[ietf-dkim] New DKIM threat analysis draft

[Earl Hood]

On October 10, 2005 at 10:07, "Hallam-Baker, Phillip" wrote:

> I think this problem goes away if we understand the purpose of DKIM to be 
> to allow parties to provide the recipient of an email with additional 
> information that allows them to make a more efficient and/or more 
> accurate determination of whether they are willing to accept it.

What is "the additional information?"  As of now, the additional
information has not been clearly defined: is it just a domain
accepting responsibiliy, is it protecting originating identities,
or something else?  Past threads have debated on DKIM's scope, but
I have not seen any firm resolution.

Past discussions tend to imply that DKIM mainly provides the ability
for a domain to claim responsibility for a message.  If this is the
case, I have not seen a good argument on how this will improve things.
For example, with reputation systems, will DKIM provide a better
(i.e. more effective) reputation system over existing ones (e.g.
network address-based) justifying the cost of deploying DKIM?

--ewh