[mnet-devel] Grid Of Trust -- I'm back

[Some Guy]

 --- Bryce Wilcox <zooko at zooko.com> wrote: > 
> > There was a more complicated idea from which this design grew.  Let's see what you guys think
> of
> > it.  I was thinking periodically to reorder on dimension at a time say once a week.  So for a
> week
> > you'd have you X and an X' based on a different salt.  To meet you new neihbors you'd have a
> bunch
> > of first-order neighbors who would have had a long term first order relationship with them. 
> The
> > idea was to keep advesarial nodes stired around.
> > 
> > The cost is of coarse that all important data would have to be transfered one hop per week.
> > 
> > New nodes might only get connected to neihbors in their own cell and slowly be able to route
> in
> > more dimensions as those are mixed.  This would force an adversary not just to boot a node to
> get
> > IPs, but actually run a productive node for a few weeks.
> > 
> > The system would need some kind of way of switching up g and D as it grew, switching up the
> salts
> > seems like it would fit in with that. 
> > 
> > Like? Don't like?
> 
> It's complicated!  I wish there were a simple design that comes with a simple 
> argument that it is attack-resistant.  Raph's advogato design has that 
> quality, but it is centralized.

I'll have to check out advogato sometime.
 
> I don't really believe in hashcash for attack resistance, since we want people 
> with $50 handhelds to easily and conveniently connect to the network while being 
> safe from people with $500,000,000,000 special-purpose computers.

I like to think of the perfect P2P net as a good civil society.  
Everyone has rights.  
The overwhelming majority supports those rights.
This majority controls the majority of power like money, press, weapons, knowlege, ect.

Of coarse if there's only $500,000,000,000 adversary and your $50 handhelds, you're screwed.  The
question is what ratio of resources of an adversary to the good guys can a system withstand.

> On the other hand, I consider friendnet for attack resistance to be more 
> plausible, for example the people with these $50 handhelds might point their 
> infrared ports at their neighbor's handhelds and click on the "that guy is okay" 
> button, thus building up a network of local trust-links that could not be 
> penetrated even with a $500,000,000,000 special-purpose computer.

How do know you can trust your "friends"?  Sorry don't want to sound paranoid, but if someone out
to get you really has $500,000,000,000, he might have ways of getting them to cooperate.

I like the idea of using real world trust.  I've considered using it for routing.  If you want to
use such links to build a global DHT, you still need a way to randomly assign responsiblity.  That
was the point of the hash cash.  The alternative to a global DHT would be something like waste.

Maybe a group of friends can do the hash cash, and share being a node.


__________________________________________________________________

Gesendet von Yahoo! Mail - http://mail.yahoo.de
Logos und Klingeltöne fürs Handy bei http://sms.yahoo.de


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
mnet-devel mailing list
mnet-devel at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mnet-devel