Re: [core] Fwd: New Version Notification for draft-groves-core-rfc6690up-00.txt

[Christian Groves <[e-mail]>]

Hello Christian,

I think its largely a non-issue. "core.*" falls under the "core" 
exception that says the CoreWG (IETF review) is responsible for 
allocating IDs in the range "core*". IANA pretty much just implements 
what the WG requests. I can add some text around the explicit 
registration of "core.*" under "core*".

Regards, Christian


On 14/11/2016 8:50 AM, Christian Amsüss wrote:
> Hello Christian,
>
> On Sat, Nov 12, 2016 at 06:48:34PM +0900, Christian Groves wrote:
>>>> [The prefix] MUST be followed by a ".".
>>> the "core" prefix claimed in RFC6690 already violates that rule. Should
>>> it be mentioned explicitly as an exception? Should the IETF "hand back"
>>> all "core[^.]" to IANA?
>> I'm not sure why RFC6690 violates the rule? RFC6690 indicates that "the
>> value starts with a lowercase alphabetic character, followed by a sequence
>> of lowercase alphabetic, numeric, ".",...". It then further recommends "that
>> the period "." character be used for dividing name segments.
>>
>> The only core value registered with IANA is core.gp so it follows the
>> recommendation in the draft. If the MUST is a problem then I'm happy to fall
>> back to the "it is recommended" text.
> "violates" was maybe too strong a word; it's just that if we limited the
> reserved names from "core*" to "core.*", "core" wouldn't need to be an
> exception any more (as it is now) but could be just one organisation
> prefix.
>
> On second thought, making 'core.' an organisation prefix would make the
> CoRE WG responsible for publishing (or at least managing) a part of the
> namespace we just handed off to IANA, so maybe it does need to stay an
> exception.
>
> On "third thought": I didn't find that it says anything ruling out still
> registering parameters with IANA inside a registered organisation
> prefix. Can a (general, but of course I'm thinking of "core." here)
> organisation prefix registrant submit a specification that says that
> paremeters need to be registered as individual parameters in the CoRE
> parameters registry (after all, that's still possible), provided it's
> within what IANA will do? (A different working group could then, for
> example, claim a "homenet." organization prefix, set the "RFC required"
> policy or maybe name their own expert without opening up a IANA new
> registry that would receive a delegation from the CoRE parameters).
>
> Is this, in general, something that the draft allows or forbids, and it
> is allowed, is it the way to go with "core."? Or do you think it's a
> non-issue anyway?
>
> Best regards
> Christian
>