I agree that the discussion on the merits/deficiencies of various accreditation
services belongs elsewhere, but this prompts another question. Once there is
at least some degree of trust in domain-level email source addresses, there
will be a need for some way to evaluate those addresses. Otherwise spammers
can just buy and use disposable domains (and advertise Sender-ID, of course).
This is what Meng is getting at in his "aspen" document.
I see a whole ecosystem of accreditation services (some new, some old)
springing up once MARID (and similar mechanisms, e.g., MASS) get adopted. I
expect they will have different business models and will use different criteria
for rating the addresses. Some might require the posting of a bond, others
might rate based on what other recipients of mail say (true reputation), and
others might be based on criteria like how long the domain has been in
existence.
It would be good to have common mechanisms for accessing this accreditation
data, so that the services can compete freely and recipients can easily use the
service that is best for them.
Does this make sense as a future work item for the (presumably rechartered)
MARID WG? Or does this sound like a separate WG? As mentioned above, MASS
will need the same thing.
-Jim