On 01/20/2010 04:25 PM, alex wrote: > I see all these difficulties, and then some. For example, cooperative index > building. Can this be completely replaced by aggregated indexes relying on a > WoT? I say this because I see a basic difference at work here. In freenet, I > understand that a btree is inserted under some key, and thus can be authored > by a single "entity" (or collection of trusted peers). However, in > traditional p2p, distributed hash tables (e.g. kademlia) are built by all > peers. Or I'm wrong here about how btrees will work in freenet?
The b-tree works on a different layer from the DHT - it's implemented on top of it. The node layer only knows about the DHT, the b-tree doesn't affect it. The individual b-tree "nodes" are just normal entries in the DHT, and these entries point to each other, these pointers forming the "tree" structure. X _______________________________________________ Devl mailing list Devl@freenetproject.org http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
