On Thursday 16 November 2006 12:59, Anthony Peacock wrote: > I realise that DCC is not a direct indicator of spamminess but an > indicator of bulkiness. And I also realise that the correct answer to > my question is 'it depends on your local needs'... > > Given that what are people's thoughts on using DCC in SA? > > DCC gives a high hit rate on SPAM here, but also contributes highly to > false positives. Since setting up DCC I seem to have lots of list > emails reported as false positives, and spend a fair amount of time > checking and tweaking whitelisting settings for these. And in most > cases a combination of DCC and a highish Bayes score is enough to tip > these over. I know I could adjust the DCC score, but was wondering what > other people do?
The thing with DCC is that it combines checking and reporting, which is why it
is an indicator of bulkiness and not spamminess, as you say. To get around
that you should whitelist all mailing lists so that mailing list mail isn't
checked against DCC, both to avoid false positives yourself and to help
others avoid false positives.
So basically you're right and I haven't added anything. What I can add is that
I don't use DCC myself, for precisely the aforementioned reason, i.e. that it
requires to much fiddling with mailing lists.
--
Magnus Holmgren [EMAIL PROTECTED]
(No Cc of list mail needed, thanks)
pgpXF7edCj7oc.pgp
Description: PGP signature
