[Gllug] Spamassassin/spamd problem

[Adrian McMenamin]

On Sun, 2004-12-19 at 23:53, Nix wrote:
> On Sat, 18 Dec 2004, Adrian McMenamin moaned:
> > Somewhere along the way I have upset my spamassassin daemon because
> > although it still reports autolearning in the header eg...
> > 
> > X-Spam-Status:  Yes, hits=16.4 required=4.5
> > tests=HTML_70_80,HTML_FONT_BIG,
> > HTML_IMAGE_ONLY_04,HTML_MESSAGE,MIME_BOUND_DD_DIGITS,
> > MIME_HTML_NO_CHARSET,MIME_HTML_ONLY,MIME_HTML_ONLY_MULTI,
> > NO_INVESTMENT,PORN_4,RCVD_NUMERIC_HELO,X_MESSAGE_INFO autolearn=no 
> 
> Er, that means `I didn't autolearn'.


I had worked that one out for myself. What I mean is that the potential
for autolearning is reported but never acted upon.
> 
> > version=2.64
> 
> You might want to upgrade to 3.02 at some point.
> 
> > (Spammers, don't cha just luv em?)
> 
> Er, no? :)
> 
> > It doesn't matter how high or how low the score is, nothing is
> > classified either as spam or ham.
> 
> Well, um, that mail was classified as spam.

I meant in terms of bayesian filtering.

> 
> Bayes didn't seem to kick off, but it may just have classified things as
> 50% instead: SA 2.64 didn't differentiate between `error' and `can't
> learn', so we can't tell from this.
> 
> Could you run the message through `spamasssin -D' and post the result?
> 
> > I can't see anything obvious in any of the config files and so i wonder
> > if there is a heuristic out there that can help :(
> 
> Ask on the spamassassin-users mailing list?
> 
> -- 
> `The sword we forged has turned upon us
>  Only now, at the end of all things do we see
>  The lamp-bearer dies; only the lamp burns on.'

-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug