[2123] in cryptography@c2.net mail archive
Re: Something really new???
daemon@ATHENA.MIT.EDU (David Jablon)
Sat Feb 7 15:58:47 1998
Date: Fri, 06 Feb 1998 23:00:39 -0500
To: cryptography@c2.net
From: David Jablon <dpj@world.std.com>
Cc: Hal Finney <hal@rain.org>
In-Reply-To: <199802062249.OAA07789@s20.term1.sb.rain.org>
At 02:49 PM 2/6/98 -0800, Hal Finney wrote:
>Regarding ...
>> http://www.economist.com/editorial/freeforall/current/index_st4443.html
>
>Sounds like [...] a layman's description of a two-string
>oblivious transfer. This is a standard cryptographic primitive in which
>Alice has two strings, and sends them to Bob in such a way that Bob only
>receives one of them, of his choice, but where Alice can't tell which one
>he chose. [...]
Oblivious transfer, hmm ... I was thrown off the track by
the original posted claim that this is "surprisingly simple":
>> [...] he has devised=20
>> a surprisingly simple way for two people to find out if both
>> possess the same piece of information without telling each other=20
>> what it is.=20
>>=20
>> Unlike more conventional cryptographic methods, which involve the=20
>> use of computers and mind-numbingly long prime numbers (that is,=20
>> numbers which have no factors except themselves and 1), this one=20
>> requires=97in its simplest manifestation=97no more than some paper, a=20
>> couple of pencils and a good supply of envelopes. And it has another=20
>> advantage over conventional cryptography: it cannot be broken, even=20
>> in theory.=20
If one ignores the "simplest manifestation" using paper envelopes,
and assumes use of oblivious transfer, it's an interesting approach.
But it seems to require much more arithmetic than the
Diffie-Hellman-based alternatives.
Can anyone point to a real reference for this work?
------------------------------------
David Jablon
Integrity Sciences, Inc.
dpj@world.std.com
<http://world.std.com/~dpj/>
