> benefit. Use of MUST (in RFC 2119 sense) invites lazy thought in the protocol > design itself, where details need not be sold as beneficial to individuals. We > should say, there is no RFC > 2119 MUST - there is only self interest.
I think you are misreading the intent behind RFC2119 a bit... The idea has always been that implementations MUST follow MUSTs -- but who is this enforced by? Vendors? They've never cared. Neither the IETF nor the Internet Society run jails that I'm aware of, nor is there any nation state who offers their jails up for folks who don't follow a MUST. The situation in the routing world is the same as it is in the cryptocurrency world -- you can make real money by not following the rules (there seems to be this general idea that the 'net is "free" -- which is wrong), and the only enforcement on the rules is the community at large. The correct interpretation of RFC2119 MUST is this: - If I write an implementation, and you send me something you shouldn't (because it's MUST), then I simply refuse to interoperate with your implementation. This (hopefully) harms the implementation financially. - If I run a network, and you produce gear that doesn't follow a MUST, then I won't buy your gear. Note that private implementations are not bound by these rules -- implementations that are intentionally not designed to interoperate with other implementations. If such implementations become large enough to care, then they go through the standardization process so everyone has a look before it becomes an issue within the community. 😊 /r _______________________________________________ Lightning-dev mailing list [email protected] https://lists.linuxfoundation.org/mailman/listinfo/lightning-dev
