All our eggs in one basket?

[Perry E. Metzger]

Jim Miller says:
> Perry Metzger writes
> 
> > What needed is a) the bank has to be able to show a third party a
> > signed request for every transaction they've performed, and b) you
> > have to be able to show a third party a signed (by the bank) receipt
> > for every transaction you've performed. In other words, you are
> > protected because the bank can't simply claim to the arbitrator "oh,
> > he withdrew all his money yesterday" because they can't show an order.
> > The bank is protected because you can't claim "oh, I deposited ten
> > million dollars yesterday" if you can't show a receipt.
> > 
> 
> 
> I'm still confused, only in a different way.  Let's let I want to withdraw  
> $10,000...
> 
> 
> 1) I send the bank a signed request to withdraw 10,000 dollars
> 
> 2) The bank withdraws the money but doesn't sends it to me.
> 
> I go to the arbitrator and say: "The bank cheated me!!"
> 
> The bank says: "We sent you the money.  Here is your withdraw request, signed
  
> by you.  You are lying."
> 
> ------
> 
> How can I prove that the bank did not send me the money?  

You can't with the partial protocol I've described thus far, but
assuming that you have to do something like signing the digital draft
in order to spend it, the bank's failure to provide a signed copy of
the draft with your signature demonstrates that you haven't spent the
money. The arbitrator needn't care if you already got the cash -- he
can order the bank to send you another copy of the draft that they
sent you.

I must admit that I haven't worked out the protocols for this yet, but
from the sketches I've made I think a quite workable system is
practical. I don't think you can cover all forms of cheating by the
bank, but I think you can construct things such that if someone tries
to cheat you you can prove it.

Perry