From: "E. ALLEN SMITH" <EALLENSMITH@ocelot.Rutgers.EDU>
Date: Wed, 22 May 1996 20:00:40 +0800
To: dsmith@prairienet.org
Subject: Re: Long-Lived Remailers
Message-ID: <01I4ZO8U2J4O8Y5IL9@mbcl.rutgers.edu>
MIME-Version: 1.0
Content-Type: text/plain


From:	IN%"dsmith@prairienet.org" 22-MAY-1996 01:33:37.30

>Actually, there's an Idea.  Set up a single address; use added
>headers in the style of:

>::
>Remailers-To-Chain: 7
>Remailers-To-Avoid: remailer@nsa.gov
>Final-Destination: tcmay@got.net

>Each remailer could construct a message that decrements the
>remailers counter, preserving the other headers.  The
>usual caveat on encrypting at each step would apply; but since
>remailers' pubkeys are available, that's a trivial concern.

	Well, if you use this for the entirety of the chain, you'll be giving
away the cleartext at each step. Not too good of an idea. You won't want
the mails to an output location to themselves go through remailers; you'd want
multiple remailers going to the same output location (or group of output
locations, which is probably preferable) via a more direct means such as POP.
Otherwise, you've simply got the same old - but good nonetheless - of adding
on new full remailers to the end of a chain, which doesn't avoid problems for
the full remailer at the end. (Re: TCMay's post in response to yours.)
	-Allen