At 4:14 AM -0500 9/29/96, David E. Smith wrote:
I dunno about that one. Even if the message itself is encrypted, every remailer has to have the address of the final recipient for your plan to work. And if you have cleartext for that message ...
Actually, as I read the original proposal, it was to only _insert_ additional links. Thus, imagine the following remailer chain, with Alice sending to Zeke through a chain, represented in parentheses (like LISP): Bob(Charles(Dora(Ed(Frank(........Zeke))))....) (All messages are encrypted, etc.) Now, imagine that Charles elects to add two additial chains, XXXX and YYYY (to make them stand out in my representation). From his point on, the chain will look like: (XXXX(YYYY(Dora(Ed(Frank(........Zeke))))....) Likewise, any of the later remailers can add more links, etc. Some dangers are that "lost Dutchman" messages will remain in the system forever. This gets fixed by probablistic criteria, to produce convergence. Or with digital postage, which causes convergence for ontological reasons. Adding new links, or adding "middleman" links, is always possible. We've had several discussions of this over the years. --Tim May We got computers, we're tapping phone lines, I know that that ain't allowed. ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay@got.net 408-728-0152 | anonymous networks, digital pseudonyms, zero W.A.S.T.E.: Corralitos, CA | knowledge, reputations, information markets, Higher Power: 2^1,257,787-1 | black markets, collapse of governments. "National borders aren't even speed bumps on the information superhighway."