[Cryptography] A little history, was What is total world transaction volume?

Patrick Chkoreff patrick at rayservers.net
Wed Feb 8 18:59:06 EST 2017


John Levine wrote on 02/08/2017 05:13 PM:

> Why anyone would want to return to this ridiculous way of using
> money baffles me.

If and when people care enough to try scaling up bitcoin transactions to
10000 per second, I'll take a guess at the default solution they will
evolve:  web wallets, which settle instantaneously because they use
standard book entry.

In other words, people will start spending liability currencies issued
by web wallet providers (e.g. Coinbase).  Those currencies will be
redeemable for actual bitcoin on the blockchain, with varying degrees of
probability.  The risk will be the technological, financial, or
fiduciary failure of the issuer.

I suspect that is how people are likely to address the problem.

Next, they'll want to make a spend to someone using a different wallet
service.  So the wallet providers will create some kind of
cross-platform spend capability.  Then the wallet providers will have to
settle up between themselves periodically, on the actual block chain.

It sounds very familiar, like a repeat of history.  That's because the
same techniques which increased efficiency of settlement 700 years ago
will likely work again today.  The fact that the underlying asset is
immutable bit patterns instead of luggable atoms is irrelevant.

I was just looking for a way to increase the number of issuers in order
to decentralize the server/provider risk.  However, that might be
technologically or humanly infeasible, and not what the market wants.
The market might actually prefer to have only 237 major service
providers around the world, and not any grand uniform
hyper-decentralized approach, such as the generalized market for bills
of exchange I described.

Actually, the market I described does NOT seem to relate very closely
with James Donald's proposal.  His proposal has to do with small locally
connected groups ("cliques" if you will) vouching for net balances among
them.  Those groups are defined entirely in terms of the small number of
transaction hops between them.  The idea does suggest a degree of trust
and mutual liability, which made me think of the Ripple-ish market for
bills of exchange, but I suppose you can't push the comparison that far.


-- Patrick



More information about the cryptography mailing list