Please would you quote properly? I've corrected your quoting and answered your points below...
On Sat, Aug 12, 2006 at 04:39:18AM -0700, Scruple Scruple wrote: > > Matthew Toseland <toad at amphibian.dyndns.org> wrote: > > On Tue, Aug 08, 2006 at 06:29:28PM -0000, anonymous freenet user wrote: > > > > > > When will 'Open-net' be deployed? > > > > > > I would like very much to migrate to 0.7, but I cannot condone doing so > > > until 'open-net' is active and making a wider anonymity set operational. > > > > Opennet might increase the total number of nodes on the network by > > making it easier to join it. In that sense it would increase your > > anonymity set. However, it does not improve your anonymity in any other > > way. 100 opennet connections is probably less anonymous in practice than > > 10 true darknet connections. Because most of those 100 connections may > > well be to your enemy, whereas treachery is less likely. > What treachery? Even if you are on an open-net and connect to a 'bad' node > the only time that would be a problem, if Freenet is working properly, is if > running Freenet would get you into trouble. You're plausible deniability > would be intact - it would be exactly as Freenet 0.5 is now. On a darknet, a peer attacking you would be called "treachery". On an opennet, since you don't know them, it's called "stupidity on your part". Either way, there are attacks that your peers can do on you. Some of them are rather powerful attacks. And yes the existence of attacks means that freenet is imperfect, and arguably broken. All I can say is it's not version 1.0 yet. > > Of course when you are on a dark-net and connect to a 'bad' node, as is > currently possible given the ugly hack that is #freenet-refs, it is far more > worse as they both know you are running Freenet and given that there are > fewer darknet connections out that leaves you and your other peers more open > to analysis. Does it? I don't think that every node having 50-200 opennet connections is a viable way forward; there are some fairly convincing arguments that we should have relatively few connections. One of those reasons is so that darknet isn't completely swamped, but there are other reasons to believe that very large numbers of opennet connections aren't helpful. > However, most people don't have many true darknet connections, so in > that sense it is better than #freenet-refs or ifreed.net. That is a short term problem. As the network matures more true darknet connections will be created. > > > > > As it is, 0.7 is not sufficiently anonymous, > > > 0.7 opennet will not be any more anonymous than 0.7 darknet. People > > don't seem to understand this, I don't know why, but I will say so > > again: On opennet you have no control (unless you are evil) over who you > > connect to. That means you are probably connected to an ubernode run by > > the bad guys pretending to be about 5000 nodes. On darknet you have to > > trust your friends; on opennet you have to trust total strangers who may > > well be colluding because they may be the same person. In version 0.8 we > > will implement premix routing and some other measures which should > > improve your anonymity even against untrusted direct connections, but it > > is quite possible that these will only work on darknet. > Open-net is more anonymous, as was said above, because it increases the > anonymity set. If Freenet is ever to amount to anything then it will have to > be used - that is the goal isn't it, for Freenet to actually be used? If it > is to be used then it will need to be easy (easier) to use. Freenet needs to have more nodes yes, this I accept, and it's the main reason why we will have to implement opennet. > > > > If Freenet isn't easy to use and continues to devolve into a GNUnet style > > > 'academic network' with 5 (50 or even 500?) people on it, that is likely > > > not a sufficient pool to continue to fund Freenet though donations (or > > > perhaps it is hoped those left are millionaires?). 500 should be able to fund a full time dev. And we have nearly that now. But we need significant growth. One way to achieve that is opennet. However we will have growth anyway, as long as the network works and does something useful. Opennet is for bootstrapping. > > > If I start a node, get some > > > refs and start inserting content, it doesn't take a whole lot of rocket > > > science to figure out that new content is probably coming from the new > > > node. > > > > Unfortunately this is exactly the same on opennet. Except that it's > > worse, because it's not just the people who you know, (or even the > > people you found from #freenet-refs) who can attack you. > Attack me how? Freenet 0.5 has been up for years without anyone, to my > knowledge and do correct me if I am wrong because I'd like to know, being > arrested for using it. That speaks pretty highly of Freenet. It's under the RADAR. 0.5 had maybe a few tens of thousands of users, and was user unfriendly and slow. 0.7 has fewer users at present but will grow. > > The only think darknet offers is protection if it is illegal to run Freenet. > That is quite useful, but should not be a replacement for an open-net. If > Freenet does become illegal to run for a user in their country then they can > migrate to darknet, or already be participating in darknet as well and simply > withdraw from open-net. *Hopefully* they will be able to withdraw to the darknet completely, and hopefully the authorities weren't already harvesting... What happens when Freenet becomes illegal in the West? Opennet is good for getting people quickly onto the network. But once they're on it they must migrate to darknet for it to have a future (and for them to have true privacy). > > > Better by far to have open-net active, this makes it a LOT easier > > > for lots more people to join the 0.7 network and create a larger crowd to > > > get lost in. > > > > It does make it easier for people to join the network. That's its sole > > redeeming feature in fact, and why we will implement opennet - but not > > yet; adding more chaos to the current utter chaos would not help > > matters. > Not only does open-net make it easier to connect to, it will be the clue that > holds the different dark-nets together. Or do you expect that darknet > dissidents in China will some how be able to magically meet and connect to > darknet dissidents in other countries? That would be easier to do given an > open-net where different dissident groups could talk to, and perhaps exchange > connections, though the open-net. They will have to, because opennet will be blocked in China. The authorities will harvest the global opennet and block all external nodes, and identify any internal nodes and block their internet access. Thus the chinese darknet will have to have *darknet connections* to other networks. But there are a lot of people in China who know people in the west and vice versa; that's one factor behind its recent economic success. > > The hope is that people on opennet will gradually get darknet > > connections as well - for security reasons, and probably for other > > reasons e.g. web of trust file index sharing etc. > I agree totally. > > Although I don't think a Freenet without an open-net will last in the long > run, or even the medium run - without usability there won't be enough people > to create enough content (pages, chat) criticle mass. Open-net is more of a short run thing for me. In the medium run Freenet will probably be illegal just about everywhere, so darknet is the only viable option. -- 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/20060812/598ad728/attachment.pgp>
