--- Ken Corson <[EMAIL PROTECTED]> wrote: > Martin Stone Davis wrote: > > Some Guy wrote: > >> --- Ian Clarke <[EMAIL PROTECTED]> wrote: > >> > >>> In recent discussions relating to how we deal with the urgent problem > >>> of load balancing, the answer to both these questions is probably > >>> "no" - meaning that negative trust can "work", in the sense that it > >>> will probably help with load-balancing, even if it doesn't "work" in > >>> the sense that it is theoretically invulnerable to circumvention. > >> > >> Theoretic Solution to you Theoretic Problem: > >> Hash cash. If you want to connect to me you, and I don't particularly > >> want to connect to you I > >> Problem solved. --- err I think. > >> > >> There are other ways you can force someone to pay resources with hash > >> cash.
Opps, what I meant to say was there are other ways to make someone pay with resources. CPU might not be the best one. Ideally you make someone pay with the same type of resource he's getting (storage for storage, bandwidth for bandwidth). > > Hoorah! Well, that certainly sounds right to me. In which case, I'm > > now on board with the people who want to use negative trust. Actually I think that idea makes the trust positive. The untrusted agent has to devote resources to gain positive trust, which can then be taken away in a heartbeat if he starts screwing around. > This sounds like a good direction for improvement. I support anything > that discourages frequent (random?) connections between nodes. In > a healthy freenet, the rate of joins and leaves should be low, and the > lifetimes of existing links should be long. Longer-lifetime links will > allow us to build better info/NGR profiles on the peers. And it is > easier to model/analyze a network with /relatively/ static links. > > Can we share some URLs for "negative trust" education ? I found this: > > (give it a chance, it's somewhat relevant) > http://www.sbaer.uca.edu/Research/1994/SSBIA/94swi031.htm > > (and here is a more relevant one) > http://www.ai.mit.edu/projects/aire/pub-repository/WAID02.pdf Hmm, not sure I see how the second one relates to trust, I'll try to look at more later. Here's a neat paper Zooko (the MNet guy) pointed out: http://citeseer.nj.nec.com/douceur02sybil.html They argue you'd need a central authority to prevent a sybil attack. I think they're wrong. Arguing in your head against the paper gives you a nice feeling on what is possible and what isn't. They have some neat ideas. If you're feeling realy studious just search citseer, they tend to have good stuff: http://citeseer.nj.nec.com/cs?q=Negative+Trust&submit=Search+Documents&cs=1 __________________________________________________________________ Gesendet von Yahoo! Mail - http://mail.yahoo.de Logos und Klingelt�ne f�rs Handy bei http://sms.yahoo.de _______________________________________________ Devl mailing list [EMAIL PROTECTED] http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
