Re: [freehaven-dev] info hiding conference

[Roger Dingledine <arma@mit.edu>]

On Tue, Nov 28, 2000 at 08:23:05PM -0500, dmolnar@belegost.mit.edu wrote:
> Another possibility might be to expand the material on paying mix-nets. 
> I don't think I've seen a fully worked out explanation of how
> mix-net nodes can actually benefit from their mixing duties. There is some
> cypherpunk speculation about how it will all be paid for with digital cash
> eventually, but as we've noticed there are some nasty details. 
> 
> This may be important, since it seems that running a mix node exposes you
> to harassment and possibly legal action. So people start a node out
> of a volunteer spirit and then let it die later. This might not happen 
> as often if remailer operators were paid. 

http://nymip.sourceforge.net/cvs/general/zks-desiderata.html#toc84

|Decentralized Payment For Service
|
|It should be possible for relays to collect payment from clients without
|using any centralized collection service, and without keeping per-client
|records of services rendered. Furthermore, providers of services other
|than relaying, such as CAs, should be able to collect reasonable payment
|for their services.
|
|Payment for relaying services should be collected in increments small
|enough that nonperformance on any increment by either party will not
|grossly burden the other.
|
|Motivation
|
|A centralized payment system is both a single point of failure and a
|serious traffic analysis liability, and billing records can be seriously
|damaging to forward secrecy.
|
|This section basically requires digital-cash-based micropayments at the
|time of service. There doesn't seem to be another way to satisfy it. What
|sort of micropayment system to use is uncertain.

I am dubious that we can solve this problem in a week or two. :)

Did you have in mind just a survey of the issues involved and a couple of
'this won't work because of this' examples?

It seems to me that if you want remailer operators to be 'paid' as
incentive to continue being an operator, then you're going to need
to use digital money with 'real value'. Does that necessarily imply an
infrastructure for turning that 'real value' into dollars, or can we come
up with some alternative (local) currency that is sufficiently useful?

--Roger