[freenet-chat] Re: the plutocratic world of freenet
Juiceman ha scritto: Yes and (mostly) no. Having many nodes/servers/clusters and pushing and pulling the data you want will have a small effect on the network, but generally speaking the more popular the requested data is, the more places it resides. Thus if you have 10 nodes/servers/clusters pushing and pulling the content you want, but 100's of other people are requesting different content, the more requested content would win. Well, It would be so if the other 100 other requesters weren't human. If you have a single machine iteratively requesting content, 24h a day, it will generate much more traffic that what can produce 100 (or 1) regular surfing humans. A human will request the content 1 time, or 2. 1 humans will request max 10.000 times, max 20.000. A single powerful machine can simulate well such a traffic! Also please note that content would be divided across the network based on hash of said content and routing done by the network. So to replace someone else's content on a node you would also need to find content with a very close hash to that of the other person's content (not entirely easy) AND make sure to request it more than the other content AND control a majority of the routes for that data hash. Even then after expending great effort and resources you probably wont entirely succeed... That would be true if there where only one or few auto-requester robots (ARR). But if governments, politicians, corporations etc... will discover a new audience in freenet, they will not think two times about setting up ARRs. Each ARR that comes into freenet, will decrease the time non-ARR content can stay in the network, and increase the time it will need to spread itself. The same effect is due to popular files. Suppose there where many big and very popular files around the network. A publisher will see that his content is very slow, and has a limited permanence, compared to the popular files. So he say: my file will difficultly become popular if it is so slow and uncertain compared to popular ones. let's set up an ARR so users will gain speed and my content will have a granted permanence and integrity. Imagine a freenet where each, *EACH* publisher auto-requests its content at minimum 100.000 times a day with a costly server. It's obvious that between them they have approximatively the same impact on the net, so they have the same right to speek. But a poor publisher, without a costly ARR, will never be able to spread its content. Because the limit of popularity is too high. The time to get it would be so enormous from a user, that many will give up, not raising popularity. The minimum requirement to have a decent reachability and permanence it would be 100.000rqs/day or little less... That's nothing new. It's the same as in real democracy. It has always been very very difficult to combine free speach, right to listen, and MONEY. ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/chat Or mailto:[EMAIL PROTECTED]
Re: [freenet-chat] the plutocratic world of freenet
On Fri, Aug 26, 2005 at 06:38:15PM -0400, Juiceman wrote: Yes and (mostly) no. Having many nodes/servers/clusters and pushing and pulling the data you want will have a small effect on the network, but generally speaking the more popular the requested data is, the more places it resides. Thus if you have 10 nodes/servers/clusters pushing and pulling the content you want, but 100's of other people are requesting different content, the more requested content would win. Also please note that content would be divided across the network based on hash of said content and routing done by the network. So to replace someone else's content on a node you would also need to find content with a very close hash to that of the other person's content (not entirely easy) AND make sure to request it more than the other content AND control a majority of the routes for that data hash. Even then after expending great effort and resources you probably wont entirely succeed... No, you'd just flood the whole network with random data. But this only works if you have high bandwidth to the whole network - which is not the case on a darknet. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/chat Or mailto:[EMAIL PROTECTED]
Re: [freenet-chat] Re: the plutocratic world of freenet
On Sat, Aug 27, 2005 at 12:32:55PM +0200, daniele wrote: That would be true if there where only one or few auto-requester robots (ARR). But if governments, politicians, corporations etc... will discover a new audience in freenet, they will not think two times about setting up ARRs. Each ARR that comes into freenet, will decrease the time non-ARR content can stay in the network, and increase the time it will need to spread itself. Flooding is possible. But on the darknet it is pretty hard to flood out the whole network. Obviously having a large network to start with would also help. The same effect is due to popular files. Suppose there where many big and very popular files around the network. A publisher will see that his content is very slow, and has a limited permanence, compared to the popular files. So he say: my file will difficultly become popular if it is so slow and uncertain compared to popular ones. let's set up an ARR so users will gain speed and my content will have a granted permanence and integrity. If his content is popular it will rapidly become available quickly. As far as deliberate attacks go, see above. Imagine a freenet where each, *EACH* publisher auto-requests its content at minimum 100.000 times a day with a costly server. It's obvious that between them they have approximatively the same impact on the net, so they have the same right to speek. But a poor publisher, without a costly ARR, will never be able to spread its content. Because the limit of popularity is too high. The time to get it would be so enormous from a user, that many will give up, not raising popularity. The minimum requirement to have a decent reachability and permanence it would be 100.000rqs/day or little less... That's nothing new. It's the same as in real democracy. It has always been very very difficult to combine free speach, right to listen, and MONEY. It's the love of money that is the root of all evil, not money itself. Flooding is possible, but it can be made expensive by having a large network, and very difficult by having a darknet. An analogy: the gap between super-rich and average isn't so huge that it is practical to buy the entire electorate, in western democracies. Because there are very many of them, and they're not *that* poor. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/chat Or mailto:[EMAIL PROTECTED]
Re: [freenet-chat] Re: Freenet mention on BBC Radio 1 tommorow?
No, my interview was for Click Online, and won't go out for around 2 weeks. Ian. On 26 Aug 2005, at 20:18, Bob wrote: Bob [EMAIL PROTECTED] writes: Hmm, sorry maybe it wasn't today after all (although it sounded like it was). There was a brief 'Internet issues' piece at 6:00pm again but it was about VoIP and if it was going to take off, no mention of anonymous networks. Bob Much as I hate to endlessly self-reply I don't want to look like I'm making stuff up :) Maybe I half-heard a reference to this? http://dodo.freenetproject.org/pipermail/iansblog/2005-August/ 53.html I am flying to London for the day tomorrow to do an interview about online anonymity for the BBC's Click Online show. I scanned through the current show (http://news.bbc.co.uk/1/hi/programmes/click_online/default.stm) this lunchtime but didn't see Ian, has it been broadcast yet? Bob ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/ listinfo/chat Or mailto:[EMAIL PROTECTED] ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/chat Or mailto:[EMAIL PROTECTED]
[freenet-chat] Re: the plutocratic world of freenet
The most convicting truth about my error in considering freenet a plutocratic network was given to me by private mail, as follow: [...] The key here is _distribution_ over the wide geographic area, which is simple for regular users, but requires a very costly and difficult to manage concerted effort on behalf of any organized adversary. And: [...] Freenet has specific properties to combat this threat. For example, after the first request, content will get cached on the nodes near the attacker, and his subsequent auto-requests will have no effect on the network whatsoever. Of course, you can always harvest the network and overload nodes one by one - but this requires _yet_ more effort and time (and, since, is more expensive). To combat _this_ threat, Freenet is proposing a darknet approach, which will make Freenet difficult to harvest, as well as difficult to establish connections to new nodes (you'll have to subvert node operator in real life to make a connection to his node). So, an auto-requester will only spread it's content in few neighbour nodes. -- - it.scienza.chimica, 25/8/05: che differenza c'è tra molarità, molalità e moralità? con le prime due puoi descrivere un politico. [dp] http://blog.daniele.homelinux.org ___ chat mailing list chat@freenetproject.org Archived: http://news.gmane.org/gmane.network.freenet.general Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/chat Or mailto:[EMAIL PROTECTED]
