The protocol specifications should be treated as the ultimate source of truth for how the protocol works, how to be implemented and what users and builders can expect on how it behaves. However, writing specs for a protocol is not a novelty and there has been multiple decades of research on how to do this well, from having a reference implementation, standards bodies and/or formal verification to ensure both code and specs behave the same. It is a hard problem and ultimately one that questions how programming is done as a whole.
0 commit comments