[tor-talk] Emulating a VPN service with Tor (security vs performance)

grarpamp grarpamp at gmail.com
Sat Nov 3 17:46:23 UTC 2012


>> I hope to test some clearnet 2 hop VPN service with double
>> encryption to feel that part.

> I thin that also one-hop works

Since Tor would not be of benefit there, I must assume there is
confusion and ask people to... please get the definitions right:

HOP = An action you take, a river you must leap, the
distance between two nodes.

ZERO hops = no motion. degenerate, no packets can move.
ONE hop = no nodes between you and the destination.
TWO hops = one node between...
N hops = (N-1) nodes between...

I also showed defined 2 and 4 hop pictures in a previous post.

> and that in general Tor is too restricted given it's threat model.

> Generally my feeling is that Tor should "evolve" into something more
> flexible within the balance of performance vs anonymity.

Yes. I don't mind restrictive options, even by default for the masses
that use it. So long as the same binary may also be configured
to do any other useful/custom things that may come up under the Tor
model... node directories, node selection, onion encryption, circuit
management, isolation models, etc. Tor is still young.

There will be a day when you can open a port on Tor and have
traffic sent to that port travel some fixed path or a dynamic
path governed by certain parameters... independant of how any
other port operates :)


More information about the tor-talk mailing list