whitelisting mailing lists with dccproc

On Tue, Sep 11, 2001 at 01:04:38PM +0100, Nicholas Piper wrote:
> 
> If I know I'm going to ignore the result I don't bother passing it
> through dccproc anymore.

Thanks for sharing your spam.  DCC only works if everybody submits
their spam for "counting".  It doesn't matter if *you* know it's spam
and therefore don't *need* to have DCC tell you that it is.  How about
contributing the spam to DCC for the rest of the community?

> If I do pass it through dccproc, I check for
> above 20 and/or many flags. If it finds it is bulk mail, I set another
> header than Mutt will later use.

Well, you will only get a >20 metric if 20 other people pass their
spam through DCC, whether they know it's spam or not (before passing
it through DCC).

If I get a spam and I know it's a spam, I still run it through
"dccproc -t many" so that others who get the spam *after* me will
be able to identify and pitch it.

> Ah. From reading about the whitelisting feature in the documentation I
> was given the impression it was best to *not* report checksums of
> solicited bulk mail (I probably thought this because it mentions it
> can be used as a "privacy" tool; to not even realise the checksums of
> my private and known-to-be-ok mail).

IMHO I don't think so.  I think everything should be sent through DCC.
What if you just happen to be the one or two people that opted into a
marketing list that is also being populated by address scraping?
Whitelisting should not be used to suppress DCC counting but should be
used to suppress DCC-metric-checking (if you know you want it).

b.

-- 
Brian J. Murrell