>To connect to server side filtering, the filtering engine on the
>server just needs to put probabilities it thinks that the message is
>spam in the headers, as well as have a standardized means for the
>client to report spam or ham. This doesn't seem that complicated: just
>a double and some sort of forwarding info to get the backchannel.
>(This assumes naive Bayes as a filter design)

Keeping in mind that upwards of 90% of mail is spam, you're going to
be downloading an order of magnitude mail if you do the filtering on
the end device.  On a desktop with a cable connection that's probably
OK.  On my phone, it's not.

>True: how much does DKIM+sender based blacklists do vs. filtering
>based on content?

In terms of volume, IP blacklists are still by far the most effective,
since they knock out most botnet spam.  Other than DMARC, which is a
separate can of worms, I don't know of anyone who does message
rejection based on DKIM signatures.  There's a whole lot of body
filtering going on.

R's,
John

_______________________________________________
Endymail mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/endymail

Reply via email to