Minor comments on draft-ietf-problem-statement-00.txt

[John C Klensin]

--On Friday, 28 February, 2003 16:09 -0500 Scott Bradner
<sob at harvard.edu> wrote:

>> they cycled a couple of times
>> including IETF spec text, and trying to keep it synced with
>> IETF versions as they moved forward ("moved around"), but
>> this was horrid.
> 
> this did happen with the ITU-T but their processes were
> reworked  so they could reference "standards" documents from
> other organizations that developed standards - this too
> quite a while for various reasons (RFC 2436 documents part
> of that)
> 
> it was not just the IETF douments that the ITU-T had trouble
> referencing this also included the ATM forum etc
> 
> it was important that it was seen that the IETF was
> developing "standards" and had clear processes to do so (the
> same was true with ISO pointing to IETF docs) - I expect its
> not important that the IETF says it is a SDO or says its a
> SDO+ or it says its an SDO & RO (research  organization) -
> but, in fact, one of the things that the IETF does is
> develop things the world sees as "standards"

Scott,

Certainly in the case of ISO and the PAS business (which we
have, IMO fortunately, never used) and for ITU and several
other groups, there is also a requirement that the
organization whose standards are being cited have some
plausible consensus process.  And then, to a greater or lesser
degree, there needs to be a consensus within the citing body
for making the citation.  

To put this differently, 

* Almost any standards-producing body will permit a
non-normative reference to almost anything.  That isn't an
issue, nor is a normative reference to something that
documents or explains a law of physics.

* To make a normative reference to another standard, everyone
requires that the other body by recognized as competent as a
standards producer.  That means, e.g., that its documents
represent:

	- Some acceptable level of quality
	
	- Some acceptable mechanisms for preserving stability
	and maintaining documents.
	
	- Some acceptable level of consensus among parties the
	referencing organization considers relevant and
	materially concerned.

* Finally, the individual document-development committee
within the standards-developing organization must determine
that a document really meets its needs in order to reference
it.

Now standards-producing organizations (including IETF) have a
history of being fairly arrogant about the criteria in the
second group.  For years, ITU and ISO assumed that no
concensus-determining procedure other than their own was
adequate since they didn't involve formal international review
by established and recognized country-based bodies.  The
former decided we were good enough, the latter generally
hasn't, partially because their procedures for deciding to
recognize someone involves a certain amount of
crawling-on-the-knees by the other body and we haven't done
that.  We've tended to recognize ISO and ITU documents all
along, but to be careful about which documents we reference,
change control, etc.

But it really isn't about terminology or about documented
procedures, but about approach and attitude.

      john