> > I'd rather have a stable, fast connection around half the globe, I
> > think, then a shaky, slow one to my neighbor. But maybe that's selfish
> > of me.
> 
> If one would propose to establish a proximity metric for choosing
> ultrapeers, I would choose the ping latency as a metric, not the hops. 

Reducing the number of hops would reduce the processing load of network
devices (think at LAN administrators) and maybe the average bandwidth
consumption.

I'm agree that the most efficient metric used to provide the best
quality to the servant is the ping latency.

I think those 2 points could benefit to the Gnutella network. The first
point is good because user will have one more favor on choosing the
Gnutella network if Gnutella agents save the networks. The second point
is good for the global latency of a request. Indeed if each node choose
the other nodes with a smaller latency, then the global latency could be
improved.


J�r�me



-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge
_______________________________________________
Gtk-gnutella-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/gtk-gnutella-devel

Reply via email to