hal at finney.org wrote: > But inserts and requests use the same routing algorithm. Therefore inserts > of data under the same key must cross, as well. Also, once requests cross > they are likely to stick to each other (unless the closeness database has > changed significantly in the interim). Therefore almost all inserts of data > under the same key will lead to the same final node, and that node will > be expected to hold entries for every data item in Freenet that uses that > key. > > QED.
I agree that inserts and requests are supposed to stick to each other after a few hops. But my understanding is that THERE IS NO FINAL NODE. There is no "sink" where all of the routes for the same key end up and stop, or wind tight circles around. The reason is that data is cached on every node along a route, so a request stops as soon as it touches the insert route. It doesn't follow it any further. That intersection point will become a new magnet for subsequent requests. Let's just consider a set of inserts and requests for the same routing key (assuming multiple inserts can share the same routing key). Imagine inserts and requests as ants and data as sugar. Initially one insert ant wanders into the network dropping sugar at various random places. Now another request ant finds the sugar. Other request ants notice the chemical trail of that ant and start following it. More insert ants arrive from different directions with sugar which they want to drop (real ants don't do this =). They follow the chemical trail too and reinforce it. But they don't just go to the end and dump all the sugar in a big pile. They drop sugar all the way along. Further request ants are now attracted to the sugar that's being deposited along all these subsidiary trails leading into the first trail. THEY DO NOT FOLLOW THE TRAIL TO THE END. Instead, they start creating new trails that lead to these subsidiary piles. These new trails grow in importance, and ants even stop visiting the original sugar piles, and the first chemical trails evaporate and disappear. Ian? Is that right? theo _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
