Ira,
I was also very surprised when it was proposed by Bert Wijnen to publish
the MIB as "Experimental" and I did question this decision. Several other
IETF members also agreed that it should be "Experimental". I actually
would prefer "Informational", but as long as it is published they can call
it a "Dumb PWG Specification".
Are you suggesting that we publish 1.0 as "Informational" and 2.0 as
"Experimental" or "Standards Track"? Or only publish 2.0? The goal has
been to get an RFC number on 1.0 and then finish 2.0 after a Bake-off.
It may make more sense to just get 1.0 published however the IETF wants
and then push for 2.0 to be standards track.
Ron Bergman
Dataproducts Corp.
On Wed, 17 Mar 1999, Ira McDonald wrote:
> Hi Ron,
>> Certainly not all (or even most) Experimental RFCs eventually
> lead to Proposed Standard RFCs, BUT only products of IETF
> chartered WGs may be published as Experimental rather
> than Informational. The IETF has NOT chartered JMP WG
> and without changes to the PMP WG charter, the JMP can't
> be a product of the PMP WG (because it's out-of-scope).
>> I think we were naive in trying to take JMP work forward
> a year ago under the Printer MIB WG charter with the IETF.
> If we get an actual free-standing Job Mon MIB WG charter
> from IETF (not so inconceivable now, in light of the
> IPP/1.1 goal of Proposed Standard, with which the IESG
> agrees), then getting published as an Experimental RFC
> makes good sense and is MUCH more likely than getting
> published (initially) as Proposed Standard.
>> If we are not going for Informational (which RFC 2400
> explicitly describes as suitable for standards from
> other organizations, such as the PWG) but rather
> Experimental, then Tom Hastings' good question about
> whether to publish Job Mon MIB v1.0 or v2.0 on that
> track is highly relevant. I personally favor
> publishing v2.0 as Experimental, but reducing the
> conformance level of the last objects added (after
> the mirror table itself) to Conditionally Mandatory,
> so that all existing v1.0 conformant implementations
> are fully compliant with the v2.0 text (such as the
> IBM implementation Harry Lewis' team built during
> the JMP development phase).
>> Cheers,
> - Ira McDonald
>>