Re: [MIB-DOCTORS] [OPSAWG][OPS-DIR] http://www.ietf.org/internet-drafts/draft-ietf-ippm-storetraceroutes-10.txtandRFC4560

Hi -

> From: "Juergen Schoenwaelder" <j.schoenwaelder@jacobs-university.de>
> To: "David B Harrington" <dbharrington@comcast.net>
> Cc: <ops-dir@ietf.org>; "'Rob Austein'" <sra@hactrn.net>; "'MIB Doctors (E-mail)'" <mib-doctors@ietf.org>; "'tom.petch'"
<cfinss@dial.pipex.com>; <opsawg@ietf.org>
> Sent: Wednesday, August 13, 2008 6:45 AM
> Subject: Re: [MIB-DOCTORS][OPSAWG][OPS-DIR] http://www.ietf.org/internet-drafts/draft-ietf-ippm-storetraceroutes -10.txtandRFC4560
>
> On Wed, Aug 13, 2008 at 09:27:29AM -0400, David B Harrington wrote:
>
> > What we found effective in the MIB community is a centralized
> > registration system (IANA) with hierarchical namespaces, in which the
> > centralized authority can delegate management of modular namespaces to
> > standards efforts (usually WGs) and to enterprises, for registering
> > descriptors and registrations within the hierarchical namspace, plus a
> > rule that all descriptors and registrations within the
> > standards-effort portion of the registration system MUST be unique.
> > The MIB system has been working quite well for twenty years.
>
> Sorry, I can't follow. There is nothing delegated to WGs as far as I
> can tell. And I am not aware of any IANA registration systems for
> module names or global TC names.
...

The CLR that IETF descriptors should be unique isn't a necessary
property of the SMI.  The CLR was added for two reasons:
  (1) *humans* generally don't qualify descriptors with module names,
        and are easily confused in ways that a properly implemented
        MIB compiler would not be confused.
  (2) back then, some MIB compilers didn't distinguish descriptors
       coming from different modules.  This is clearly a bug, but a very
       common and wide-spread one.

So Juergen's quite right about the technical non-need for descriptor
uniqueness in SMI.  Though reason (1) might be worth considering
as a basis for a possible similar CLR in XML-based work, I do hope
that we can avoid (2), which is ultimately much more pernicious.

Randy

_______________________________________________
MIB-DOCTORS mailing list
MIB-DOCTORS@ietf.org
https://www.ietf.org/mailman/listinfo/mib-doctors