Re: PR Problem?

On Mon, Nov 13, 2006 at 05:10:49AM -0500, Adam Back wrote:

> The recipients filter wont know how long the recipient list is and
> cant detect reuse across different recipients.

I agree that it can't detect reuse across multiple systems. I'm not sure
that's the showstopper you think it is if you're checking stamps at the
MTA. If you could maintain a site-wide register of received stamps, then
you'd be assured that only one MTA transaction per unique message +
message stamp would take place.

As for the recipients list, the recipient *does* see the To: and Cc:
headers, and could determine how many names were in each. What you can't
see are Bcc: headers, which is why they should have a higher minting
cost.

Anyway, I'm not wedded to the idea. What I'm really trying to get at is
the need for an RFC-like plan for how to implement sensible hashcash
header checks--a best-practices document, as it were--and the need for
hashcash to generate appropriate stamps based on the recommended
policies.

I only got on my hobby horse because the OP was asking why hashcash
doesn't have more uptake. My comments weren't intended as criticism of
the previous work done, or of the technical merits of the ideas behind
it. 

-- 
Unabashedly littering the information superhighway with detritus like
this for over 15 years now.