RDF in XHTML Task Force

Manu: Mark's proposal for how we allow Microformats-like markup is interesting. He wants to do this:

Manu:prefix [ : [ resource] ]   instead of this:

Manu:[ prefix [ : ] ] resource <--- this is what we have right now in the CURIE spec. That would allow you to do this:

Manu:prefix="hcard=http://microformats.org/vocab#hcard" and then <div typeof="hcard">...</div>

benadida: Don't know if I like that.

ShaneM1: Me neither.

<Steven> Well, I think it is quite clever. I like it, it solves a problem in a neat way

benadida: This is the hGRDDL approach that we talked about some time ago - we shouldn't stuff it into the prefix mapping.

Steven: I think it's clever. We do something special already with a CURIE that doesn't have a value without a prefix.

Manu: I think the approach could work if we merge it with the @profile approach.

ShaneM1: It's inconsistent with the definition of CURIE in the spec.

benadida: Mark's saying that we flip it around, but why isn't there a way to create a auto-prefix, like Pascal "with" statement.

Steven: That is another solution.
... The thing with Mark's solution is that you don't need more syntax.
... There is a problem to be solved, and I thought that Mark's solution, while it's too late, is a neat one.

benadida: Forget that it might be too late, it's still important, but it requires that we define a whole bunch of prefixes.

<ShaneM1> prefix extension proposal is at http://www.rdfa.info/wiki/RDFa_Vocabularies

ShaneM1: This proposal stands whether you use @profile or @prefix.

benadida: You're building a DTD for HTML4?

ShaneM1: That's done.
... We need to separate the discussions.
... How do we generate documents that support HTML4 is one discussion.
... How do we support Microformats-like approaches, is the other.

<ShaneM1> http://www.rdfa.info/wiki/RDFainHTML4

benadida: What do we do with HTML4? How do you validate? Do we want to make it more official?

ShaneM1: We're producing a member submission, which is a profile of RDFa for HTML4.

benadida: Creative Commons would be happy to support as a member.
... This is a great approach - we should move faster on it.
... We're doing this to allow people that want validation in HTML4 and want to use RDFa.

Action Items

<scribe> ACTION: Ben ask SWD to approve publication of an updated RDFa Primer [CONTINUES] [DONE] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action01]

<scribe> ACTION: Jeremy review and consider expanding the description of TopBraid in the RDFa wiki [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action02]

<scribe> ACTION: Jeremy to demonstrate GRDDL with XHTML/RDFa once the NS URI is set up. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action03]

<scribe> ACTION: Manu talk with Jamie McCarthy about an AskSlashdot piece [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action04]

<scribe> ACTION: Manu talk with Michael Smethurst at BBC about RDFa [DONE] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action05]

<scribe> ACTION: Manu to create test cases for testing relative URI resolution (href/CURIEs/etc). [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action06]

<scribe> ACTION: Manu to upload test harness source code to W3C CVS. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action07]

<scribe> ACTION: Manu to work with Microformats community to address RDFa as unified markup for uFs. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action08]

<Steven> "Note. For versions of HTTP that define a Link header, user agents should handle these headers exactly as LINK elements in the document. HTTP 1.1 as defined by [RFC2616] does not include a Link header field (refer to section 19.6.3)."

<Steven> http://www.w3.org/TR/html401/struct/links.html#h-12.4.1

<scribe> ACTION: Manu to write summary for Semantic Web Use Cases for Ivan. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action09]

<scribe> ACTION: Manu write a pending test case for literal property and no child nodes. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action10]

<scribe> ACTION: Manu write the perl code for Slashdot. [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action11]

<scribe> ACTION: Mark create base wizard suitable for cloning [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action12]

<scribe> ACTION: Mark write foaf examples for wiki [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action13]

<scribe> ACTION: Michael to create 'RDFa for uF users' on RDFa Wiki [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action14]

<scribe> ACTION: Ralph think about RSS+RDFa [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action15]

<scribe> ACTION: Ralph to make http://www.w3.org/2008/07/rdfa-xslt happen [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action16]

<scribe> ACTION: Shane to update XHTML ns document to point to new XSLT URI [CONTINUES] [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action17]

TAG finding (NoahM)

benadida: This is just to make sure we're in sync with TAG, great response to Noah, Steven.

Steven: I don't quite grasp the issue he is having.

benadida: It's critical - if TAG finding notes that we're not compliant, even if we are, it would be bad.
... I don't think there is an issue.
... Noah is asking, how do you follow your nose.
... We declare @version at the top of the document.
... I think we said that we're modifying the XHTML namespace?
... It's been agreed upon that it will be done.
... That's how we follow our nose, right? We go to the namespace document and we're done, right?

Steven: Is this machine follow your nose, or human follow your nose?

ShaneM1: Machine.

Steven: The namespace document is pure english at the moment, it's not yet RDFa annotated, but it will be.

ShaneM1: nope. The change to the XHTML document is a GRDDL-compliant @profile addition - that's the machine part of this.

benadida: It is follow-your-nose in-so-far as GRDDL is follow-your-nose.

ShaneM1: That's what TimBL told us to do.
... It doesn't point to RDFa, it points to XSLT transformation.
... Tim's declaration is that everything with an xmlns declaration is RDF.

benadida: is there another path to the RDFa specification.
... There is @version.

ShaneM1: There is a doctype declaration.

benadida: What's the normative one?
... The canonical one is the XHTML namespace.
... I'll follow it up, but we want to make sure that they know it is follow-your-nose compliant.
... I believe that we use version="XHTML+RDFa1.0"

ShaneM1: I think Noah's argument is more arcane. His reading of the RFC is that it doesn't permit the interpretation of documents of that media type as containing RDF.

benadida: If that's the case, GRDDL is broken, because GRDDL is supposed to apply to XHTML.
... What we're doing with the namespace document is 100% GRDDL compliant.
... GRDDL is REC and we're GRDDL compliant.
... If GRDDL is okay, why is RDFa not okay?
... @profile says this is GRDDL, rel="transformation" states that is how you extract triples.

ShaneM1: There is no normative Schema document for XHTML 1.0

benadida: Then it's not the same thing as the namespace document.

ShaneM1: The namespace document is the thing at the end of @xmlns

<benadida> ACTION: Ben to figure out *how* to do the namespace-doc GRDDL thing [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action18]

<ShaneM1> http://www.w3.org/TR/grddl/#ns-bind

benadida: I will continue TAG discussion.

ShaneM1: RFC alludes to modularization, but doesn't point to it.
... Modularization doesn't point to RDFa.
... I think that is Noah's argument.

benadida: Yes, but GRDDL.

ShaneM1: He hasn't incorporated GRDDL into his thinking.

benadida: We'll talk about Danny's comment later.

<Steven> (Good call)