RE: sa-learn question

[Paul Hutchings <pa...@mira.co.uk>]

Perhaps I'm along but users do dumb things.. I have a "spam" public folder
(we use Outlook) and usually whenever I look at the contents before
exporting and learning it I can bet money there will be virus alerts,
undeliverables, stuff they accidentally dragged in there etc..

I guess it depends on your structure but I prefer a manual process even if
it means it happens a little less often.

Paul

> -----Original Message-----
> From: kler [mailto:kler@pcm3153.dti.supsi.ch]
> Sent: Wednesday, June 23, 2004 2:23 PM
> To: spamassassin-users@incubator.apache.org
> Subject: sa-learn question
> 
> 
> Hi all,
> I hope someone can help me. I've searched google without much 
> chance ...
> 
> On a mail server is installed SA 2.63 and we have set up a 
> global bayes
> db. We would like to allow users who get spam or ham messages to send
> them back the messages to us and run sa-learn through. 
> Is there a way to make sa-learn analyze only the body of the mail?
> (Since the header part comes modified?). 
> Or is there a better solution?
> 
> Any suggestion is well accepted!
> Thanks,
> Chiara
>