--- Ian Clarke <[EMAIL PROTECTED]> wrote: 
> Some Guy wrote:
> 
> > Long term if we implement something like premix routing, we could
> > change the underlieing network to some more efficient DHT as long as
> > it was secure against DSA and other censorship attacks dirrected at
> > the data.
> 
> I have never seen any convincing argument that a DHT will perform better 
> than pre-NGR routing under realistic network conditions, so it is far 
> from accepted that migrating to a DHT is a desirable goal, in the short 
> or the long term.

I'm not pushing Chord, CAN, Pastry, or some pet DHT of mine.  What I'm saying is that 
if we don't
have to worry about send/reciever annoymity, we're free to try other underlying nets 
be they NG,
old-style, or what ever.

That said you're right I have no conclusive evidence that any DHT peforms better than 
NG.  However
the only evidence I have to the contrary is that freenet is way more popular than the 
DHT projects
out there.  That doesn't convince the hell of me.  Freenet maybe more populare because 
it claims
to provide annoymity.


> Furthermore, I suspect that NGR-based routing will be far superior to 
> DHTs under realistic network conditions, as DHTs tend to be rather 
> brittle.  Someone should really do a proper comparison to resolve this 
> question once and for all.

Well, Ken and I are discussing doing some things like this.  I'm thinking of a simple 
experiment
of NG kind of like in the 2000 paper you wrote.  I just want to see if delivers a nice 
O(f(N))
route time and how long it takes an unspecialized net to converge.  We probably should 
have done
this a long time ago.

As far as real life P2P systems being to brittle for DHTs.  You may be right even 
though they
claim things like O(log(n)) connection work, thier conectivity stuff is way 
complicated.  My
reaction is about like yours "way complicated"="way broken" most of the time.

What ever DHT (count NG as a DHT) we used.  It'd be free to focus on censorship 
resistance.  That
would mean fighting:
* the targetted DoS attacks
* cancer nodes
* the other stuff people will think of


__________________________________________________________________

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

Reply via email to