spam 
filters), there will occationally be the false positive.   Automatically 
deleting false positives can have disasterous effects.


I guess what I am asking is for a way to be MM to process a 
.procmailrc - specific to each list - before distributing to the list.


as an exanoke, this is what I put in my own .procmailrc to handle the 
spam header tags:

# We have 3 spam filters running on the server:
# 1) SpamAssassin (SA) adds the headers "^X-Spam-Level:" and 
"X-Spam-Status:"
#      and if the score is >5, adds "[Spam]" to the Subject line
# 2) BogoFilter adds the header "^X-Bogosity:" with values of "Yes, No, 
Unsure"
# 3) QSF adds the header "X-Spam:" with values of "YES, No".


# Begin recipes for table shown in
# http://physics.tamu.edu/services/comp_support/docs/AntiSpam.html

# Row 2, col 4
# all agree it is ham, no need for learning, but still need the rule
# so processing of this script stops.  Do this one first since 
(hopefully)
# most of the incoming email will match and speed the processing of this 
script
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\No
* ^X-Spam-Status:.\No,.\score=-
$DEFAULT

# Not in the table, but if SA scores higher than 8, put directly in Spam 
folder
# A score of 8 or more from SA is enough on it's own to be sure it IS 
spam
:0
* ^X-Spam-Level:.\*\*\*\*\*\*\*\*
.Spam/


# Row 1, col 1
# Bogo Spam, Qsf Spam, SA ham --> not sure
:0
* ^X-Bogosity:.\Spam
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\No,.\score=-
.Unsure/

# Row 1, col 2 & 3
# Bogo Spam, Qsf Spam, SA between 0 & 8 --> spam
:0
* ^X-Bogosity:.\Spam
* ^X-Spam:.\Yes
# no need to check SA score here
.Spam/

# Row 1, col 4
# Bogo yes, both SA and Qsf say Ham --> ham folder for learning
:0
* ^X-Bogosity:.\Spam
* ^X-Spam:.\No
* ^X-Spam-Status:.\No,.\score=-
.Ham/

# Row 1, col 5
# Bogo yes, SA 0-5, Qsf no  --> We're not sure
:0
* ^X-Bogosity:.\Spam
* ^X-Spam:.\No
* ^X-Spam-Status:.\No
.Unsure/

# Row 1, col 6
# Bogo yes, SA score is 5-8 (yes), put directly in Spam folder
# Qsf ranking doesn't matter if Bogo & SA are that sure.
:0
* ^X-Bogosity:.\Spam
* ^X-Spam-Level:.\*\*\*\*\*
.Spam/

# Row 2, col 1
# Bogo and SA are sure it's ham, Qsf Spam --> ham folder for learning
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\No,.\score=-
.Ham/

# Row 2, col 2
# Bogo Ham, SA 0-5, Qsf yes --> We're not sure
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\No
.Unsure/

# Row 2, col 3
# Bogo Ham, SA 5-8, Qsf yes --> We're not sure
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\Yes
.Unsure/

# Row 2, col 5
# Bogo no, SA 0-5, Qsf no  --> Ham for learning
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\No
* ^X-Spam-Status:.\No
.Ham/

# Row 2, col 6
# Bogo no, SA 5-8, Qsf no  --> Not sure
:0
* ^X-Bogosity:.\Ham
* ^X-Spam:.\No
* ^X-Spam-Status:.\Yes
.Unsure/

# Row 3, col 1
# Bogo Unsure, SA negative, Qsf yes --> Unsure
:0
* ^X-Bogosity:.\Unsure
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\No,.\score=-
.Unsure/

# Row 3, col 2
# Bogo Unsure, SA 0-5, Qsf yes --> more likely, but still Unsure
:0
* ^X-Bogosity:.\Unsure
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\No
.Unsure/

# Row 3, col 3
# Bogo Unsure, SA 5-8, Qsf yes --> sure enough to put in Spam
:0
* ^X-Bogosity:.\Unsure
* ^X-Spam:.\Yes
* ^X-Spam-Status:.\Yes
.Spam/

# Row 3, col 4 & 5 (together)
# Bogo unsure, SA & Qsf say ham --> Ham for learning
:0
* ^X-Bogosity:.\Unsure
* ^X-Spam:.\No
* ^X-Spam-Status:.\No
.Ham/

# Row 3, col 6
# Bogo unsure, SA 5-8, Qsf No --> Unsure (only SA says yes)
:0
* ^X-Bogosity:.\Unsure
* ^X-Spam:.\No
* ^X-Spam-Status:.\Yes
.Unsure/


--

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Chris Barnes                           AOL IM: CNBarnes
chris-barnes at tamu.edu                Yahoo IM: chrisnbarnes