From: Hal <hfinney@shell.portal.com>
Date: Wed, 1 Nov 1995 03:33:11 +0800
To: cypherpunks@toad.com
Subject: Re: ecash remailer
In-Reply-To: <199510311625.IAA07105@infinity.c2.org>
Message-ID: <199510311703.JAA14785@jobe.shell.portal.com>
MIME-Version: 1.0
Content-Type: text/plain


sameer <sameer@c2.org> writes:

>To provide payee anonymity:

>Enterprising cypherpunk, Ed sets up the Ecash Remailer.

>Alice pays Bob e$15. Alice is anonymous.
>Bob sends Ed the e$15
>Ed cashes the e$ into his ecash mint account, withdraws e$13.50
>then pays Bob those e$14 .. Bob can now spend those e$ at will.

>	Bob is now anonymous.

>	This requires some trust-in-Ed, but Bob could be anonymous
>from Ed just as Ed is anonymous.

>	Perhaps this is the scheme lucky mentioned. I will have to get
>an account and implement it.

I think this is basically the scheme Lucky mentioned.  A more elaborate
version would have Bob sending Ed blinded proto-coins to be used in the
withdrawal.  However this would require hacking the ecash protocols to
work differently than intended, which would probably infringe the
patents.

What about this, though: Alice did not mean to pay Bob, but rather
Charlie, and Bob stole the coins.  He launders them through Ed's
service.  Charlie never got the cash, and Alice complains to the bank
that the coins were stolen.  The bank says, fine, we can identify the
perpetrator, let's see... it's Ed.  Ed is now charged with theft and
has an expensive and uncertain legal experience ahead of him.

Are you sure you want to put yourself in this position?  You might win,
but it could still be expensive (ask PRZ).  And if your service is seen
as a fencing operation to receive stolen goods with the legitimate uses
just a "cover", you could lose.

Also, I believe in normal use Digicash coins are marked as being for a
specific recipient.  This is not certain since no details have been
released.  And apparently it can be worked-around by the spender by
marking the recipient as just "@" (or some such string).  If this feature
is present in the Mark Twain cash then the payee-anonymity service may
not be very effective.

Hal