[Cryptography] What is total world transaction volume?

[Peter Gutmann]

James A. Donald <jamesd at echeque.com> writes:

>Bitcoin cannot replace government money because there is a soft limit on the
>number of transactions, and we are already hitting that limit.
>
>Also because final settlement is a bit slow.

"A bit slow" is an understatement.  About a decade ago the Visa/MC payment
gateways were doing 5K transactions a second, which included ignoring some of
the crypto verification that was being used because it couldn't be processed
in time.

Then there's the case of ignoring the CVV2/CVC2, because most of the time when
it's wrong it's because the cardholder mistypes it or tries to guess it and
gets it wrong rather than fraud, or ignoring the ARQC because the software
stack gets it wrong, and what you lose in fraud is a fraction of what you lose
in rejecting otherwise valid transactions.

Some years ago during a discussion of another (proposed) e-payment system, a
banking guy joked that if they adopted it (even at the low volumes being used
back then), they'd be able to accept transactions for an hour and then have to
shut down and spend the remaining 23 hours processing them all so they could
have their next transaction acceptance hour the following day.

So you need a payment system that can run at the rate of (say) 10K/second and
that has a tunable level of fault-tolerance in order to avoid rejecting valid
but slightly incorrect transactions.

A small amount of fast symmetric crypto and credit cards is the closest we've
got to meeting that requirement.

Peter.