On May 21, 2003 03:09 am, Tracy R Reed wrote: > On Tue, May 20, 2003 at 06:21:51PM -0700, Ian Clarke spake thusly: > > Amuzingly many of the BitTorrent indexes have gone down due to > > slashdotting. This may be a good time to advocate Freenet as a > > more-or-less Slashdot-proof BitTorrent alternative with the added bonus > > of anonymity. > > I have friends who are really into bittorrent. bittorrent does not have > the problem of data falling out of the network so easily (ie. much more > reliable), having to deal with java, and it is generally faster. I > downloaded some "video and audio content" the other night at 140k/s for > one stream and around 70 for another. I have seen freenet get up to 90 but > not lately because I haven't been able to assemble a full splitfile of any > size in a month or two.
I will second this. Routing in freenet is, imho, working quite poorly at present. Think we need to revert to the old method which is known to work (no its not perfect) or implement the new ideas. What we have now is not effective. It would also be interesting to understand why the current method performs so poorly. This would allow us to fix simulations making them actually predict how freenet will respond... One reason might be that nodes do not really know much about other nodes. In my routing table, only 16 (of 50) nodes have info on more than 10 keys. Maybe some method of having nodes exchange key info would help? ie implement a what keys do you know node xxx has type message. > Freenet clearly has bittorrent beat on anonymity > and being decentralized and overall I think freenet has the potential to > be better technology for even just plain file transfers but it has quite a > ways to go. Ed Tomlinson _______________________________________________ devl mailing list devl at freenetproject.org http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl
