Chapter 8.12: Future Remailers

From THE CYPHERNOMICON

generally, closer to the goals outlined in Chaum's 1981 paper on "Untraceable E-Mail"
- Anonymity
- Digital Postage, pay as you go, ,market pricing
- Traffic Analysis foiled
- Bulletproof Sites:
Having offshore (out of the U.S.) sites is nice, but having sites resistant to pressures from universities and corporate site administrators is of even greater practical consequence. The commercial providers, like Netcom, Portal, and Panix, cannot be counted on to stand and fight should pressures mount (this is just my guess, not an aspersion against their backbones, whether organic or Internet).
Locating remailers in many non-U.S. countries is a Good Idea. As with money-laundering, lots of countries means lots of jurisdictions, and the near impossibility of control by one country.
- Digital Postage, or Pay-as-you-Go Services:
Some fee for the service. Just like phone service, modem time, real postage, etc. (But unlike highway driving, whose usage is largely subsidized.)
This will reduce spamming, will incentivize remailer services to better maintain their systems, and will
- Rates would be set by market process, in the usual way.
  "What the traffic will bear." Discounts, favored customers, rebates, coupons, etc. Those that don't wish to charge, don't have to (they'll have to deal with the problems).
- Generations
  - 1st Gen–Today's Remailer:
  - 2nd Gen–Near Future (c. 1995)
  - 3rd Gen
  - 4th Gen-

Dean Tribble has proposed…
"It sounds like the plan is to provide a convenient mail filtering tool which provides remailer capability as a SIDE EFFECT! What a great way to spread remailers!" [Hal Finney, 93-01-03]

Habryn, 94-04]

"Yes, but it's not running for real yet. Give me a few months until I get the computer + netlink for it. (It's running for testing though, so if you want to test it, mail me, but it's not running for real, so don't use it.)" [Sameer Parekh, 94-04-03]

to make there be a cost to flakiness (expulsion) and a benefit to robustness, quality, reliability, etc. (increased business)
- pings, tests, cooperative remailing
- spreading the traffic to reduce effectiveness of attacks
- which execute protocols
e.g., to share the traffic at the last hop, to reduce attacks on any single remailer

spy networks can be run safely, untraceably, undetectably
- anonymous contacts, pseudonyms

digital dead drops, all done electronically…no chance of being picked up, revealed as an "illegal" (a spy with no diplomatic cover to save him) and shot

so many degrees of freedom in communications that controlling all of them is essentially impossible

unless crypto is blocked–and relatively quickly and ruthlessly–the situation described here is unstoppable

what some call "espionage" others would just call free communication
(Some important lessons for keeping corporate or business secrets…basically, you can't.)

for example, if a remailer has a strict policy of
accumulating N messages, then reordering and remailing them, an attacker can send N - 1 messages in and know which of the N messages leaving is the message they want to follow; some uncertainly helps here

the mathematics of how this small amount of uncertainty, or scatter, could help is something that needs a detailed analysis
it may be that leaving some uncertainty, as with the keylength issue, can help

the "remailer@csua.berkeley.edu" remailer now adds actual paragraphs, such as this recent example:

"I fixed the SKS. It came with a scope and a Russian night scope. It's killer. My friend knows about a really good gunsmith who has a machineshop and knows how to convert stuff to automatic."
- How effective this ploy is is debatable

Anonymous AIDS testing.
under FDA review for 5
delayed release on the
badly and perhaps kill
test result…they want the existing system to prevail. mention this to show that anonymous systems are somtimes opposed for ideological reasons.)