On Friday 25 Jul 2003 16:50, Tom Kaitchuck wrote:

> Ether way could work, but the optimal solution is somewhere in between. The
> question is how do you DEFINE the optimal network? Each node could balance
> these two extreams to meet it's current needs, but what do you optimize
> for? If you optimize for local near term routing speed, you'll always get
> the first configuration. We currently use a version of the first system
> with lots of built in fudge factors. But what is the BEST solution? When
> does redundancy become more important than capacity?

Hmm... Would it be possible to take some metrics and make nodes optimize their 
optimization depending on the recent network behaviour? e.g. if the network 
has been slow but reliable, optimize for speed, but if it has been fast but 
unreliable optimize for redundancy? What happens when it becomes slow and 
unreliable? :-O

And how do you tell those conditions apart?

Gordan
_______________________________________________
devl mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl

Reply via email to