On Thu, May 03, 2001 at 02:33:38AM -0400, Tavin Cole wrote:
> On Wed, May 02, 2001 at 11:21:40PM -0700, Dev Random wrote:
> > What about a Whitelist plus HashCash for people not on the list?
> > The recipient could specify how much HashCash they want.
> >
> > Also, I've been thinking of an extension to HashCash which I call
> > TuringCash. It has the advantage that it would be much harder to
> > accelerate with specialized hardware.
> >
> > TuringCash works like this. Specify a little virtual machine with a very
> > clean instruction set. The email recipient publishes the algorithm to
> > run and the expected running time. The output of the algorithm must
> > be submitted (just like the hash collision in the case of HashCash).
> > The virtual machine sandbox would be provably secure (not hard with a
> > clean instruction set).
>
> Interesting. Can you give an example of such an algorithm?
HashCash itself would be such an algorithm. The point is that any
algorithm would be suitable. If someone manages to optimize for a
particular algorithm, just change it (with only a configuration change).
I feel TuringCash is a bit off-topic. Food for thought...
I think a HashCash implementation would be a good start for spam
protection and resource usage control in general.
PGP signature
