Re: An idea about Java and remailer clients and servers...
Phil Fraering writes:
Why not "charge" for the ability to send an anonymous message with the duty to have for a short time (maybe an hour or two) running on your machine a node in a remailer network?
User X on Machine A sends a form via HTTP (or a variant- SHTTP, HTTPS, etc.) to Machine B. User Y on Machine C receives an anonymous mail from Machine B. Suspecting User X, User Y sends a mail to be anonymized and sent back to himself to User X. User X's temporary remailer does as it's told. User Y now has a strong reason to suspect User X has sent the said mail. Cpunks write code and all, but I don't think this one's going to work. :-(
Phil Fraering writes:
Why not "charge" for the ability to send an anonymous message with the duty to have for a short time (maybe an hour or two) running on your machine a node in a remailer network?
User X on Machine A sends a form via HTTP (or a variant- SHTTP, HTTPS, etc.) to Machine B. User Y on Machine C receives an anonymous mail from Machine B. Suspecting User X, User Y sends a mail to be anonymized and sent back to himself to User X. User X's temporary remailer does as it's told. User Y
I forgot to add. There is no reason User X has to run his remailer immediately. His software could simply commit to running a remailer for 1 hour at some specified future date < some threshold. Any messages sent to him for remailing would be queued until that time. Therefore, all your technique would tell you is that the user remailed a message sometime between date X and date Y. if Y-X > few days to week or two, the intelligence gathered on User X is miniscule. Traffic analysis would detect User X using the remailer network anyway. -Ray
Anonymous writes, concerning the "you want a remailer, you run one" idea: User X on Machine A sends a form via HTTP (or a variant- SHTTP, HTTPS, etc.) to Machine B. User Y on Machine C receives an anonymous mail from Machine B. Suspecting User X, User Y sends a mail to be anonymized and sent back to himself to User X. User X's temporary remailer does as it's told. User Y now has a strong reason to suspect User X has sent the said mail. Cpunks write code and all, but I don't think this one's going to work. :-( I was thinking in terms of User X running one node in a mixmaster network. AFAIK, mixmaster doesn't work in one-bounce mode; otherwise, why would it go through all the trouble of breaking up the messages, etc.? Phil
Phil Fraering writes:
I was thinking in terms of User X running one node in a mixmaster network. AFAIK, mixmaster doesn't work in one-bounce mode;
The sender can ask the client to set the number of hops as low as 1, if she/it/he so desires. I think Doug Barnes has suggested the best protocol for handling "co-op remailing" ;) -Futplex <futplex@pseudonym.com>
Phil Fraering writes:
Why not "charge" for the ability to send an anonymous message with the duty to have for a short time (maybe an hour or two) running on your machine a node in a remailer network?
User X on Machine A sends a form via HTTP (or a variant- SHTTP, HTTPS, etc.) to Machine B. User Y on Machine C receives an anonymous mail from Machine B. Suspecting User X, User Y sends a mail to be anonymized and sent back to himself to User X. User X's temporary remailer does as it's told. User Y now has a strong reason to suspect User X has sent the said mail.
If the "duty" cycle is 1 hour and there are 10000 users utilizing the network, that tells you nothing. All it does it confirm that User X sent a remailer message within the last hour. One could just as easily finger User X and use the same reasoning. And if one has to suspect User X in the first place, User X has already blown his cover partially (either by writing style or other leaks) -Ray
participants (4)
-
anon-remailer＠utopia.hacktic.nl -
L. McCarthy -
Phil Fraering -
Ray Cromwell