Quote from: wiebemarten on November 13, 2017, 11:48:45 AM

practical applications for this knowledge

I am inclined to think that there are many practical applications. One application would be high-speed, high-resolution, secure digital timestamping. If you define a formal protocol (e.g. specified by a finite-state machine), you can build a distributed network that processes timestamp requests in a deterministic way. The only catch is that there is no built-in way to measure real-world time (so to order timestamp requests). One way to solve this would be to just cheat off the blockchain and use it as a secure timestamping service. The downside of this is that you are back to 10 minute time-resolution. A self-contained solution would be to use proof-of-work not as a mining system but, rather, as a self-contained time-service within the network. So, the network proceeds along its protocol FSM until it reaches a PoW sync-point. At this point, everyone just waits until a PoW is mined before continuing. The network-wide "nonce" is reset using this PoW so that all old transactions can be instantly recognized as old by checking their nonce. In this way, the network does not need to track its past history yet it can remain secure. The protocol would be able to adjust its PoW but there is no reason that PoW mining would become extraordinarily intense because its only purpose is to serialize the network to keep everyone on roughly the same clock. If you do it right, you can also dynamically partition the network so that timestamp requests can be processed in parallel and aggregated together. Basically, by generating a PoW (separate from the syncing PoW), you can use this as a "partitioning token" so that you can be trusted to generate a timestamp without the entire network having to validate it. This prevents Sybil attacks, while allowing the network to scale horizontally. You could incentivize the whole system by using a cryptocurrency to apportion user fees in some reasonable way to everyone who contributes to processing the network. LN would be perfect for arranging these payments. The real-world applications are literally limitless. Unlike the absurdity of corporations using "blockchain technology" internally, a timestamping service built in this way would make sense for corporations to use internally because it shifts the burden of securely ordering/syncing events onto the timestamping-service (which specializes in this). If the network is designed to achieve sufficient horizontal scaling, it could even handle use-cases like secure, high-resolution timestamping of events in an MMO game. This would alleviate a lot of headaches facing game designers.