RE: RDF Web Apps ISSUE-124 (was: Re: Request for Review of RDFa Lite FPWD)

Thank you for opening an issue on my comment.  Here is a more detailed rationale for my view.

I think a more realistic representation of "the dependencies" is the following since HTML5 + RDFa has BOTH HTML5 and RDFa 1.1 as normative references in http://dev.w3.org/html5/rdfa/#normative-references:

+----------------------------+
|         HTML+RDFa 1.1      |
+-------------+-------------+
|    HTML5    | RDFa 1.1    |
+-------------+-------------+

Therefore if you replace RDFa 1.1 with RDFa 1.1 Lite in this diagram since it is a proper subset or profile then in my view HTML + RDFa 1.1 should be modified to be dependent on both RDFa 1.1 and RDFa 1.1 Lite.

+--------------------------------+
|              HTML+RDFa 1.1        |
+-------------+-----------------+
|    HTML5    |   RDFa 1.1         |
|                      |         or                |
|                      | RDFa 1.1 Lite   |
+-------------+-----------------+

FTR, I am NOT trying to stop or delay the publication of the FPWD of RDFa 1.1 Lite especially now that you have opened an issue for my comment(s).  

/paulc

Paul Cotton, Microsoft Canada
17 Eleanor Drive, Ottawa, Ontario K2E 6A3
Tel: (425) 705-9596 Fax: (425) 936-7329


-----Original Message-----
From: Manu Sporny [mailto:msporny@digitalbazaar.com] 
Sent: Wednesday, December 07, 2011 8:50 PM
To: public-html@w3.org
Subject: RDF Web Apps ISSUE-124 (was: Re: Request for Review of RDFa Lite FPWD)

On 12/07/2011 03:33 PM, Paul Cotton wrote:
> Thank you for forwarding this review request to the HTML WG.

Thanks for reviewing the specification, Paul. :)

> 1. Document conformance
>
> Section 3.1 "Document Conformance" states: "In order for a document
> to claim that it is a conforming HTML+RDFa Lite document".   To me
> this means that the RDFa Lite 1.1 specification not only specifies a 
> proper subset (aka profile) of RDFa 1.1 but it also profiles the HTML 
> + RDFa Working Draft [1] and provides a new definition of document 
> conformance for the combination of HTML5 and RDFa 1.1 Lite.
>
> I am not sure if this is the best model.  For example the RDFa 1.1 
> specification itself does not define "document conformance" but this 
> is found in the HTML + RDFa Working Draft.  Another alternative 
> instead would be to define this new document conformance level in the 
> HTML +RDFa Working Draft by adding a reference from the HTML
> +RDFa Working Draft to the RDFa 1.1 Lite specification.    Did you
> consider this way of defining the two different "document 
> conformances" for HTML + RDFa and HTML +RDFa Lite?

This is now an open issue:

https://www.w3.org/2010/02/rdfa/track/issues/124

To clarify - RDFa Lite 1.1 only specifies a proper subset (profile) of
HTML+RDFa 1.1. RDFa Core 1.1 does not specify any document conformance
requirements because only a RDFa host language specification (such as HTML5, or XHTML1) can do that.

We had discussed whether or not to even have the document conformance section in the RDFa Lite specification since the requirements are exactly the same as the HTML+RDFa Document Conformance requirements.
Google wanted something definitive stated as to what constitutes a legal RDFa Lite document and this was the best we could come up with before the FPWD vote. The conformance requirements for a legal RDFa Lite document are exactly the same as the conformance requirements for a legal HTML+RDFa document. We will probably go back to Google and ask them to see if this is what they wanted, or if we could simplify things by just leaving this section out.

The RDF Web Apps WG would like to not complicate matters by having functionally different conformance criteria from what is already expressed in the HTML+RDFa spec. We had considered your suggestion, but we did not want any pointers going from the HTML+RDFa spec to the RDFa Lite spec. RDFa Lite is meant to be built on top of HTML+RDFa 1.1 because it has largely been the HTML community (working with RDFa) that has asked for a lighter version of RDFa (in that language). The specs are layered like so, for now:

+--------------------------+
|       RDFa Lite 1.1      |
+--------------------------+
|       HTML+RDFa 1.1      |
+---------------+----------+
| RDFa Core 1.1 |  HTML5   |
+---------------+----------+

> 2. RDFa Lite Conformance
>
> In addition shouldn't Section 3 Conformance explicitly state that only 
> the attributes included in RDFa Lite specification can be used in a 
> HTML + RDFa Lite document, etc.

We definitely do not want to create different conformance criteria between HTML+RDFa and RDFa Lite. This would complicate implementations, would complicate authoring, and complicate the validator suite. We want any RDFa used in HTML to be conforming and simply leave it at that.

RDFa Lite is meant express a best practices subset for beginners while giving Google the chance to say that they're following an official REC-track specification and not something that they just made up. We're trying to find a way to balance all of these requirements without creating a document that further complicates RDFa.

-- manu

--
Manu Sporny (skype: msporny, twitter: manusporny) Founder/CEO - Digital Bazaar, Inc.
blog: The Need for Data-Driven Standards http://manu.sporny.org/2011/data-driven-standards/

Received on Thursday, 8 December 2011 16:30:15 UTC