Rats. It does seem like everything getting through lately is scoring 0% in
Bayes. I was hoping it was just clever spammers. I guess it's time to
purge my Bayes database and start over.

--
Owen B. Mehegan (owen@nerdnetworks.org)

> On Tue, Aug 12, 2008 at 12:41:17PM -0700, Owen Mehegan wrote:
>> Here are two more that got through today. Even several hours later,
>> these haven't shown up in blacklists. Do anyone else's rules catch
>> these?

>
> Your main problem is that both messages hit BAYES_00:
>
>> X-Spam-Status: No, score=2.0 required=5.0 tests=BAYES_00,FREEMAIL_FROM,
>> HTML_MESSAGE autolearn=no version=3.2.1

> ...
>> X-Spam-Status: No, score=4.7 required=5.0
>> tests=BAYES_00,BOTNET_SERVERWORDS,
>> FREEMAIL_FROM,GEO_QUERY_STRING,HTML_MESSAGE autolearn=no version=3.2.1

>
> Versus:
>
> $ spamassassin -D check < spam1 > /dev/null
> [10048] dbg: check: is spam? score=7.001 required=5
> [10048] dbg: check:
> tests=BAYES_99,DKIM_SIGNED,DKIM_VERIFIED,HTML_MESS AGE,RCVD_IN_BL_SPAMCOP_NET
> $ spamassassin -D check < spam2 > /dev/null
> [10069] dbg: check: is spam? score=6.197 required=5
> [10069] dbg: check:
> tests=BAYES_99,DKIM_SIGNED,DKIM_VERIFIED,GEO_QUERY _STRING,HTML_MESSAGE
>
> --
> Randomly Selected Tagline:
> "Oh gee, there it is, too bad." - Prof. Farr
>