"$MTP" and Extra Added Crypto(TM) (was Re: Throw away the Internet and start over?)

R. A. Hettinga rah at shipwright.com
Wed Apr 23 10:36:25 PDT 2003

Hash: SHA1

At 2:04 PM +0100 4/23/03, Justin Mason wrote:
>What about:
>  - spammers impersonating your friends, using forged To and
> Received 
>    headers?
>  - your friends' mailing list, which sends mail as
>    fork-bounces at lists.xent.com?
>  - spam sent to the aforementioned list?
>These ideas (cf. also hashcash, whitelisting in general) generally
>run into those issues in SMTP -- namely there is no usable way to
>associate a sender with an authenticated identity in current mail
>use, unless
>you can persuade everyone to adopt S/MIME or PGP/GPG.

Yes. That's exactly the idea.

SMTP is conjoined, out of band, with lots of other protocols these
days, onerous and otherwise: SSL, DNS, MMX-lookups, and so on, mostly
to prevent spam. Some future "$MTP" would have to include some kind
of cryptographic authentication as a matter of course, probably hooks
for something out of band.

Frankly, if you required a signed message you'd probably kill 80% of
all current spam. If you required the encryption of each message to
the recipient's key, you'd vastly overload the outbound production
capacity of even the biggest spammers and kill 80% of the rest.
Including a blind-signature stamp, payable to the recipient, in a
signed and encrypted message would price all current abusers of
"free" email out of the market. You might get occasional unsolicited
commercial messages from people with legitimate economic
propositions, people for whom the return on the cost of postage is
worth something, but that's fine, compared to what you see right now,
and, frankly, the proposition of exchanges of competitive advantage
is what makes the world go round. Also, if your mail load from people
you don't know is too high, just you raise your price receive mail
from people you don't know. The resulting market for internet message
transit is end-to-end, geodesic, like the internet was designed to
be, and the sender, the person who creates the marginal cost, is the
person who pays that cost, as it should be.

All that, of course, requires real-time decryption, authentication
(check my keyring as a whitelist), and, for the resulting occasional
mail from strangers, an on-line redeem/reissue of any blind-signature
stamp attached to the message. Not as hard as it looks, and,
eventually, if not currently, less machine time than the process of
filtering *all* of my inbound mail for spam, sometimes requiring two
separate POP downloads, depending the filtering application. 

Like I said, this just for *individual* inbound SMTP, so it could
start small, in scope, and in market size, and work up from there:
Signatures (or encryption, whatever's easier), then signatures and
encryption, then signatures, encryption and postage. 

It would take some testing to see if the economics/processing power
were there, but I bet we're close, if not there already. 


Version: PGP 8.0 - not licensed for commercial use: www.pgp.com


R. A. Hettinga <mailto: rah at ibuc.com>
The Internet Bearer Underwriting Corporation <http://www.ibuc.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'

More information about the FoRK mailing list