On Sun, Jul 09, 2006 at 11:26:31AM -0700, Ian Clarke wrote: > > I don't see why it is necessary to provide artificial incentives for > people to run a darknet node, other than the real incentive of making > it much harder for people to know that you are running a Freenet > node. I'm also a bit concerned that turning Freenet into some kind > of generic JXTA-like platform for packet-forwarding is a potentially > distracting form of mission-creep.
I do. At the present time there is ZERO benefit for somebody who started a node on opennet to get darknet connections. There is no security benefit, because all connections are treated equally, and the user's node is visible for as long as it is connected to the opennet. We have to improve on that, or there will be no migration from opennet to darknet, and Freenet will remain almost exclusively opennet. With the result that it cannot be safely used in hostile regimes - hostile regimes which will include the West in IMHO literally the next few years. We can deploy opennet when a) it has been simulated, b) there are real incentives for using the darknet, even if you are also on the opennet, c) we have made some progress on the basic problems of storage and load limiting/balancing, and d) it is easy to add darknet connections (this has arguably already been fulfilled). > > Our key barriers to adoption right now are usability ones, and the > simple reality is that the lack of an opennet solution is excluding > by-far the most popular use-case for Freenet, which are people that > just want to get connected conveniently and who don't really care all > that much about who knows whether they are running Freenet. > > Providing support for the opennet use-case is not to the exclusion of > the darknet use-case, indeed given that it will bring more users, > more developer interest, and more donations, it will provide us with > more resources for working on both the darknet and the opennet. We have a lot of developer interest right now. And we're not too badly off for donations either. But if you want to make a bold (and ill thought through) step forward, I suggest deploying mrogers' token passing load balancing. It has not been simulated properly yet, but neither has opennet. > > Ian. -- 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/20060710/9f15abc6/attachment.pgp>
