RDF in XHTML Task Force

Action Items

<scribe> ACTION: Ben to transfer wiki issues to tracker [recorded in http://www.w3.org/2009/09/03-rdfa-minutes.html#action05] [CONTINUES]

<scribe> ACTION: Shane to produce proposed diff re: XMLLiteral change [recorded in http://www.w3.org/2009/07/30-rdfa-minutes.html#action03] [DONE]

<scribe> ACTION: Ben to create RDFa WG charter template. [recorded in http://www.w3.org/2009/08/20-rdfa-minutes.html#action04] [CONTINUES]

Ben Adida: Revoke action to transfer wiki issues to tracker
... Any other topics to add to the agenda?

Manu: Sam might want to talk about process moving forward in HTML+RDFa

Process for moving HTML+RDFa forward

Ben Adida: Let's talk about the HTML WG process and how we can make things easier for moving HTML+RDFa forward.

Sam Ruby: Manu has put out two drafts now.
... The biggest issues seems to be xmlns:, so addressing that head on would be good.
... Splitting out Microdata might be good to work on too.

Ben Adida: My worry is if it gets rejected based on Ian's draft vs. Manu's draft.
... What if there is no opportunity to discuss things on a point-by-point basis?

Sam Ruby: I don't think there will be an outright rejection.
... I think that the only reason we won't publish it as a WD is because it is not ready yet.
... Both Paul and Maciej are fine with the draft and don't see an issue moving forward.

Manu: Are we going to have a poll before publishing?

Sam Ruby: I'd much rather not have a poll/vote and publish if there are no objections (lazy consensus).

Ben Adida: I believe that xmlns: is a broader issue than RDFa.

Sam Ruby: Yes, it is.

Ben Adida: Should we wait off on some of these distributed extensibility issues or push the issue now with RDFa?

Sam Ruby: Wait a bit, but contribute what you can when the conversations come up.

ivan: One procedural question - is the first public working draft dependent on the xmlns: issue?
... I'm concerned that there would be an objection to publishing the document based on xmlns:

Sam Ruby: There's no requirement that publishing requires that there is broad consensus.
... at least for WDs.

Ben Adida: Any other general advice?

Sam Ruby: Nope, glad how it's moving forward.

HTML5+RDFa issues

<msporny> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2009Sep/0087.html

Manu: sent out email earlier this week (url above),
... major issue remaining is xmlns, lots of discussion

<ivan> ivan's addendum

Manu: couple of other small items which also affect RDFa in XHTML 1.1

Ivan: put xmlns aside for now based on Sam's input?

Manu: still need to discuss language in the doc
... hard deadline for 9/18 FPWD
... we need to be on the same page for what goes into the doc
... XMLLiteral issue with preservation with bare xmlns

Shane: your implementation doesn't preserve the default namespace?

Manu: No, it doesn't. I'm talking about the default namespace.

Ben: I think the spec says that default namespace should be preserved.

Manu: in that case, the test cases are wrong.

<msporny> ACTION: Manu to review test cases on default namespace preservation [recorded in http://www.w3.org/2009/09/10-rdfa-minutes.html#action04]

Ben: if we have to change the test cases, is that an errata?

Shane: no, it's just a wrong test case.

Ben: We have consensus that the spec says to preserve xml literals, and the only way to do so is to preserve the default namespace.

Ivan: should the text be more explicit?

<msporny> The preservation of xmlns as well as xmlns:-prefixed attributes:

<msporny> http://html5.digitalbazaar.com/specs/rdfa.html#invalid-xmlliteral-values

<ShaneM> oh - for the record my implementation DOES preserve the default namespace for elements within XML Literals

<scribe> ACTION: Shane to add clarification in errata regarding preservation of default namespace [recorded in http://www.w3.org/2009/09/10-rdfa-minutes.html#action05]

<msporny> The preference of the profile link type to the profile attribute:

<msporny> http://html5.digitalbazaar.com/specs/rdfa.html#the-profile-attribute-and-link-type

<msporny> The XMLLiteral discussion and what sort of errata and changes to

<msporny> HTML+RDFa should be made

Ivan: let's finish the xml literal discussion.

Ivan: the examples and test cases do not have canonical XML for XML literals.
... only RDF/XML has a canonical XML transformation specified.

<ShaneM> Our spec defines the behavior in http://www.w3.org/TR/rdfa-syntax/#s_xml_literals

Ivan: we can say this is not our problem, up to SPARQL, etc.
... or we can canonicalize our examples.

Ben: we said we expected SPARQL to canonicalize

Ivan: absolutely, but it's not spec'ed by SPARQL.
... so we can punt or write something unnecessary but correct.

Ben: what's the cost of punting?

Ivan: maybe an errata clarification?

Shane: we don't normatively talk about canonicalization.

<ShaneM> I have written the following errata text:

<ShaneM> <li>Section 5.5 step 9 - The text of step 9 reads in part "The

<ShaneM> value of the [XML literal] is a string created by serializing to

<ShaneM> text, all nodes that are descendants of the [current element],

<ShaneM> i.e., not including the element itself, and giving it a datatype

<ShaneM> of rdf:XMLLiteral.". For the avoidance of doubt, this means in

<ShaneM> part that the current default namespace of each element MUST also

<ShaneM> be included in the emitted XML literal.

<ShaneM> </li>

<ivan> start of the thread

Ben:what about this: "2.00"^^xs:float

Ben:vs. this: "2.0"^^xs:float

<ShaneM> and "2"^^xs:float

<scribe> ACTION: Ben to send Philip a "consensus of the task force" email that there is an issue, but not RDFa's. [recorded in http://www.w3.org/2009/09/10-rdfa-minutes.html#action06]

<msporny> The xmlns: discussion and clarifications to the RDFa spec:

<msporny> http://rdfa.info/wiki/html5-rdfa-wd-issues#4.3_The_xmlns:_attribute

<msporny> http://rdfa.info/wiki/html5-rdfa-wd-issues#5.2_Preservation_and_Validation_of_xmlns:

Manu: let's deal with the xmlns issue
... first issue, xmlns is "not an attribute"
... second issue, "stop talking about XML in HTML5"
... define processing rules in a form that doesn't talk about namespaces or XML namespaces.
... Henri wants us to define how an infoset, non-DOM processor would parse this?

Ben: sounds out of scope

Manu: maybe an implementors' guide?

Ivan: if I have DOM1 impl, then no namespace, but if DOM2 impl, then other calls need to be made to get at the namespace info. Implementation approach is very different.
... I think that's Henri's problem.
... so an implementation guide might be useful and good.

Manu: that's exactly right.
... there's no *one* way to implement this in a web browser.
... but it is a very straight-forward algorithm.

Shane: it's not just a browser issue.
... but that is a bit of a red herring.
... implementors' guide on wiki?

Ben: +1

Manu: wouldn't be too difficult to write the algorithm for web browsers.

Ivan: perfectly okay to put open issues in FPWD.

Shane: re: what language should be in the doc, for normative purposes?

Ben: we can minimize the # of references to xmlns.

Shane: we already have processing rules specified for RDFa syntax.
... it's okay to include things by reference.

Ivan: touching the processing rules is the most dangerous thing we can do.
... there is a small section that talks about xml namespaces. But most of the spec does not reference xmlns.

<ShaneM> unfortunate text: Mappings are provided by @xmlns.

<ShaneM> Should be "Mappings are provided by [XMLNS]"

Ben Adida: Let's try to come to some loose consensus on these issues.
... xmlns: is bigger than us at this point, maybe we should wait until Tuesday.
... to see what the status of the xmlns: discussion is at that point.
... implementers guide should be on the wiki

<ivan> actually, shane, the whole of section 2 in the processing rule is a little bit out of place there, we could just refer to the CURIE and URI processing...

Ben Adida: let's not do implementation language in the spec.

<ShaneM> In a recent email I said:

<ShaneM> Further, since the RDFa Syntax Recommendation is only concerned about the "syntax" of those prefix declarations, and has no semantic requirements beyond that for the use of XML Namespaces, it should be clear that parts of the Namespaces in XML Recommendation that deal with how XML Namespaces effect the declaration of elements and attributes is irrelevant for an RDFa Syntax - conforming processor.

<ShaneM> (Note - I would be very comfortable adding such language in the RDFa Syntax Errata document immediately.  I will bring it up at the next Task Force call.)

Manu: I'm concerned about any language in the XHTML+RDFa spec, that contains references to XML.

ivan: In the processing rules, section 2, which refers to XMLNS is a bit out of place.
... Maybe we could just refer to CURIE and URI processing.

Ben Adida: I think Mark and Shane spec'ed CURIE to be abstract in its mapping implementation.

ShaneM: That step identifies URI prefix mappings.

Ben Adida: So, we should write up implementation notes on the wiki?

Manu: yes.
... sounds good.

<Ben Adida> ACTION: Ben to put up JS code that implements the xmlns algorithm on "RDFa Implementors' Guide" wiki [recorded in http://www.w3.org/2009/09/10-rdfa-minutes.html#action07]

<ShaneM> http://www.w3.org/MarkUp/xhtml2/wiki/CurieJavascript

ShaneM: That's the CURIE resolution processing implementation in javascript.