Re: Closure on Extensibility & RDF

[Henry Story <henry.story@xxxxxxxxxxx>]

Here is one suggestion I was thinking of to move along, quickly and 
seamlessly I hope.

  1. Atom will have an associated machine readable OWL document that 
defines each
   of the objects and properties described in the Atom syntax spec, 
with language
   that mirrors that of the spec.
    1.1 The IETF document is the final arbiter of any disputes

  2. The Atom-OWL document will be placed at the namespace location of 
the Atom spec.
    2.1 it will be retrievable in full by requesting that http location 
using the application/rdf+owl mime type
    2.2 any of the atom properties or Entities definition will be 
retrievable by
requesting the url of those entities in a web browser
    2.3 a link to the atom-owl document will be placed from the IETF 
spec to the
        atom-owl document, specifying that a machine readable version 
of the spec
        is available at that location.
    2.4 And vice versa the Atom-owl document will point to the IETF 
spec as the
        final arbiter of the meaning of the Atom-OWL document


  3. Evolution of the Atom-OWL and Atom IETF specs
    3.1 The aim is that any IETF Atom conformant document, be an RDF 
document
        as described by something like the transformation in [1]
    3.1 during the evolution of the atom OWL development, the atom-owl 
document
       will be under the obligation to do anything it can within the 
syntax
       allowed by rdf and owl to map precisely the IETF document
    3.2 If it is absolutely impossible for some aspect of atom-owl to 
map the
        IETF document, this should give some cause to think of revising 
the IETF
        spec so that such a mapping does become possible. This should 
be rare.

  4. An Atom extended document is an Atom document with additional name 
spaced
    elements the whole of which can be mapped into an RDF graph.


Henry Story

[1] http://www.imc.org/atom-syntax/mail-archive/msg11850.html



On 8 Jan 2005, at 19:49, Tim Bray wrote:

> [On behalf of Paul and myself:]
>
> The opinion has been forcefully expressed that Atom should adopt an 
> extensibility framework based partly or wholly, directly or 
> indirectly, on RDF.  This idea is not unreasonable on the face of it.  
> Thus, the time has now come to put this into a concrete proposal.  
> Since we've talking about it for months, this shouldn't take long.
>
> Would the advocates of this view please, by Monday January 17th, bring 
> the WG a proposal for its consideration.  This could take one of these 
> forms:
>
> - A Pace proposing extensibility-framework changes to the language of 
> the format draft.  (Rob/Joe, any chance of getting that -04 draft I 
> know you're working on published Real Soon Now as a basis?)
> - A new Internet-Draft that co-exists with the existing Atom drafts 
> that describes an extensibility framework.
> - Both.
>
> Note that if you do a Pace and it fails to obtain quick consensus, the 
> ideas may fall by the wayside; on the other hand, if you do an I-D 
> this may find a life of its own and wind its way to RFC-hood even if 
> people in the Atom community remain unconvinced that they need it.  
> Also, work on an I-D would be free of the distraction of Paul and I 
> snarling at everyone to finish it up and ship it right now :)
>
> On the other hand, if the extensibility framework requires only a few 
> simple additions and/or changes to the existing drafts, and there's a 
> chance the WG will be OK with them, then that's probably a much better 
> way to go.
>
> In any case, we've invested enough time arguing in the abstract.  
> Discussions on this subject which are not framed in terms of specific 
> language in specific documents will be ruled out of order by the 
> co-chairs starting Monday January 17th.
>
>  -Tim