Do I need to repeat myself?
> Seems like a security disaster, ripe for a multitude of race conditions and consensus attacks on supply, state, and immutability, architecturally completely an autistic dream that is not possible to soundly implement.
I don’t think we should change the entire transaction validation mechanism to add excessive complexity.
Shouldn’t HTLCs accommodate these additional L2s? I don’t understand why, if you want a chain with alternative features and functions, you can’t just use HTLCs like Lightning does. Lightning already does about a billion dollars worth of transactions monthly, seems like the concept of using HTLCs to interface with an additional layer is pretty proven. It seems the necessary interface is already there. I guess I am just too stupid to understand.
BIP300 is one of those “ideas guy” moments where something very bold is proposed without a lot of mind put towards the implementation being completely nightmarish.
