Not really, because the only reason they have a location to test against is because the connection looks like it is coming from the vpn server location. They don’t have any other location data to test against, and even if they decided to then run the test against every possible location on the planet, they still have the issue that their data is heavily skewed by the fact your traffic is flowing through a vpn, so your latency is not going to be perfectly matching their test servers unless they force the test servers’ traffic through the same vpn server.
Nothing about this is setup to find your location on the other side of a vpn - it is basically testing if you are using a vpn or otherwise “spoofing” your location and returning a yes or a no.
I was like 3 paragraphs into a writeup about response times, latency, probabilities, etc but I realized you already have all the information and can’t be reasoned with.
You do know what I mean by “response time” right? The recieving computer gets the packets and sends word back. NOT the VPN node, the VPN is not unencrypting traffic to emulate a real computer, it’s instead just relaying the packets TO YOUR MACHINE. VPNs are not the perfect black box void immune to complicated analysis.
Do you not think a VPN will affect response time?? I implore you re-read the paper you keep referring to because they spell it out pretty basically what they are doing - and finding a person’s actual physical location behind a VPN is not it.
I am not claiming a VPN is a perfect or complete solution… The modern web has an absolute ton of ways to track you even through a VPN, but CPV isn’t it.
If you can prove where people aren’t then you can prove where they are.
Not really, because the only reason they have a location to test against is because the connection looks like it is coming from the vpn server location. They don’t have any other location data to test against, and even if they decided to then run the test against every possible location on the planet, they still have the issue that their data is heavily skewed by the fact your traffic is flowing through a vpn, so your latency is not going to be perfectly matching their test servers unless they force the test servers’ traffic through the same vpn server.
Nothing about this is setup to find your location on the other side of a vpn - it is basically testing if you are using a vpn or otherwise “spoofing” your location and returning a yes or a no.
I was like 3 paragraphs into a writeup about response times, latency, probabilities, etc but I realized you already have all the information and can’t be reasoned with.
You do know what I mean by “response time” right? The recieving computer gets the packets and sends word back. NOT the VPN node, the VPN is not unencrypting traffic to emulate a real computer, it’s instead just relaying the packets TO YOUR MACHINE. VPNs are not the perfect black box void immune to complicated analysis.
Do you not think a VPN will affect response time?? I implore you re-read the paper you keep referring to because they spell it out pretty basically what they are doing - and finding a person’s actual physical location behind a VPN is not it.
I am not claiming a VPN is a perfect or complete solution… The modern web has an absolute ton of ways to track you even through a VPN, but CPV isn’t it.
VPN will effect response time and that effect on response time is easily measurable.