Good point about the error, though I wonder whether in one respect the actual darknet design (all nodes belong to one network, which also includes opennet nodes) might be more vulnerable than the design they imagine (many small networks): a powerful attacker could harvest opennet addresses, then monitor those addresses to find hybrid nodes with darknet neighbours, then monitor their darknet neighbours, and so on. Whereas with isolated darknets there'd be no public starting point.
A related thought experiment: let's say I'm a darknet user in China, and the government harvests opennet addresses and blocks them, so my darknet's now cut off from the rest of the world. What happens to my darknet? Does location swapping, together with the occasional resetting of locations to random values, eventually cause it to form an isolated mini-Freenet? Cheers, Michael On 28/07/10 14:35, Ian Clarke wrote: > There is an error in the paper: > > Section 3.1: > > "Note that be- cause darknet nodes do not communicate with opennet > nodes, there may be many disconnected darknets instead of one large > network." > > Darknet nodes can communicate with an opennet node provided there is an > out-of-band agreement between them, the hope being that this will > prevent the problem they describe. > > Ian. > > On Wed, Jul 28, 2010 at 6:44 AM, Michael Rogers <m.rogers at cs.ucl.ac.uk > <mailto:m.rogers at cs.ucl.ac.uk>> wrote: > > Sorry if this has already been discussed - I was off the list at the > time the paper was published. > > Researchers at the University of Minnesota carried out an address > harvesting attack on Freenet opennet. They estimate the size of the > opennet at around 2,500 nodes at any time, from a long-term population > of around 11,100. It took a single node 2.5 hours to collect the > addresses of nearly all active opennet nodes. > > Obviously the concept behind this attack isn't news to anyone here, and > I realise it's not something opennet tries to protect against, but I was > interested to see some empirical figures. > > http://www-users.cs.umn.edu/~hopper/mcon-ccs.pdf > > Cheers, > Michael > _______________________________________________ > Tech mailing list > Tech at freenetproject.org <mailto:Tech at freenetproject.org> > http://freenetproject.org/cgi-bin/mailman/listinfo/tech > > > > > -- > Ian Clarke > CEO, SenseArray > Email: ian at sensearray.com <mailto:ian at sensearray.com> > Ph: +1 512 422 3588 > > > > _______________________________________________ > Tech mailing list > Tech at freenetproject.org > http://freenetproject.org/cgi-bin/mailman/listinfo/tech
