[ietf-dkim] Draft summary of SSP functionality

[Dave Crocker]

Michael Thomas wrote:
>   I can't speak for Spamassassin, I'm not one of their developers
>   and I don't know the intricacies of how they choose values. It
>   appears that you're just trying to be argumentative here since
>   you've seemingly dismissed the larger point that anti-spam filters
>   rarely use individual pieces of information as a silver bullet.


You think it is argumentative that, for example, a protocol element that 
directs a recipient to Reject mail will typically be adhered to and that 
conforming to SSP but not adhering to that explicit request is unlikely?

This sort of heuristic approach to protocol specification and interpretation 
usually has a poor outcome in Internet-scale services.

d/

ps. Another question that has solidly emerged is whether you can discuss this 
topic without ad hominems.

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net