On Sat, 2007-06-16 at 17:01 -0400, Michael B Allen wrote:
> Hi,
>
> I just setup a new server with vanilla SA


What version?

> on CentOS 5 and a lot of obvious
> drug/stock/foreign stuff is getting through. I have verified that DNSBL
> is being used. In general, I would like to know what the prevailing
> wisdom is as to increasing the agressiveness of my filter.


Add the SARE rules. They tend to kill most of the drug and stock stuff.
>
> Are there certain plugins that I need to make sure are working? If so
> what are they?


That depends.
>
> Will SA get better as it considers the input?


If you have bayes enabled.
>
> Also, if I drag spam from the inbox into the Spam folder, will SA learn
> from that? If I drag non-spam out of the Spam folder will SA learn
> from that?


That's up to your MUA , but not likely.

> Is there a way to add the X-Spam-Report to regular messages for a while
> so that I can see exactly why it's getting through?

Yes. See
http://spamassassin.apache.org/full/...agging_options

>
> How do I properly activate filtering based on character encodings used
> in messages? Basically I want to severely penalize non-Latin1 encodings.


In 3.1.x, just set ok_locales en
in 3.2.x, set ok_locales and also enable the Textcat plugin.

Details in
http://spamassassin.apache.org/full/...nguage_options
> Mike

--
Daniel J McDonald, CCIE # 2495, CISSP # 78281, CNX
Austin Energy
http://www.austinenergy.com