RHX@CORNELLC.CCS.CORNELL.EDU (Dick Cogger) (05/07/88)
John, Sounds like a good Idea to identify the protocol. How about "IP:nnn.nnn.nnn.nnn" as a format. Understand your objection to gateway discovery and obtaining subnet mask, but maybe we who want those things will establish a spec for it and label it optional in the spec. Then at least we can have interoperable implementations for now. Things would always work for workstations with hard-set parameters for their own IP addresses, masks, and default gateway. If as and when there's a better way, we can convert. Seems better to have a handful of implementations which are unoptimal but interoperable than to have a handful of unique ones. -Dick