MailScanner SpamAssassin issue

[Ugo Bellavance]

<x-flowed>
augustin siaens wrote:

> Le lun 16/08/2004 à 16:06, Ugo Bellavance a écrit :
> 
>>augustin siaens wrote:
>>
>>
>>>Dear all,
>>>
>>>I face the following problem: SpamAssassin works fairly well in command
>>>mode (when I do debug with -t -D and a spam.txt) but that it fails to
>>>scan something when using in the system with MailScanner. I've tested my
>>>MailScanner config and I didn't find anything weird.
>>>
>>>here's a report I get when using the command line and the debug mode.
>>>
>>>The same e-mail pass trough without any problem if I use the e-mail
>>>system. No header modification.
>>>
>>>Do you have any idea of the origin of the problem?
>>
>>You have Use Spamassassin = yes?
> 
> Yes, it is ok on that point. Apparently SpamAssassin may be working after all but the scores are too low (even with a 5.0 treshold)

This probably mainly because bayes is not used whey you run SA from 
Mailscanner.  Once it will have seen 200 hams and 200 spams it will 
start working and you'll see good scores.

You can try the bayes starters DB at www.fsl.com/support.

And maybe see the optimization section of the Most Asked Questions page.

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).

</x-flowed>