Benoit, > To me, that's a grey area (in the process). Changing an IPFIX IE > description is non consequential (and I guess welcome, as it > clarifies), but what if now, an IPFIX IE type is changed? > I propose that the OPSAWG chairs discuss, engage with IANA, IPFIX IE, > and the OPS ADs here. We'll share the conclusions.
IANA must have encountered this situation before, in other registries. The definitions in IANA are part of the IPFix standard. We don't know who else might already have implemented the fields per the existing definitions - so only backwards compatible and interoperable changes can be allowed, with due process. Don't request fields which might change. I would prefer to reject early assignment requests without an RFC. If the new requests are incompatible with the existing IANA definitions, then the existing fields could be deprecated and new fields added. That should be done in an RFC. P. _______________________________________________ OPSAWG mailing list -- opsawg@ietf.org To unsubscribe send an email to opsawg-le...@ietf.org
