P2P wanted?

[Ben Wilhelm]

The issue with hashcash is that it gives massive advantages to anyone 
with huge zombie botnets, and guess who's most likely to have those? 
Malicious users. (Sigh.)

It's almost an anti-improvement. :P

-Ben

Glymr Darkmoon wrote:
> the most likely mechanism i've heard proposed for agoric load balancing
> is the use of 'hashcash' which is asymmetrically expensive, it costs the
> sender more than the receiver, to be sure, not a huge amount, but it's a
> way to allow users originating traffic from a server node to be
> prioritised by the network at the cost of computation. exactly how to
> implement it in tor has yet to be defined, but it seemed to me like a
> damn good solution to the problem.
> 
> On Fri, 23 Dec 2005 22:37:33 -0500, "Roger Dingledine" <arma at mit.edu>
> said:
> 
>>On Fri, Dec 23, 2005 at 11:21:34PM -0400, Matt Thorne wrote:
>>
>>>I know that no one wants to do this, but out of curiosity could a tit
>>>for tat mechanism be built into the tor client? I realize that this
>>>wouldn't be desirable, just curiosity.
>>
>>It would be desirable, but it's a bit tricky. We need to a) design a tit
>>for tat mechanism that encourages good behavior, b) assess its impact
>>on anonymity and fix it until we're happy, and c) do some (possibly
>>significant) architecture changes on Tor so we can provide better service
>>to some servers or on some circuits.
>>
>>Please read section 5.1 of
>>http://tor.eff.org/doc/design-paper/challenges.pdf
>>
>>I'm collaborating with Dan Wallach at Rice University on a small NSF
>>project to look into this. I hope to have some answers in a month or two.
>>
>>--Roger
>>