W3C home > Mailing lists > Public > public-swd-wg@w3.org > July 2007

Re: [RDFa] ISSUE-8: RDF containers in RDFa

From: Ben Adida <ben@adida.net>
Date: Thu, 12 Jul 2007 07:03:30 -0700
Message-ID: <469634B2.6020005@adida.net>
To: mark.birbeck@x-port.net
CC: RDFa <public-rdf-in-xhtml-tf@w3.org>, SWD WG <public-swd-wg@w3.org>

Mark Birbeck wrote:
> I think this could easily be left to a future version, so my vote
> would be for that. :)

I agree that solving it all will be tough, but having at least a
solution for *some* kind of list seems pretty crucial, or else we're
tossing lots of RDF graphs out the window. Given the awkwardness of
rdf:Bag and rdf:Seq, I lean towards looking mostly at RDF Collections,
and leaving containers out for now. But maybe we can find something nice
that implements everything.

Keith said:
> Personally, I would  prefer to be explicit, instead of having syntactic sugar for lists.
> 
> In general, I think it would be good for the RDFa rules to be as simple and consistent as possible, so that it is easy to understand how to write it, and also easy to write parsers.
> 
> With regards to lists in particular, I find lists in HTML much more useful than lists in RDF, and would rather that the two were not so tightly coupled that I couldn't write an HTML list without it generating triples. 

How would you make a collection entirely explicit without using *some*
syntactic sugar? It's not doable without horrible ugliness, you have to
bootstrap off *something* in the HTML.

That said, I'm not proposing that every UL or OL generate a list. But if
you put a @rel on it, for example, then maybe. So as the author, you're
definitely choosing to do this.

The concept of lists is, in my experience, almost always something you
have to somewhat special case. I suspect authors will have to think
about this explicitly, too: "oh, now how do I make a *list* of authors"
so having a specific syntax for lists seems natural to me, and
bootstrapping off HTML lists is also quite natural.

If you have another proposal for achieving lists naturally, definitely
bring it up.

Ivan said:
> Unfortunately, your proposal of 2006 has two major deficiencies:

Oh yes, I should have been clear that I don't think that proposal is
good enough as is, or has the right focus. I agree with you that we
should focus on collections rather than containers.

I lean towards the following, simple solution: we do collections only
for now, where UL and OL both create collections *if* they have been
turned into a bnode using @rel or @rev. @href on the LIs generates URI
elements of the list, otherwise it's literals.

So:

<ul>
<li>foo</li>
<li>bar</li>
</ul>

yields nothing.

<ul about="#paper" rel="dc:creator">
<li>Mark</li>
<li content="Steven">Steve</li>
<li>Ralph</li>
</ul>

yields:

<#paper> dc:creator ["Mark", "Steven", "Ralph"] .

and

<ul about="#paper" rel="dc:creator">
<li href="#mark">Mark</li>
<li href="#steven">Steven</li>
<li href="#ralph">Ralph</li>
</ul>

yields:

<#paper> dc:creator [<#mark>, <#steven>, <#ralph>] .

Feel free to shoot it down :)

-Ben
Received on Thursday, 12 July 2007 14:03:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:29 GMT