Thanks for your comments on the paper. We will certainly be taking these and others into account when producing the final version.

There's a lot to take on with your discussion, but I'm particularly interested in your P3. Do you think this is an achievable goal for payments protocols? With your example of the Ricardian contract, the human readable part could be, but what about questions as to whether the digital signature was correct and questions as to what the presence of a digital signature means (e.g when was the document signed; does it imply the user saw the contract/agreed to the contract/had an opportunity to amend the contract)?

At the moment, disputes which make it to court are expensive because experts do need to get involved to answer such questions. I would however argue that this is because of a failure to follow our principles 2 and 5. There are no procedures so experts have to be creative and deal with building consensus with the opposing side on a case-by-case basis.

If there were widely agreed procedures to follow when resolving disputes, it would be much quicker and cheaper to resolve disputes, possibly avoiding the need for an expert to get involved at all if some of the steps can be automated.

Interesting and relevant thoughts.

The world clearly needs to have complete trust in the basic infrastructure that we all use - including communication networks and payment gateways.

But let's make no mistake: for this to happen we have to rely on something stronger than today's standards (network protocols, encryption algorithms, random numbers generators), or use a provably-safe layer to wrap these provably-unsafe standards.

So far, the only purpose of those standards has been to comfort the masses in a false sense of security to better abuse them - the point you are trying to address.

The only way to get a legitimacy in this process is to start using systems that can be proved inbreakable and stop using systems which vulnerability is a by-design feature.