At 7:20 PM -0500 2002/08/21, J.A. Terranson wrote:
Presenting a computationally difficult problem to a connecting MTA moves the requirement for the CPU power to the sender while keeping the recipient site unfettered. Let's face it, the spam problem is merely one of cost shifting from sender to reciever, and this proposal shifts the load back. Any site that wishes to maintain the current system of email subsidies to the sender domain need only provide a computationally simple token.
Now this is more plausible. You'd still need something akin to a PKI to distribute the computationally simple tokens, and you'd need a way to easily revoke them. But if this was implemented by default in the standard MTAs, you would go from hundreds or thousands of message deliveries per minute to five or more minutes per un-authenticated message delivery. This is something that might be worth discussing in the appropriate forums, such as the SMTP-related working groups of the IETF. -- Brad Knowles, <brad.knowles@skynet.be> "They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania. GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E W+++(--) N+ !w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++) tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)