Re: [apps-discuss] I-D Action: draft-ietf-appsawg-media-type-suffix-regs-00.txt

[Tony Hansen <[e-mail]>]

On 5/17/2012 1:42 PM, Henry S. Thompson wrote:
> Further also to the recent review.
>
> With my informal W3C TAG and XML Core WG hats on, I want to flag again
> our concern about section 8:
>
> 8.  IANA Considerations
>
>     See the Message Type Structured Syntax Suffix registration forms in
>     Section 3 - Section 7.
>
>     The existing Structured Syntax Suffix registration for "+xml"
>     should be modified to include the following
>
>     Fragment identifier considerations Media types using "+xml" SHOULD
>                         process any fragment identifiers defined for
>                         "application/xml" in the same way as defined
>                         for that media type.  (At publication of this
>                         document, the fragment identification syntax
>                         considerations for "application/xml" are
>                         defined in [RFC3023].) Specific media types
>                         using "+xml" MAY identify additional fragment
>                         identifier considerations, MAY define
>                         processing for fragment identifiers that are
>                         classed as errors for "application/xml" and MAY
>                         designate fragment identifiers defined for
>                         "application/xml" that SHOULD NOT be used.
>
> RFC3023 does not, in fact, embarassingly, define a fragment
> identifier syntax for application/xml documents, so this paragraph is
> at best misleading.

Please re-read the text. It does not say that fragment identifier syntax 
is defined in 3023, but instead says that fragment identifier syntax 
 >>considerations<< are defined there. In particular, sections 5 and 7 
discuss how fragment identifiers are to be treated.

> Also, it's perhaps inappropriate to attempt to override-at-a-distance in
> this way, given that this document obviously will not be documented as
> superseding 3023.

Interesting point -- I should have added an "updates 3023" when I added 
that update to +xml's media type registration. Done.

     Tony Hansen

> The overall intent of this is clearly a Good Thing, but the timing is
> tricky. . .
>
> I will try to get some concrete information about the chances of a
> getting a draft RFC intended to replace 3023 as a matter of urgency.
>
> ht