You seemed to indicate that you'd start immediately? So how's the simulation going? Are you anywhere close to a model which can include load yet? We have several of our own simulations for 0.7, but they don't include load...
As I see it, the first priority is to make a reasonable simulation of the proposed algorithms (and the current one), taking load into account. If the proposed algorithm handles it better, we need to tune it, and then deploy it. If it doesn't it's back to the drawing board. If the proposed algorithm works perfectly, then there's another thing to investigate: Can we bias routing to give less keyspace to really slow nodes? IMHO it is important that - if possible - we give modem nodes something useful to do, and therefore plausible deniability. What's the current status? If you can't start work seriously for another month that's fine, but you were asking questions, so...? -- Matthew J Toseland - toad at amphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/tech/attachments/20060602/9a677712/attachment.pgp>
