On Tuesday 30 Jul 2013 23:38:09 Steve Dougherty wrote: > On 07/30/2013 05:05 PM, Robert Hailey wrote: > > On 2013/07/30 (Jul), at 3:40 PM, Steve Dougherty wrote: > >> ... > > I'm sure Tor does something similar, it would be interesting to know > > how Tor nodes bootstrap. > > As far as I understand it tor bootstrapping is also centralized. I don't > know any specifics. Tor does have obfsproxy (and bridges) to handle > blocking, which seems conceptually equivalent to a darknet node > connecting to a peer which connects to the wider network. [0][1]
Right. These are somewhat harder to block than our opennet seed nodes, because there are more of them, and you have to use a unique email address to get them. But they are pretty easy to harvest nonetheless for only a very small cost. We could do something similar, but anything like this is going to be possible to harvest and block reasonably cheaply. Also they had a protocol exploit that let the chinese block all of them, I'm sure that's fixed now, but it's something we discussed in 2000! On Tuesday 30 Jul 2013 21:40:24 Steve Dougherty wrote: > On 07/30/2013 02:58 PM, Robert Hailey wrote: > > > > On 2013/07/30 (Jul), at 1:21 PM, Matthew Toseland wrote: > > > >> Too long? Too honest? IMHO we need to err on the side of honesty. > > > > IMO it is missing a few Tor downers. > > Good points. > > > (1) people can & have run "exit nodes" to collect passwords > > Could be framed as "attacks on unencrypted connections are still easy." > Is it valid to say that in general usage getting someone to proxy > through your machine (to abuse their connection) is more difficult than > getting them to use your Tor exit node? True. > > > (2) Tor generally uses only 3 hops (e.g. easier Sybil & traffic analysis) > > Is correlation between activity on the same circuit also relevant? This is dubious. Right now we use no tunnels at all. PISCES will use longish tunnels during setup, but shorten them a bit after that (how much is to be determined, there's a tradeoff between tunnel performance and how many hops of the local darknet you see IP addresses for). Opennet will probably use short tunnels eventually. > > > (3) people can & have been busted using Tor-hosted services > > Would it be appropriate to link to the "Hidden Services need some love" > post for more details? [0] Probably. This boils down to it being hard to configure a web server correctly to sit behind a hidden service - surely there are packaged solutions by now? > > [0] https://blog.torproject.org/blog/hidden-services-need-some-love > [1] https://downloads.freenetproject.org/alpha/opennet/seednodes.fref
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl