Geoffrey Spear wrote: >RFC 2119 is intended to govern technical specifications for software >and communications protocols
which is a situation where precision and clarity are of great importance. Human-directed law is another such situation. We have debates about the meaning of the rules all the time; do you really think we'd be better off if the rules were *less* clear? > There's no reason I can think of to allow a >discouraged or "deprecated" action in a nomic at all, Look at the current uses of recommendation and deprecation. We deprecate rather than forbid (or recommend rather than require) in cases where it would be a major pain to decide canonically whether the condition is met. Allowing some discretion is good in some cases. I think there's more call for permitted-but-deprecated behaviour in nomic than in network protocols. MMI isn't a blind cloning of RFC 2119. It's using the same structure, but applying it to the issues that arise in nomic. -zefram
