Hacker News Reader

Comment by pontifier | Hacker News Reader

I believe the problem here can be used as a proof of proximity in a blockchain. Imagine 2 specially designed devices that can receive a string, perform a public key encryption on it, and transmit it with very low turnaround time.

A bitcoin block hash is found, and 2 participants begin a back and forth hashing session in which billions of round trips are performed...every minute or so, a transaction containing the latest result is submitted to the network, and eventually one is locked in.

An interested party could then verify that the transaction could not have occurred unless the 2 keys involved were within a certain physical distance from one another.

Not sure what it could be used for yet, but it's something that feels like it could be important for some purpose.

View on HN

Only works if there's no router delays in the middle. The vast majority of consumer Internet connections fail this - I've already got an 8ms ping just to get to the first Comcast node, from my home wi-fi router. It's near instantaneous within Comcast's network, but then loses a couple milliseconds handing off to their fiber backbone provider.

He could use timeout as a proxy for distance because all the e-mail recipients were within the same university, with a direct-switched network. Once it got on the wire, the only latency was from speed-of-light.

You could use this to prove that N devices are spread fairly geographically evenly near the line between two devices you know the location of, if the transport latency was fairly constant.

If you cross this line over multiple countries, you now have a system where you can prove that the devices are spread across some number of the different countries, which probably has some useful properties, like making nation-state attacks slightly more difficult.

I don't think you can prove they have a minimal distance between them, just maximal. Thus, any data stream may have always came from a single country, you can only impose an upper limit.

Sure, you can prove a minimal distance. Think of this problem as a metal chain. Each link of the chain enforces a maximal distance on the nodes it adjoins. The chain is bound on either end to some fixed point.

In an non-constrained case, sure, there's no-minimal link distance. You just have a pile of chain.

However, if you make the chain taut between the two points, say by decreasing the number of nodes in the chain or the maximal distance of links, the minimal distance between nodes then approaches the maximal distance between nodes. When you have a chain constrained to the straight line between the two points, the maximal distance between nodes is equal to the minimal distance between nodes.

That would certainly demonstrate that the keys were brought geographically near each other for some amount of time, which is interesting, but of what use would it be? It can't usefully test where I am, just my key. I can rent a cloud machine anywhere in the world for 15 minutes and put my key there temporarily.

If you had a way of projecting a pattern onto a surface you are taking a picture of, you could receive a hash, project that out onto whatever surface, take a picture, and send back the signed hash of the picture you just took, thus proving that a particular picture was taken in a particular place.

If that picture is of you, you can thus prove that you were in a particular place at a particular time.

photographic proofs are not trustworthy... even with projections.

https://youtu.be/p5_tpq5ejFQ

Reply on news.ycombinator.com