[apps-discuss] Comments/questions about draft-akagiri-mail-divide-00

["Kurt Andersen (b)" <[e-mail]>]

After the presentation about this proposal in APPSAWG this morning, I've
read through the proposal and have the following questions:

1) What is the benefit to recipients (not the receivers) of this system? It
seems that, at best, some coarse categorization of mail types can happen,
but I don't see how it is useful to me as a recipient to rely on a sender's
characterization of their email.
2) I don't see any benefit for receivers (the MSO/ISP who runs the mail
system). This seems like it causes them to have to deploy a bunch (up to
8x) of receiving systems.
3) As a sender, I also don't see how this provides any benefit. If I am a
good sender, then I will have a good reputation in any reliable reputation
service and if I'm malicious, this system seems like it gives me a variety
of options to try to subvert the process. Having different receivers
specify individual reputation services that they want me to certify with
does not seem scalable. How would a sender such as Yahoo even consider such
a proposal?

Even if this was limited to something like trying to automate/publicly
publish a "trusted sender" channel such as some receivers have available,
I'm not sure that scaling such a thing is compatible with the intent of
limited access.

--Kurt Andersen