On Mon, Jun 25, 2001 at 04:58:40PM -0400, Tavin Cole wrote: > On Mon, Jun 25, 2001 at 10:49:18PM +0200, Oskar Sandberg wrote: > > On Mon, Jun 25, 2001 at 04:24:57PM -0400, Tavin Cole wrote: > > > On Mon, Jun 25, 2001 at 10:08:12PM +0200, Oskar Sandberg wrote: > > > > On Mon, Jun 25, 2001 at 11:58:22AM -0700, Scott Miller wrote: > > > > > > > > > I agree that it may be too strong. Perhaps you should drop it if > > > > > > > it > > > > > > > exceeds some percentage of the total number of refs. I do like > > > > > > > the > > > > > > > 'deleting from the top end' though. > > > > > > > > > > > > Et tu Brute! > > > > > Come on, you have to admit its an interesting solution to both traffic > > > > > balancing and the ubernode problem. It might also shake up the > > > > > system a > > > > > little by causing entrenched nodes to re-evaluate the routes by having > > > > > to find alternate routes for keys going to the most popular route in > > > > > their datastore. > > > > > > > > The percentage thing is too arbitrary, it rubs me the wrong way. > > > > > > > > My reaction, though, was to both of you presenting these as solutions to > > > > the honest cancer issue, given the first rule. > > > > > > Oh, come on. This is not about "distrusting" a node... > > > > If you use it as a way to keep a node from fucking with you, then yes it > > is. > > Naah, this is simply about preventing any node in your routing table > from having more references than the other nodes. Which is why I > still don't think it's overly strong. Particularly because I value > the "shaking up" effect Scott noted.
Yes, but it doesn't stop one party from having more references than other parties, so it cannot be considered to address honest cancers at all. -- 'DeCSS would be fine. Where is it?' 'Here,' Montag touched his head. 'Ah,' Granger smiled and nodded. Oskar Sandberg oskar at freenetproject.org _______________________________________________ Devl mailing list Devl at freenetproject.org http://lists.freenetproject.org/mailman/listinfo/devl
