"You write a standard, and get everyone to agree on it, before implementing it." There is a reason things like IETF require "rough consensus and running code"

"You write a standard, and get everyone to agree on it, before implementing it." There is a reason things like IETF require "rough consensus and running code" - because doing the way you suggests sounds really wonderful, but is pretty much is guaranteed 100% to not work in practice. It is hard to find good examples where it led to either good standards, or compatible implementations. Design by committee is what brought us POSIX, etc. This has worked out .... so well in practice.