Re: Tor exit nodes by country

[Paul Syverson <syverson@xxxxxxxxxxxxxxxx>]

On Thu, Jan 13, 2005 at 08:19:50AM +0000, Ferdinand O. Tempel wrote:
> Quoting Mike Perry <mikepery@xxxxxxxxxx>:
> 
> > Thus spake Roger Dingledine (arma@xxxxxxx):
> > 
> > > If you want to ask about specific ports, you can do something like:
> > > http://serifos.eecs.harvard.edu:8000/cgi-bin/exit.pl?ports=25,9030
> > > 
> > > (Your Tor client already parses all of this internally and picks
> > > appropriate exit nodes for your destination IP and port -- so at this
> > > point this is just a more visual representation of the network.)
> > 
> > Not to look a gift horse in the mouth (and this is pretty awesome),
> 
> I, like any geek in town, like statistics and numbers and all that.
> But doesn't all this statistics keeping kinda mitigate the use of tor
> altogether? Publicizing all the (location) information of all the exit and
> intermediate nodes makes it very easy to attack those nodes and their owners
> would one like to. One of the key selling points of this kind of network is
> anonymity. Shouldn't that also count for the nodes and exit points in some way
> (I know it won't be that easy to achieve technically)?
> 

You are right that it is difficult to design a usable workable
low-latency widescale system that hides its nodes but provides OK
anonymity to its users. Until such a design is worked out by someone,
Tor is less ambitious. The Tor network is not meant to be anonymous
itself or provide censorship-resistant publishing (location hidden
servers notwithstanding), rather it provides (some degree of a type
of) anonymity to its users. (Location hidden servers could be thought
of as part of the network, but it is probably better to think of these
as one application of the network.) Tor's main resistance to directed
attacks against its nodes and their operators is to have enough nodes
in enough jurisdictions that these sorts of pointed attacks become,
well pointless.

aloha,
Paul