From: Scott Brickner <sjb@universe.digex.net>
Date: Sat, 4 Nov 1995 21:16:47 +0800
To: lvhove&amp;vnet3.vub.ac.be@vnet3.vub.ac.be (Leo Van Hove)
Subject: Re: Info needed on observers
In-Reply-To: <v01530501acbea2598fe2@[134.184.16.171]>
Message-ID: <199511032049.PAA13455@universe.digex.net>
MIME-Version: 1.0
Content-Type: text/plain


Leo Van Hove writes:
>I am currently preparing a paper on payment systems for the Internet
>and related issues.  I recently found out that so-called observers could
>be placed on smart cards (in order to prevent double spending, or even to
>prevent people from cheating on taxes...).
>
>To be honest that's about all I know about observers and I am very eager
>to find out more.  I am, however, having a hard time finding interesting
>references on this subject.
>
>Hence, all suggestions to get me started would be extremely welcome.
>Questions of particular interest are: possible functions (and limits),
>how 'real' are they already?, what about privacy?, ...

With ecash it's possible to duplicate a coin and spend it twice ---
it's not really any different than copying a file and printing each
copy, after all.  In most ecash schemes, the double-spender can be
detected and identified when the coins are deposited, but this could be
long enough for him to disappear, leaving the bank or the payee holding
the bag.

An "observer" is designed to create a prior restraint to
double-spending.  The observer is a tamper-resistant smart card that
must be involved in all coin transfers.  When the coin is withdrawn
from the bank, the observer's signature is needed to validate it.  When
the coin is spent, the *same* observer's signature is again needed to
validate it.  The observer won't sign the same coin twice during the
spend protocol.

To double-spend with an observer present, you'd need to bypass the
tamper resistance.