1994-08-26 - Re: In Search of Genuine DigiCash

Header Data

From: rah@shipwright.com (Robert Hettinga)
To: Jason W Solinsky <solman@MIT.EDU>
Message Hash: d56cc60e4a050124d27c21925c987152d51f29bcf2357a6af971d841ffe854bc
Message ID: <199408260050.UAA04812@zork.tiac.net>
Reply To: N/A
UTC Datetime: 1994-08-26 00:51:56 UTC
Raw Date: Thu, 25 Aug 94 17:51:56 PDT

Raw message

From: rah@shipwright.com (Robert Hettinga)
Date: Thu, 25 Aug 94 17:51:56 PDT
To: Jason W Solinsky <solman@MIT.EDU>
Subject: Re: In Search of Genuine DigiCash
Message-ID: <199408260050.UAA04812@zork.tiac.net>
MIME-Version: 1.0
Content-Type: text/plain

> Making a digital cash system secure, scalable and distributed
>is a non-trivial task, making it anonymous is still more difficult.

But it's done already, right? At least as far as scoping out the problems
and having reasonable solutions which now need to be tried out, I mean.

If we're talking about problems of marketing, access to and from the rest
of the financial infrastructure, enforcement for fraud (double spending), I
think the problems are manageable enough for a market test, and I've done
my best to talk about that here and elsewhere.

If you're talking about algorithms, they're done already. They're to the
point where people are writing better ones that do the same thing more
efficiently. If you're talking about the operations stuff (handling cash,
keeping track of unspent cash, identifying offenders, etc) I contend that
the code is being written now or has been written already.

>Guaranteeing anonymity creates alot of problems as was brought out in a
>previous discussion on license based cash in which it was pointed out
>that by colluding with consumers a bank can still "mark" bills.

I am under the impression that technology exists that allows anonymous
offline digital cash transactions where double spenders are identified. Is
that not the case?

>I am yet to see a single anonymous digital cash system which could not be
>implemented more simply if the requirement on anonymity were not made. I
>would be pleased to be proven wrong.

How simply? Are the computational requirements for generating a piece of
anonymous offline digital cash, for discovering double spenders, for
holding and spending cash Chaum's digital cash, or, say, Brand's version,
excessive when compared to non-secure methods, especially on-line methods?
Why do you suppose these guys are spending money developing the stuff?  Are
they wasting their time? I've heard sums in the range of at least $10
million that Chaum has spent already...

>What I'm really asking is for an example of this overhead that is being

First of all, I'd like you to talk a little about offline digital cash
systems which are non-anonymous, and have all the other features of Chaum
or Brands. Pointers to the information is fine, unless others want to see
it here.

If you're talking about *any* online system, I've got you hammered. I just
won that fight on another list (or at least I think I did ;-)), and I'll
dump it all to you through e-mail, if you want. I'll give you hint: my
jumping off point was Eric's observation that offline methods don't involve
the banking system to effect every transaction.

>> There are other reasons
>> for not doing on-line transactions. Including credit checks, interest
>> calculations on outstanding balances, vendor reserve requirements,
>> transaction threading, on-line wait states and bandwidth, etc.  It's
>> considerable.
>And its going to get more considerable when we have communities of agents
>arguing with each other. I think we want to solve the problems created by
>these requirements, not shy away from them.

Agreed, but I don't think avoiding them all together by doing transactions
offline is shy, I think it's economically necessary.

Bob Hettinga

Robert Hettinga  (rah@shipwright.com) "There is no difference between someone
Shipwright Development Corporation     who eats too little and sees Heaven and
44 Farquhar Street                       someone who drinks too much and sees
Boston, MA 02331 USA                       snakes." -- Bertrand Russell
(617) 323-7923