regrets/late and input on items [Re: Telecon Agenda, Thursday 17 April 2008, 1500 UTC]

At 11:51 AM 4/16/2008 -0700, Ben Adida wrote:
>Our telecon is tomorrow, 1500 UTC, 11am Boston.

Regrets; I'm likely to arrive late, as a (~semi-annual) W3C Chairs
teleconference has been scheduled that overlaps the first 30
minutes of our telecon time.  I'll try to monitor irc.

Remember that we have regrets from Shane too (I'm preety sure).

>Please at least skim the Primer in time for the meeting!

Just (barely) did that.

http://www.w3.org/2006/07/SWD/RDFa/primer
 $Id: Overview.xml,v 1.42 2008/04/16 17:41:05 adida Exp $

While I like the new organization (focusing on use cases),
I remain astonished at how much the document has shrunk.
I haven't yet really thought about whether the smaller version
has enough coverage of the material for my taste.

>2) RDFa Primer: ready to go to WG?
>http://www.w3.org/2006/07/SWD/RDFa/primer/

I'll accept a consensus of {Manu, Mark, Steve} on this.

>3) XHTML namespace quick resolution
>- did we agree it would be a *GRDDL* pointer?

I don't quite understand this phrasing of the question.
I've been expecting us to take Fabien's namespace document [1]
and revise it to be suitable as the XHTML namespace document.
(I've been meaning to do this myself and offer the result to the group
to evaluate, but I've not managed to make the time to do this yet.)

  [1] http://ns.inria.fr/grddl/rdfa/

>4) ISSUE-109 and ISSUE-110
>http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Apr/0025.html

I don't think Mark's response [2] is sufficient.  I'd add something like

"We haven't discussed @cite as a group.  When we have discussed
adding semantics to old HTML attributes we find that there are deeper
considerations and the answer is not as clear as first impressions
might lead one to expect.  Mark indicated some of this in his
response [2].  We prefer to defer consideration of @cite for a future
version of RDFa.  Regarding @src, in fact in earlier designs the
Group did consider different semantics for this attribute.  After
considerable discussion, we concluded that in fact it was more
useful to authors to have @src specify the subject of a triple."

Basically, our response is "We didn't think about @cite in detail
and we'd prefer to defer it.  We did think about @src in great
detail and we feel our current design is the more useful one,
for reasons illustrated in [2]."

We could acknowledge the possible confusion in the way
an author needs to override @src and @href and consider
whether the Primer might need to discuss this more.

  [2] http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Mar/0311.html

>6) ISSUE-112: RDFa described in purely functional terms?
>http://www.w3.org/2006/07/SWD/track/issues/112

Personally, I much prefer functional descriptions.  But I'm not
the document editor :).  I don't think it's worth taking the
months now that it would take to reach consensus on a
normative functional description and insure that it is
completely consistent with the current procedural specification.
(I'm not expressing an opinion on the technical difficultty of
this task but rather the practical difficulty we have demonstrated
in producing spec language in a short amount of time.)

>7) ISSUE-114: RDFa for dynamic content?
>http://www.w3.org/2006/07/SWD/track/issues/114

The key question is "Has the use of RDFa for dynamic
information in XHTML pages been considered?"
to which I think we can safely answer "no, but thanks for
the suggestion.  We will add this to a list of things to consider
in a future revision of RDFa."

>8) ISSUE-116: Safe CURIEs using "curie:dc:creator" ?
>http://www.w3.org/2006/07/SWD/track/issues/116

  "This specification does not address using RDFa in HTML
  documents."

True


  "... the mapping of CURIE prefixes will have to be rewritten
  for HTML documents ..."

very likely True.  The separate CURIE spec anticipates this.

We design for our Host Language.  At present, our (only)
Host Language is XHTML.  XHTML has a mechanism that
does exactly what we need for creating (prefix, URI) pairs
and that mechanism is supported by existing DOM
implementations.  We're *far* better off leveraging that
existing implementation than inventing a parallel
mechanism that duplicates xmlns: (and have to consider
proposals for changing the inheritance rules than I'm
quite confident will be made) for RDFa version 1.

-Ralph

Received on Thursday, 17 April 2008 14:04:27 UTC