Re: [salud] Updates to section 6

[Paul Kyzivat <[e-mail]>]

Hi Dale,

Comments inline.

On 3/28/13 8:39 AM, Dale R. Worley wrote:
> I suggest the following edits to section 6 to clarify the registration
> process and to align the syntax terminology with the new ABNF.  The
> changes are marked in the left margin with "-" and "+".
>
> I have a number of questions and comments at the beginning of this
> text.
>
> Comments?
>
> Dale
> ----------------------------------------------------------------------
> --- 07-section-6	2013-03-27 20:09:39.713133697 -0400
> +++ 07-updated-section-6	2013-03-27 20:35:12.872213564 -0400
> @@ -1,224 +1,250 @@
> +[[ Where did this rule come from?
> +     Each <alert-category> or <alert-indication> label MUST NOT exceed 27
> +     characters.
> +   Can/should we maintain this rule given that we allow punycode labels
> +   in FQDNs?
> +   (I have a memory that there is a restriction on host name labels to
> +   be no longer than 27 characters, but I could not find that stated
> +   in any RFC.)

I don't know. I see no reason to keep it.

> +I've removed "and the rules to combine them [i.e., <alert-label>s]"
> +from the description at the beginning of section 6.1 because for
> +registered URNs, the only rule needed is "use only the combinations
> +listed in the registry".

OK.

> +We say "The policy for adding <alert-indication>s may differ for each
> +<alert-category> and MUST be defined by the document describing the
> +corresponding <alert-category>."  But we do not state the policy for
> +the standard <alert-category>s.  And I do not remember what we decided
> +that policy should be.

Note I commented on a related issue re "local" below. *That* category 
needs a different rule from all the others, in that subordinates to 
"local:country" may not be registered via the iana process. But I 
suppose alternatives to default and country may be added as in any other.

Otherwise I suggest we just specify that adding <alert-indication>s 
subordinate to any of the <alert-category>s defined in *this* document 
is Standards Action.

Or we could take the easy way out and simply say that its Standards 
Action for everything. (Except for private ones of course.)

> +I've removed the entries like this:
> +
> +   service:<private-name>         RFC XXXX  Reserved for private
> +                                            extensions
> +
> +because they are not enabled by the registration process, but rather
> +by the extension rules of section 7.

OK

[snip]

>   6.2.6.  The "locale"  alert-category and alert-identifiers
>
>      The following table contains the initial IANA registration for the
>      "locale" <alert-category> and <alert-identifier>s.  The value of this
> -   indicator provides information about the location of the user at the
> -   other side of the call.
> +   indicator provides information suggests that alerting signals
> +   characteristic of the specified location should be used.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>   Liess, et al.             Expires April 5, 2013                [Page 21]
>
>   Internet-Draft               Alert-Info URNs                October 2012
>
>
> - <alert-category>/             Reference  Description
> + <alert-category> or           Reference  Description
>    <alert-identifier>
>    -----------------------------------------------------------
>    locale                        RFC XXXX  <alert-category>
>                                            for "locale" <alert-identifier>
>    locale:default                RFC XXXX  Alerting not location
>                                            specific
>                                            (default value)
>    locale:country:<ISO 3166-1 country code>
>                                  RFC XXXX  Country-specific alerting
> - locale:<private-name>         RFC XXXX  Reserved for private
> -                                         extensions
> ----------------------------------------------------------------------

I'm wondering about the registration for country codes. I don't know if 
it is clear here what IANA should do. Does this mean:

- IANA should go out and get a copy of the ISO 3166-1 list and populate 
it into their own table?

- Specific country codes must be registered by some specification, but 
must conform to 3166-1?

Clearly we *intend* that individual country codes will not be listed in 
the IANA table, but rather are implicitly included by reference. How do 
we say that?

	Thanks,
	Paul