Registration was Re: draft-ietf-forces-model-07 update

RFC2434 specifes other options, inclduing Expert Review:-).

I lack a sense of how big this LFB class namespace is initially and how big it
might grow, which makes my suggestions a little vague; even an order of
magnitude would help (and might provoke debate).  Also, I am unclear where it is
initially being populated from; this I-D just says documents..

Tom Petch

----- Original Message -----
From: "Joel M. Halpern" <[email protected]>
To: <[email protected]>
Sent: Thursday, October 12, 2006 4:42 PM
Subject: Re: draft-ietf-forces-model-07 update


> You raise two interesting questions here.
>
> On the first, if I am understanding you, I think you are absolutely
> correct.  We should specifically require that any LFB registration
> have a definition in XML in accordance with the model
> requirements.  That means it has to match the XML schema, and it has
> to match the non-schema requirements for being a valid LFB Class
> definition.  (We did not capture all the documented constraints in
> the XML Schema.)
> We can and should add words to that effect.
>
> That does raise the question of whether there needs to be a reviewer
> to check such things.  And if there is a reviewer, is the reviewer
> checking just the formalities, or more than that?
>
> With regard to the names, I rather like the idea of requiring that
> names in the non-RFC section start with a specific sequence.  I am
> not sure I like X-, since people tend to think that they can make up
> and use any X- name (whether they can or not.) prefixes that occur to
> be are VND-, PRV-, or FCFS-.  Anyone have ideas / preferences?
>
> Thank you,
> Joel M. Halpern
>
> At 03:48 AM 10/12/2006, tom.petch wrote:
> >I would like to see tighter control specified in the IANA Considerations.
> >
> >I see that section as the 'Data Definition' for the database that
> >IANA is being
> >told to set up and so should be as specific as possible, in terms of the
names
> >to be used by IANA and in constraints on what can be put there.
> >
> >I note that the only requirement on a new LFB class is that it be documented,
> >not that it be documented in eg XML; and if XML, which version thereof?
Where
> >an RFC is required, will any do eg Information, with its low level
> >of review or
> >should it be standards track (there have been on or two post-publication
> >disagreements over assignments that slipped through the individual
submissions
> >track bypassing a WG and of course, the submission was purely on behalf of
the
> >individual and not on behalf of the company whose name appears in his e-mail
> >address)?
> >
> >On a semantic point, I think it wrong to allow a free-for-all with LFB Class
> >names, expecting this to lead to chaos and confusion; most registries
> >differentiate Standards Track approved names from not, eg by the
> >presence of X-
> >(SMTP) or @ (SSH), and many impose a rudimentary structure (although I accept
> >that most name spaces are grossly over-structured).
> >
> >Tom Petch
> >
> >----- Original Message -----
> >From: "Deleganes, Ellen M" <[email protected]>
> >To: <[email protected]>
> >Sent: Wednesday, October 04, 2006 1:01 AM
> >Subject: draft-ietf-forces-model-07 update
> >
> >
> >The latest version of the ForCES Model draft is attached.
> >
> >Regards,
> >Ellen Deleganes