[RDFa] ISSUE-42: the @src attribute

Hi all,

Today, we discussed the @src attribute on IMG, though as pointed out by
Shane we should consider what happens on SCRIPT, STYLE, and OBJECT, too.
Let's attempt to approach this generically, where X in {SCRIPT, STYLE,
OBJECT, IMG}.

<div about="s"
<X rel="p" src="o">
  ... stuff ...
</X>
</div>

clearly yields the expected triple:

s p o .


Introducing @about inside the same element

<X about="s" rel="p" src="o">
 ... stuff ...
</X>

yields the same thing:

s p o .

Now what happens if we introduce @new_attribute, syntactic sugar for
rdf:type? First, without the @about:

<div about="s"
<X rel="p" src="o" new_attribute="t">
  ... stuff ...
</X>
</div>

What does the type refer to? It seem natural that it would be "o" that
is of type "t", meaning the triples are:

s p o .
o rdf:type t .

If we merge the @about into the same element, it seems to me that we
should keep the triples exactly as above. This is the same reasoning
that would apply if we have:

<span about="#me" rel="foaf:knows"
      href="/ralph" new_attribute="w3c:RDFMaster">
...
</span>

I would want the above to claim

<#me> foaf:knows </ralph> .
</ralph> rdf:type w3c:RDFMaster .


In effect, @src should behave like @href.


The remaining question is what to do for contained content inside
<OBJECT>. <SCRIPT> and <STYLE> can't contain HTML elements, so we don't
need to worry about them right now.

Since the contents of <OBJECT> are typically parameters given to the
object, nothing that is rendered, I propose that we postpone this
decision, effectively not parsing RDFa inside <OBJECT>. I can see
arguments both ways for setting @src as the subject or not, and I'd
rather not bias the future of RDFa one way or the other.

Thoughts?

-Ben

Received on Thursday, 12 July 2007 19:01:10 UTC