TrustBucks

-----BEGIN PGP SIGNED MESSAGE-----
 
tcmay@got.net (Timothy C. May):
> You want to elaborate on this contention?
>
> If Alice transfers Digicash-type money to Bob, this is because Alice either
> bought the DC-money someplace, or already had it, or otherwise arranged
> with a bank to make the transaction. Maybe the bank "loaned" her money she
 
Mistaken terminology, mea culpa. The scheme I was thinking of is
actually called Private Currency. Someone mistakenly labelled the
writeup Digicash and I cut and pasted without thinking. I do know the
difference when my brain is on.
 
s/Digicash/Private Currency. Apologies to Digicash.
 
I'll explain Private Currency and why it's good and bad. In Private
Currency you don't "buy the [money] someplace". You mint it when paying.
Alice and Bob check each other's public debt and if neither is scared
off by the other's high debt, they mint a debt for Alice and money for
Bob. They publish a record of the transaction, which is how they knew
each other's public debt in the first step. So in theory the amount of
currency in existance is exactly 0. In practice I wouldn't trust anyone
for a debt that I didn't trust directly for that amount. I believe the
scheme would stall.
 
I conceived TrustBucks as an alternative that would retain the
decentralization but work.
 
> No protocols. No anonymity. No protection against double-spending.
>
> Looks promising. Keep us informed.
 
Fine. I doubt my mechanisms will be optimal but here you go.
 
Restraint on double-spending: Each participant publishes a list of the
    ID and value all outstanding TrustBucks of their own variety. Value
    of the notes can be obscured so it can only be verified by someone
    who has seen the note itself.
 
    What if some participant doesn't publish a complete list? Well, who
    are they robbing? People who directly trusted them for that amount
    and now won't ever again.
 
Anonymity: Each participant identifies their currency by a randomly
    chosen ID-number instead of name and publishes their ID/value under
    that number.
 
    However, it's pretty pointless, since the chain of trust has a hard
    time extending beyond people who directly know each other anyways.
 
 
A better objection would have been that it's hard to identify a chain of
mutually trusting links between two strangers who want to make a
transaction.
 
 
-----BEGIN PGP SIGNATURE-----
Version: 2.6.1
 
iQCVAwUBMgPoDJi7GCxryNrZAQHeEgQAmHsJithWMhmRv4y3IjnCBFKAgmZLCQ+i
NVYGDBVJ19iwAOTTwqHgcYMGEYdKBLUaBMRAczJDfGRbsB6WbFLKyiESHT8gpV7R
6CVesb7XpRaVDBylgTvoE/NNXfNrLrTfWOeVWtivMSVkDRKJC6BbONR1J5juhQjv
A9s1wa/uwSw=
=hsSY
-----END PGP SIGNATURE-----

An interesting idea.  It reminds me of a barter system, with the similar
problem of trying to put together a complex trade which is mutually
acceptable.  I wonder whether it could be automated if people posted
their holdings and what they would accept.  Then software could go into
this database and try to put together a set of trades that will let
someone make a purchase.  However it would seem to be very harmful to
privacy to have to post all this information.

There are some "lightweight payment" schemes out there which have the
property that people only accept cash that is "for them".  Sometimes
there is a broker involved who actually issues the cash on behalf of the
merchant (the merchant trusts the broker to do this) so that customers
need only go to a smaller number of brokers.  Then these systems can be
based on heavier payment systems like digicash or credit cards which
people use to open accounts with the brokers.

I do like the decentralization idea, but these lightweight schemes have
some of the same advantages.

Hal

At 19:11 -0500 8/3/96, TrustBuckFella wrote:

>I'll explain Private Currency and why it's good and bad. In Private
>Currency you don't "buy the [money] someplace". You mint it when paying.

IOW Electronic IOUs/Markers.

>Restraint on double-spending: Each participant publishes a list of the
>    ID and value all outstanding TrustBucks of their own variety. Value
>    of the notes can be obscured so it can only be verified by someone
>    who has seen the note itself.
>
>    What if some participant doesn't publish a complete list? Well, who
>    are they robbing? People who directly trusted them for that amount
>    and now won't ever again.

If the value is obscured there is still no verification of how much they
have outstanding. So long as all of the TrustBucks are listed (with the
amounts listed correctly but obscured), there is no way to verify that the
claimed total is accurate unless you monitor their list before the swap and
after it and there is only one new TrustBuck listed (with the correct
amount added to the outstanding total). Listing phony $0 notes on the list
and removing them later (and dropping the claimed outstanding balance)
would be possible. With this method, anyone who I give my note to will see
the correct amount in THOSE notes as well as seeing the outstanding balance
go up the correct amount but that amount can be manipulated (as I stated)
due to there being no complete disclosure of the amounts.