RE: MIB transition

["David B Harrington" <dbharrington@comcast.net>]

Hi,

Comments in line. 

> 
> I understand this concern.  However, I have severe misgivings 
> about any
> plan that does not at the outset make provisions for 
> maintenance work.  We
> in the IETF have conceded that we no longer have the 
> resources to maintain
> the bridging-related MIB modules.  So we need to make a plan 
> for the IEEE
> to do it.  And that plan should allow for reasonably fast turn
around.

I don't quite agree with that we made this concession. 

We lost our Bridge-related MIB resources when the only thing we were
allowed to work on was the transltion of documents from SMIv1 to
SMIv2, which provided no new capabilities for management over RFC1493
and RFC2674, and the WG had finished the technical design of the
RSTP-MIB years ago, but we didn't get it published because we were
fixing rfc-editor's nits and boilerplate updates for MIBs and IPR and
on and on. The dragging-on nits with no practical benefits almost
drove me to walk away from the IETF altogether as well.

Our process prevented taking on new work that would have helped people
manage the newly-developed 802.1 technologies, such as 802.1s
(Multiple spanning tree), for which we had multiple people
volunteering to help by submitting MIB modules they has developed
within their enterprises, discussing them on the mailing list, and
some offered to edit (and is now acting as editor in the IEEE for the
MSTP-MIB). 

It's not that we don't have resources; we use them so unwisely that
many companies refuse to sponsor people to do editing in the IETF,
especially for something as low priority to most companies as MIB
modules.

The IEEE seems to have a process that is less bogged down with CLRs,
and has a system to ensure active participation by members, and the
ability to deliver new work much more effectively than the IETF.

That's the reason I've been driving to have 802.1 pick up the work,
and the reason 802.1 wants to pick up the work - they want results,
not endless editing cycles.

> 
> I said before that the IEEE 802.1 WG should have the same 
> authority as an
> IETF WG would have.  But as we all know there are conditions 
> attached the
> authority that an IETF WG has to modify MIB modules -- among 
> them is the
> condition that the mods have to pass MIB Doctor scrutiny.  Would
your
> concerns be addresses if the IEEE could be persuaded to allow 
> review by
> an IETF Designated Expert (read: MIB Doctor) for any actual 
> modifications
> they wish to the IETF bridge MIB modules?  It sounds like 
> they want to do
> that anyway for their own MIB modules.

I don't think we have the right to tell them how their process should
work, and that they MUST pass MIB Doctor review before making changes
to our standards. We would have no authority to force that. If we do
not transfer the rights to the existing documents, then we CAN force
any changes to the IETF standard MIB modules to be put through MIB
Doctor review.

Since they are voluntarily seeking MIB Doctor reviews during
development, their documents are likely to pass such an IETF review
very quickly, so there should be no delaying of IEEE work.  

If, on the other hand, their work does not pass IETF MIB Doctor
review, I would not like to see them publish a poorly designed MIB
module as a competing standard, and declare our documents HISTORIC.
> 
> //cmh
> 
> 
>