RE: registration OIDs: do the values matter?

["C. M. Heard" <heard@pobox.com>]

On Mon, 24 Feb 2003, Wijnen, Bert (Bert) wrote:
> Oh... an idea might be that we ask IANA to assign
> a 999999 under mib-2 and under experimental and that
> we name it something like
> 
>   999999 -- placeholder for new modules, not a real OID branch. 

One hears in this an echo of the suggestions in

http://www.ietf.org/internet-drafts/draft-narten-iana-experimental-allocations-03.txt

> That way it is still kind of risky, but we then at least have
> documented the intention of that number.
> 
> Just thinking aloud here.

I tend to think that having placeholder OIDs is not really necessary
since the OID number space is practically unlimited (unlike the
situations contemplated in the above-referenced I-D).  We have the
whole experimental subtree (or rather the part of it that remains
unallocated) as a playpen, so if we need experimental codepoints we
can just go get them.

> I had once suggested that all MIB compilers should accept xxx
> as a acceptable number... but that seems not to work.

I see the non-compilability of { mib-2 xxx } and friends as a
feature, not a bug.  It makes it clear that proper number has not
been allocated and that the stuff in the module can be changed
arbitrarily.  Putting in a temporary replacement for compilation
purposes is is not so hard.  Indeed, the fact that you don't have to
do that with real OIDs that will later need to be changed makes it
easy to miss the change when it does come.  That comment applies
equally to { mib-2 999999 } IANA-allocated experimental nodes.

//cmh