On Thu, Nov 21, 2002 at 03:19:06PM +0100, Oskar Sandberg wrote: > On Thu, Nov 21, 2002 at 03:01:03PM +0200, Costas Dokolas wrote: > <> > > There is indeed a need to improve speed (try saying that 10 times, fast), > > but let's not break freenet, OK? > > Let's not get overly dramatic here. Anything like this is a tradeoff: > duplicated data creates more work for the network, but so does making > ten requests when one would suffice. Clearly having all the data anyone > ever wants under a single key isn't a great idea, but then nor is > inserting each 256 byte values under a key each (imagine all the data in > freenet stored in a single 256 byte datastore!). > > Our current observations, which are all we have to go by, show that > traffic is a larger problem in freenet at the moment then a lack of > capacity to store all the data inserted - therefore it makes sense to > bundle small pieces of data just like we split very large ones. I'm not sure about that. The network is unbalanced, but most nodes don't have a traffic problem. Nevertheless it makes sense to do it this way if only from the point of view that making a site atomic is something useful to be able to do. > > Unlike Matthew's flooding, I would not characterise this as a > pissing-in-your-pants-to-stay-warm feature. > > -- > > Oskar Sandberg > oskar at freenetproject.org >
-- Matthew Toseland toad at amphibian.dyndns.org amphibian at users.sourceforge.net Freenet/Coldstore open source hacker. Employed full time by Freenet Project Inc. from 11/9/02 to 11/1/03 http://freenetproject.org/ -------------- 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/20021121/7b638698/attachment.pgp>
