"Hallam-Baker, Phillip" <pbaker(_at_)verisign(_dot_)com> wrote:
The fact we perform 5 billion transactions with no charge mechanism in an
entirely decentralized fashion does not provide an existence proof for
being able to resolve 5 billion charge transactions.
One way to lower the number of messages is to aggregate the costs
per ISP over time. e.g. Rather than accounting for each transaction,
account for N transactions between you and ISP X over Y hours.
The real charges can then be based on differentials, and not on
absolute numbers. "You're charging me $1000 for my mail, but I'm
charging you $900 for mail you sent to me, so I've only got to pay you
$100."
The end result of this system, though, ends up looking a lot like
accreditation, where the accreditation is "ISP is/is-not
revenue-neutral to the rest of the net." i.e. has a balanced flow of
traffic, OR pays his bills on time.
We already know that spammers hide their traffic on zombie machines,
so this system won't let us bill the spammers. Rather, it lets the
spammed recipients bill the ISP originating the spam for not fixing
their system. The ISP is then free to push those bills onto their
customers, who are running zombied machines.
That's probably the one real utility of any e-postage system: Permit
better accounting of the cost of spam.
Alan DeKok.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg