spam/notspam w/sa-learn

Tue May 6 16:05:22 IST 2003

The improvement that we are discussing is the ability to merely allow any
mail user to
FORWARD a message to spam or notspam and have it processed through
SpamAssassin's sa-learn
script and treated accordingly.

The original problem, as Julian stated, is that Outlook, Outlook Express
remove the headers
when a message is forwarded so instead of the message being marked as spam
or ham, depending
on which email address (spam/notspam) you sent it to, it would incorrectly
mark the end user.
This is not the desired affect.

I was mulling over the possibility of any user sending a email to either
spam or notspam which
merely contains an address (or series of addresses).  Then a script runs
which processes this messages
through the sa-learn script which, in turn, teaches SpamAssassin about them.

Unfortunately, although I pretty good in C and PHP, I'm not much of a Perl
wizard.
Perhaps what I describe above is exactly what your perl script does?

If so, why not share it with the rest of us so we don't have to re-invent
the wheel :)

Many thanks.

N

-----Original Message-----
From: Andrea Cogliati [mailto:AndreaC at GOTECH.IT]
Sent: 6-May-03 4:14 AM
To: MAILSCANNER at JISCMAIL.AC.UK
Subject: Re: spam/notspam w/sa-learn

Kevin,

I'm currently using Exchange 2K Public Folders exactly for this purpose
and I wrote a very simple Perl script for getting spam&ham from there.

What 'improvement' are you talking about? Am I missing something?

Bye,

Andrea

-----Original Message-----
From: Kevin Spicer [mailto:kevins at BMRB.CO.UK]
Sent: Monday, May 05, 2003 9:26 PM
To: MAILSCANNER at JISCMAIL.AC.UK
Subject: Re: spam/notspam w/sa-learn

A couple of us on the list have been experimenting with ways of using
the 'public folders' feature in Exchange to get round this problem with
Outlook.  I understand this works with Exchange 5.5.  But due to an
'improvement' Microsoft made in Exchange 2000 it doesn't work with that
(which I personally find damn annoying). The basic way it works is by
creating two public folders for spam and ham then using a script on the
MailScanner machine to grab the messages using IMAP.

The most important thing when feeding messages to sa-learn is that the
message ID should not be changed (since sa-learn tracks which messages
it has learned using this).  Should the message ID change, which is
likely to happen when forwarding, SA may well learn the same message as
both ham and spam!

On Mon, 2003-05-05 at 19:33, Feasey, Nicholas wrote:
Hmm?  So then, by forwarding I would place my address in the
spam/notspam mailboxes and be identified as such. Any ideas on how to
"fool" the app so that it will work with Outlook or am I going to have
to try to write some sort of script - if, in fact, that is possible.

N
    -----Original Message-----
    From: Julian Field [mailto:mailscanner at ECS.SOTON.AC.UK]
    Sent: 5-May-03 2:22 PM
    To: MAILSCANNER at JISCMAIL.AC.UK
    Subject: Re: spam/notspam w/sa-learn

    At 19:16 05/05/2003, you wrote:

        Forgive my ignorance/stupidity on this subject but what do you
        mean by redirect?
    Different mail apps call it different things, but it is usually
    bounce or redirect. It sends the message on with the headers intact,
    so that replies go back to the original sender and not the person
    who redirected it.

    Outlook and Outlook Express cannot do it.
    Why, I don't know, it's the simplest job in the book :-(

        Yes, the cron job is in place, just want to give my users
        CORRECT instructions on using the "auto-learn" feature.

        N

            -----Original Message-----

            From: Julian Field [mailto:mailscanner at ECS.SOTON.AC.UK]

            Sent: 5-May-03 2:07 PM

            To: MAILSCANNER at JISCMAIL.AC.UK

            Subject: Re: spam/notspam w/sa-learn

            At 18:40 05/05/2003, you wrote:

                I just want to be clear on the usage of the spam and not
                spam accounts used in conjunction with the sa-learn
                script.

                I can have my users forward any message received to
                either spam (to have it identified as spam) or notspam
                (to have it identified as ham) and it will

                be automatcially learned (sa-learn)??

            It is important that they "redirect" and not "forward" their
            mail to the addresses, as forwarding will destroy the
            headers and make it appear that your users are the spammers.

                Is this correct?

            You do, of course, need my cron job script to do the actual
            work, mailboxes aren't magic :-)

                I don't want my users showing up as spammers because
                they are listed as the forwarder.

            --

            Julian Field

            www.MailScanner.info

            Professional Support Services at www.MailScanner.biz

            MailScanner thanks transtec Computers for their support

    --
    Julian Field
    www.MailScanner.info
    Professional Support Services at www.MailScanner.biz
    MailScanner thanks transtec Computers for their support

BMRB International
http://www.bmrb.co.uk
+44 (0)20 8566 5000
_________________________________________________________________
This message (and any attachment) is intended only for the recipient and
may contain confidential and/or privileged material.  If you have
received this in error, please contact the sender and delete this
message immediately.  Disclosure, copying or other action taken in
respect of this email or in reliance on it is prohibited.  BMRB
International Limited accepts no liability in relation to any personal
emails, or content of any email which does not directly relate to our
business.