Keith Moore wrote:
[summary of approaches...]
which ones am I missing?
Hashcash?
http://www.cypherspace.org/hashcash/
the idea is that a sender has to compute a token which takes some
tunable amount of cpu resources, and the recipient can verify quickly.
The recipient will then delete/bounce/filter according to the presence
or absence of such tokens.
See also CAMRAM http://www.camram.org which is an anti-spam proposal
based around hashcash and signatures.
The obvious extension to reduce the CPU requirements is to only
require hashcash tokens on first mail to a new recipient that has
never replied to a mail from you (an unsolicited mail). A reply then
automatically includes a reusable token allowing subsequent messages
to be sent without further tokens.
The recipient can verify the token locally without reference to any
remote databases. The tokens are not reusable as they are only valid
for the recipient. Then tokens can not be used to send multiple
messages to the same recipient because his hashcash aware client will
reject them as double spent.
Adam
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg