[EMAIL PROTECTED] wrote:
>
> > It's "MicroMint", by the way, Adam.
>
> Doh! I know that. (Excuse the braino -- the landscape is littered
> with dozens of epayments candidates with similar vaguely currency
> related names, to the point of name confusion).
>
> > Which, besides the fact that it's the only currently available
> > method for generating economically non-forgeable streaming bearer
> > coins at transaction sizes of $0.001, (a 10th of a penny for those
> > of you Rio Linda) or less, is why we're looking at using it at the
> > Internet Bearer Underwriting Corporation.
>
> It could be that there are patent issues I am not considering. But
> for example in the same paper that they present MicroMint, Rivest &
> Shamir discuss PayWord, which sounds a much more economically viable
> system using signatures and hashchains. You can tune the size of the
> signatures down if individual signed values are very low value, buy
> hardware accelerators, optionally use elliptic curve to speed things
> up perhaps.
As I calculated for EFCE, 512-bit Lucre signatures _in Java 1.1_ on a
P2/300 cost .00004p each to generate (naturally we don't care how much
the blinding costs, because the mint doesn't have to do it, but its of
the same order, of course). I'd say that was economically non-forgeable
for $0.001 transactions, and a damn sight more robust than MicroMint.
Even at 1024 bits, $0.01 is economic. And they aren't going to become
vapour-money in the forseeable future.
Cheers,
Ben.
--
http://www.apache-ssl.org/ben.html
Coming to ApacheCon Europe 2000? http://apachecon.com/
