On Wednesday 19 September 2007 15:05, you wrote: > There doesn't seem to be much in the way of freenet .7 opennet design > documents available yet, so please pardon me if I'm asking foolish > questions.
We are all idiots most of the time. Inexperience doesn't make you more of an idiot in any statistically meaningful way. (Thanks Scott Adams for pointing out this basic fact about human nature!). > > For hybrid nodes with both opennet and darknet peers does it really > make sense that opennet links participate in location swapping? Yes. It's been simulated (admittedly not in great detail), and seems to work. And I don't really see what the alternative is, given that many peers will be simultaneously opennet and darknet on a single network, and given that we must take into account opennet peers when deciding whether to swap on darknet. > > It seems to me that a nodes numerous opennet links may exert excessive > influence over it's location, moving it off the position which would > be correct for the fixed darknet topology. Most likely the "fixed darknet topology" isn't even fully connected. We need a single network. > > Would it not make more sense if nodes connected to both simply slowly > changed who they link to on the opennet to achieve a desirable > distance distribution from their swapping derived location on the > darknet. They do change who they link to. That's the whole point of opennet. > > I'm seeing quite a few opennet links with very high latency. Because > opennet is already is 'less secure' and exists mostly as a way to > bootstrap freenet, would it be acceptable to mostly prefer peers with > lower latency? Opennet will eventually purge the slow nodes as it is based on dropping the least-recently-successful node. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20070920/4d256b23/attachment.pgp>
