Re: Comments about the RDFa representation in Pingback

On Tue, 22 May 2012 13:21:02 +0200
Dominik Tomaszuk <ddooss@wp.pl> wrote:

> So an example in [2] isn't correct.
> [2]
> http://www.w3.org/community/rww/wiki/Pingback#Writing_a_Pingback_Service

Well, the example doesn't include a @value attribute at all, so whether
@value is supported is irrelevant.

The example would produce these triples:

	@prefix pingback: <http://purl.org/net/pingback/> .
	<> a pingback:Container ;
	   pingback:source ""@en ;
	   pingback:target ""@en .

If the <input> elements had @value attributes with values "foo" and
"bar" it would be:

	@prefix pingback: <http://purl.org/net/pingback/> .
	<> a pingback:Container ;
	   pingback:source "foo"@en ;
	   pingback:target "bar"@en .

Given that both of those properties are owl:ObjectPropertys according
to the pingback namespace document (by the way, the RDF/XML version of
this appears to be an empty file - Turtle is fine), a literal value is
probably not what you want.

> How it can be solved? Ivan, you're member of RDFa Working Group. Do
> you think that support for (X)HTML forms should be added in RDFa
> processing rules?

It's unlikely that any changes will be made to XHTML+RDFa 1.1 apart
from minor editorial ones. Anything major would bust the spec down from
Proposed Recommendation status to a mere Working Draft, and not a lot
of working group members would be too happy about that.

You're probably best off pinning your hopes on HTML+RDFa 1.1 (which
also, counter-intuitively, defines the processing rules for the XHTML
serialisation of HTML5). As I said, @value is already supported there.

To cope with the literal versus resource thing, the processing rules
could possibly be further modified to deal with:

	<input type="url">

HTML5 has a whole raft of input types which could potentially be
handled in different ways; an idea which seems to me to be both
beautiful and horrific simultaneously. (But, we're already doing
regex matching on <time datetime> to determine which date-related
datatype is being used, and it's no more horrific than that.)

-- 
Toby A Inkster
<mailto:mail@tobyinkster.co.uk>
<http://tobyinkster.co.uk>

Received on Tuesday, 22 May 2012 11:54:56 UTC