Anonymous mail service up for alpha testing

-----BEGIN PGP SIGNED MESSAGE-----

	I've written a small anonymous mail service, and it's now
available for testing. There's no security, and I'll be keeping logs,
so don't think that it's secure, in any way. It's also running on a
PPP link which isn't connected all the time, so it's rather flaky.
	(I'll set it up as a real service once I get a real link-- if
anyone else wants to do it, they're welcome to use my code.)

	How to get an anonymous account:

	Send mail to admin@infinity.hip.berkeley.edu -- include in the
message a login, a "Full Name", a choice of remailer, and an encrypted
return address block encrypted with that remailer's public key.
	I'll set it up.

	How the anonymous account works:

	Someone will send mail to
login@infinity.hip.berkeley.edu. Then the system looks up in a table
which remailer is associated with that login. It then sends out mail
to that remailer, starting with the contents of the encrypted return
address block, then a "##" and then all of the message to
login@infinity, with "Received" lines taken out.

	Thus once the message gets to the last remailer of the chain
in the encrypted return block, the ## pasts the identifying
information of the person mailing to login@infinity.hip in the header
of the message. (It *should* do that...) If the person mailing to the
infinity address would like anonymity he/she should use an anon-mailer
on his/her end.

The encrypted-return address you send me should look like:

::
Encrypted: PGP

- -----BEGIN PGP MESSAGE-----
etc.

	Make sure you include that ::/Encrypted or the remailer which
gets it won't know that it's PGP encrypted.

	Remember, this is just setup for testing. Don't use it for
real applications.

- -Sameer

-----BEGIN PGP SIGNATURE-----
Version: 2.3a

iQCVAgUBLU9jrni7eNFdXppdAQH/FwP/b9pllDYnW6L4x0y1dVnC6km9TQ9lTw2x
U/ea87JnguYSHYRxOk6lZoBBx5ZH/A48OCHJztzWHaSP2Tq69Oro4FTrtRcpTjbf
ti8L97x9+Xvx1A6/Vkw1nuS5MRJ8SoPUV4bDKFdf80Ykhik5bk8b0WOUew1uF6dq
QJzyDsKDFQU=
=2EIr
-----END PGP SIGNATURE-----

On Wed, 2 Feb 1994, Sameer wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> 
> 	I've written a small anonymous mail service, and it's now
> available for testing. There's no security, and I'll be keeping logs,
> so don't think that it's secure, in any way. 

Excellently and well done, Sameer!  Ah, the pleasure of seeing a budding 
cypherpunk do us all some good.  Keep us posted.

Mucho Obligado, Amigo,
brianB

[Here is the conclusion to my QC paper, unfortunately I can't get the 
whole file into a PS format because of the faulty file translators in the 
Mac applications.]

Conclusion
	Quantum cryptography has proven to be an interesting and novel 
application of quantum physics.   It does posses some severe limitations 
that I have considered.
	Optimistic predictions of its affective area is still far below 
100 km.  This may of course change depending on technological 
development.  It has been suggested to me that one could have secure 
stations where interception and reception of the message would be 
allowed. [10]  This is possible, but weakens the absoluteness that is 
the appeal of quantum cryptography.  
	A basic assumption is made previous to the research mentioned: 
that Eve will not interfere on the public channel.  It could be very 
possible that Eve would set herself up between Alice and Bob on the 
quantum and private channels, and act as a relay station that I mentioned 
in the first point.  She would have to impersonate both Alice and Bob, 
who in reality might not even be on the same public and quantum channels, 
but merely think they are.  Public key methods could be used for 
authentication, but this destroys the motivation for the use of quantum 
cryptography.
		I feel the solution here is in the definition of 
public.  Meaning a random and public switching of public channels, 
phone numbers and such.  Even this may be subverted by a very powerful 
Eve who may also control the phone companys switching circuits.  Perhaps 
further thought can resolve this issue, but the problem of identification 
and authentication on the public channel is severe.
	Further, quantum cryptography is subject to a denial of service 
attack.  If Eve wishes, she may destroy the unique and expensive quantum 
channel, or merely observe everything that goes by, not caring to read 
the information, just making it unsuitable for use by Alice and Bob.  
Ekerts concept of keeping shared EPR pairs in permanent storage (perhaps 
using a superconductor to warehouse keys when the quantum channel is 
open) is not yet feasible, and it will be necessary to keep these keys 
somewhere , but the security of keys is not a problem unique to quantum 
cryptography.  
		I look forward to the resolution of these issues and the 
further  development of the technology that will allow quantum 
cryptography to become a practical security mechanism.


1.	C. Bennett. Science.. vol. 257, p. 752 (August, 1992).
2,	C. Bennett, G. Brassard, and A. Ekert.  Scientific American. p. 	
	50 (Oct., 1992)
3.	A. Ekert, Phys. Rev. Lett. vol. 67, p. 661 (1991)
4.	C. Bennet, and G. Brassard, Phys. Rev. Lett. vol. 68, p. 557 	(1992)
5.	A. Ekert, J. Rarity, P. Tapster, and G. Palma, Phys. Rev. Lett. 
	vol. 	69, p. 1293, (1993).
6.	A. Muller, J. Breguet, and N. Gisin. Europhs. Lett., vol. 23 (6), 
p. 	383 (1993).
7. 	S. Barnett, and S. Phoenix. Phys. Rev. A, vol 48 (1), p. R5, 
	(July, 	1993).
8.	C. Bennett. Phys. Rev. Lett. vol 68 (21), p. 3121 (1992)
9.	D. Denning. Cryptography and Data Security.
10.     Personal e-mail as a follow-up to a posting to sci.crypt.  I have 	
	unfortunately lost the persons name.