RE: What to do when moving from experimental to PS

["Romascanu, Dan (Dan)" <dromasca@avaya.com>]

I do not have a good argument. Just bad ones ;-) Here are two, and both seem to be on your side

- The problem here seems to be the lack of compatibility between an 'Experimental' standard, and its successor at Proposed. There is no such requirement - if I read RFC 2026, Section 4.2.1, it refers to Experimental RFCs as the result of experimental or research work, but not as preliminary phase for standardization. Maybe this is unfortunate, and needs to be changed - but in this case the change needs to be done at another level. RFC 1155 seems to be consistent with this approach, when mentioning that:

The mgmt(2) subtree is used to identify objects which are defined in
   IAB-approved documents.  Administration of the mgmt(2) subtree is
   delegated by the IAB to the Internet Assigned Numbers Authority for
   the Internet.  As RFCs which define new versions of the Internet-
   standard Management Information Base are approved, they are assigned
   an OBJECT IDENTIFIER by the Internet Assigned Numbers Authority for
   identifying the objects defined by that memo.

I read this actually as being allowed to define MIBs with experimental status under mgmt(2) if the IAB approves it, but not the other way. Maybe this is the exception which could have been used by the authors of 2786, if they knew that their MIB will need one day to enter the standards track

- Keeping son-of-2786 MIB under experimental, creates another problem - call it a marketing problem. Some of the customers organizations have very strict criteria of certification, and take very seriously the IETF standards level. Some require a protocol or MIB to be at Draft or even Full Standard before they deploy. Proposing them a 'standard' MIB under the experimental(4) branch might be a problematic offer for the vendors.

Regards,

Dan




> -----Original Message-----
> From: Wijnen, Bert (Bert) [mailto:bwijnen@lucent.com]
> Sent: Sunday, October 20, 2002 3:55 PM
> To: Mreview (E-mail)
> Subject: What to do when moving from experimental to PS
> 
> 
> MIB reviewers/doctors:
> 
> If you take a look at:
> http://www.ietf.org/internet-drafts/draft-ietf-ops-rfc2786std-00.txt
> 
> then you can see that the author wants to keep the MIB under
> experimental, even when we were to approve it as PS.
> I think this is wrong... but author keeps pushing back
> 
> Here is my comment to author and his response:
> 
> >>- I think that when we move this MIB to stds track, that it SHOULD
> >>   be moved from the experimental branch to the mib-2 branch.
> >>   I know that that may cause incompatibilities, but that is exactly
> >>   what the IESG note on RFC2786 warns for.
> >
> >Like I said - there are no substantive changes (and I'll 
> deal with your 
> >comments below in a minute) - why force people to make this 
> change?  Its 
> >the biggest problem I've encountered with trying to get MIBs 
> through the 
> >process.....
> >
> 
> I think my position is supported by RFC1155, section 3.1 and 
> subsections,
> although it does not explicitly state that one could never have a stds
> track MIB Module under the experimental tree.
> 
> Any supporters (with explanation why) for one or the other approach?
> 
> Thanks,
> Bert 
> 
>