[0wing]

It sounds like Lightning is attempting to solve an unsolvable NP-hard math problem regarding routing. [1]

The Lightening network is mostly a gimmick attempt to save the design failures of scaling from the BTC communities refusal to increase network bandwidth in order to save disk space.

That and for the Lightening network to work without opening up a dozen "deposit channels" into everywhere you shop, users will be pushed into larger liquidity hubs, i.e. payment processor middlemen which is ironic given this is exactly what Satoshi's white paper sought to avoid.

[1] https://en.wikipedia.org/wiki/Travelling_salesman_problem

[rspeer]

I am not a bitcoin fan but I can't imagine anything like this would be affected by the NP-hardness of the traveling salesman problem.

It's not like the system breaks if someone finds a way that a payment could have been routed faster. There are quick approximations to the TSP. Also, a payment network doesn't sound like a single object that has to traverse a graph with various costs, it sounds like a network flow problem.

[DalasNoin]

i don't think 0wing got this one 100% right. he is maybe referring to a claim about lightning, that it is going to be private. but this is not true, there will be a full map of the lightning network. david gerard and emin gün sirer(one of the greatest blockchain researchers) have pointed this out:

https://twitter.com/davidgerard/status/940311875820179456

[rspeer]

I see what David Gerard is saying, it's just that having a map is very different from having to find the optimal solution to TSP on that map. We look at maps all the time and don't solve TSP on them.

[DalasNoin]

there is a lot of missinformation going around in the crypto-currency world. i think that after this privacy claim was debunked it morphed into the idea, that lightning was mathematically impossible. somewhere along the line TSP was added to the explanation. 0wing says TCP is "unsolvable NP-hard" which doesnt make much sense. well, its hard to understand the path of rumors. lightning might yield a constant factor scaling improvement with worse security and worse convenience, but it can work.