>From XHTML 2's standpoint, all you need is the URI for your roles. If that's
backed up with some RDFS/OWL statements that tell you more about it, that's
a great bonus, and it's a consequence of XHTML 2 getting it's relationship
to RDF sorted out. But it is still external to XHTML 2.

Misha and the IPTC's interest though is in putting subject codes in the form
of QNames into the object part of the statement, not the predicate part. So
whereas we can easily do this:

  <meta property="dc:subject">Chess</meta>

we unfortunately have to do this:

  <link rel="dc:subject" href="" />

In other words you have URIs in one place and QNames in another.

(Note that if we're really committed to using DC in XHTML 2, then we don't
need to invent iptc:subject. ;) )

I don't think the solution is particularly difficult, but those interested
in the discussion should be aware that it is also not completely

For example, the most likely solution would just involve adding a @resource
attribute (or some other name) to be used with @property:

  <meta property="dc:subject" resource="codes:Chess" />

It could be used with @rel/@rev instead, but if we keep it with @property
then we have effectively compartmentalised the attributes that concern
'real' linking from those that are about metadata.

However, the moment we introduce this, we get the @about attribute shouting
and waving and saying "I want to be a QName, too". This is required, because
otherwise you can't make statements about the objects you have used in your
other statements, unless you once again use two different syntaxes.

We therefore additionally need to introduce a QName attribute for the
subject position. Since there is no obvious name for this attribute, perhaps
we should reconsider @resource, and call all of the QName versions of our
attributes by the same name as their corresponding attribute but with a 'q'
prefix. This would give us @qabout for the subject, and @qhref for the

But then to be consistent, @rel, @rev, @role and @property should really be
called @qrel, @qrev, @qrole and @qproperty. @rel, @rev, @role and @property
could be retained, but they must now contain URIs instead of QNames. We
therefore end up with:

 Subject:   qabout    about

 Predicate: qrel      rel
            qrev      rev
            qproperty property

 Object:    qhref     href
            qrole     role

(Note that @role is a special type of object identifier where the predicate
is always <>.)

I don't think that using two attributes to represent the same
concept--differing only by syntax--is great.

Personally I would prefer to see a syntax for URIs that copes with both
QNames and URIs in the same attribute--it's what I'd hoped when I first
started work on RDF/A, and is in effect what N3 does. Then it ceases to be
the responsibility of the mark-up, and moves to the realm of the attribute



Mark Birbeck
CEO Ltd.

t: +44 (0) 20 7689 9232

Download our XForms processor from 

> -----Original Message-----
> From: 
> [] On Behalf Of Al Gilman
> Sent: 08 June 2005 14:59
> To: Misha Wolf;
> Cc:;; 
> Subject: Re: DC in XHTML2
> At 6:04 PM +0100 6/7/05, Misha Wolf wrote:
> >Is anyone else in the DC community looking at the use of DC in
> >XHTML2 and thinking along similar lines to us?  We would be 
> very happy 
> >to work with others on this issue.
> It sounds as though you could use @property with the same 
> sort of intermediate dictionary that we think we can do for 
> extensions to the repertory of values for @role.
> Here (If I have it straight) we will have a role dictionary 
> (separate URI from the XHTML 2.0 namespace or the current 
> document instance) which has a named entry which gathers a 
> collection of assertions and then a qualified name referring 
> to this named entry in the dictionary will assert those 
> assertions about the element bearing that QName as a value of @role.
> In the process of developing this we will definitely need 
> assistance from SW [Best Practices?] and would be delighted 
> to compare notes with "the people who brought you NewsML."
> Keep in touch with Lisa or post progress updates or questions 
> to <> for public discussion and 
> <> if W3C member-confidential for any reason.
> Al
> >As some of you will know, the XHTML2 draft [1] introduces a very 
> >interesting syntax for RDF, which many would consider to be an 
> >improvement on the traditional RDF/XML syntax.  No doubt, 
> many content 
> >authors will be seeking to add DC metadata to their XHTML2 content 
> >using these new mechanisms.
> >
> >The News Architecture Working Party of the International Press 
> >Telecommunications Council (IPTC) is investigating the use of
> >XHTML2 for expressing DC and other metadata.  A major 
> problem for us is 
> >the lack of support in the current XHTML2 draft (as in 
> RDF/XML) for the 
> >use of QNames to express terms in controlled vocabularies 
> (aka values 
> >of properties).
> >
> >At the moment, the XHTML2 @content attribute takes PCDATA 
> and the @href 
> >attribute takes IRIs.  There is no attribute available for QNames.
> >
> >We want to be able to use, eg, <dc:subject> with a QName as 
> a value (ie 
> >the object of the RDF statement).  The reasons include 
> legibility and 
> >compactness.  News items (and news headlines) often carry numerous 
> >subject codes, hence the need for compactness.
> >
> >Is anyone else in the DC community looking at the use of DC in
> >XHTML2 and thinking along similar lines to us?  We would be 
> very happy 
> >to work with others on this issue.
> >
> >[1]
> >
> >Misha Wolf
> >Standards Manager, Reuters
> >Chair, News Metadata Framework WG, IPTC Vice-Chair, News 
> Architecture 
> >
> >
> >
> >
> >-----------------------------------------------------------------
> >         Visit our Internet site at
> >
> >To find out more about Reuters Products and Services visit 
> >
> >
> >Any views expressed in this message are those of  the  individual 
> >sender,  except  where  the sender specifically states them 
> to be the 
> >views of Reuters Ltd.

Received on Thursday, 9 June 2005 14:03:16 UTC