> > As for closeness, is the purpose of this to choose which node to send > a request which could not be served locally to? I looked in the > source code, and it appears that "closeness" is literally how close > the keys are to each other, not some algorithm to determine which > nodes are most likely to have the data available on them. Why not > just keep data on nodes and their likelyhood of having data available > on or through them, and try use nodes in descending order of > reliability? That is what I was planning to do with nfreenetd. Of > course, such a node choosing algorithm would result in nodes with > large datastores and many connections to other nodes getting much > heavier use that other nodes.
Holy shit Travis. I suggest you really stop coding nfreenetd now and spend a *lot* more time studying existing code, the protocol, and any other documents you can get your hands on. You've totally missed the boat here. Key-closeness is what makes Freenet routing work. Its not just some random select whatever-the-hell node I want next to find data, you *have* to choose the node that has in the past had keys close to the one you're looking for. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20000804/826e4804/attachment.pgp>
