That's what I was trying to say... The researchers are deanonymizing transactions from non-Tor connected hosts. So why are we talking about Tor limitations in response to this? Shouldn't we be discussing how to address the issues in Bitcoin proper?

M

On 11/27/2014 9:30 PM, Gregory Maxwell wrote:
On Thu, Nov 27, 2014 at 5:44 PM, <misterbg6@gmail.com> wrote:

I might be mistaken, but it seems to me this paper discusses unintended ways
of obtaining the IP addresses of clients involved in transactions on the
core Bitcoin network.

You're mistaken. :)

If a node is used exclusively via tor it effectively doesn't have a IP address.

(short of bugs of a class that aren't discussed here)

The paper is about fingerprinting approaches that probabilistically
connect transactions to hosts that you can already identify their IPs.