RE: MIB module numbering for TBD MIB roots

[Bill Fenner <fenner@research.att.com>]

>There is an inconsistency here I think - last time I tried, some other IANA
>assignments (protocol codepoints) were performed before any IESG approval,
>usually at the request of WG chairs. Has this policy changed?

Different registries have different allocation polices.  Some (like
port numbers) are simply first-come first-served.

I think the real point here is that there is already a procedure,
published in a full standard, for getting an OID for MIBs that are working
drafts.  It has a couple of undocumented pitfalls which we came across
when doing the IPMROUTE-MIB, notably that some MIB compilers couldn't
handle having the same module name for the experimental and standard
version, so we had to rename it when moving to the standards track.

But, back to the point - if there's a documented procedure, and many
people violate it now, why would we think that if we documented a new
procedure everyone would follow that one?

  Bill