Hal <74076.1041@compuserve.com> said:
My understanding was that everyone who tried to talk to him would get two aliases assigned automatically.
Actually, what I expected to happen was this: the acknowledgement of an anonymous id on penet for the anonymous id on pax would be generated, and this ack would be sent via the chain back to me (barrus@tree) since I established the chain. That is, I had linked anon.435@pax to an5022@penet to barrus@tree. Then, upon mailing to anon.435@pax from my other account (which already has an anonymous id established, so another would not be generated), the mail would proceed to an5022@penet, which would create an id and send it back to anon.435@pax. Now I expected this ack to then be turned right around, send to an5022@penet and then on to barrus@tree. So I was expecting the creation of another anonymous id, but the acknowledgement didn't go to barrus@tree. My original thinking was that once I established the id's in both direction, when someone responded to anon.435@pax, they would be allocated an id if they didn't have one. And since penet had by this time seen anon.435@pax, no new id would be made, and the mail would proceed on the me. Anyway, that was an experiment that seems to lead to explosive anonymous id growth :-) I agree with Matthew that not mailing back an ack would help cut down the flurry of mail, but it still results in all sorts of extra id's. I was hoping the whole thing would be like a pointer: mail to id1@pax forwards to id2@penet and then on to me with no extraneous account manufactured. But since we have our own cryptographically protected remailers, we cypherpunks can make our own remailing chains (Hal's constructed anonymous addresses). This way, you can decided on the path of your outgoing mail and your return mail: create the appropriate header, and attach your message on the end. To receive responses, just send a response header with the return path encrypted along with instructions to your recipient to cut the response header into a new file, add a message to the bottom, and mail to the appropriate remailer. (Note: I've used this method successfully twice, so it isn't too hard to do). Just remember that if your recipient doesn't have pgp, don't route your mail through extropia or their message will be blocked. If I get a chance I'll work on a program that will generate the appropriate header given routing input. /-----------------------------------\ | Karl L. Barrus | | barrus@tree.egr.uh.edu (NeXTMail) | | elee9sf@menudo.uh.edu | \-----------------------------------/