W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > May 2007

Re: comment on using RDFa

From: Mark Birbeck <mark.birbeck@x-port.net>
Date: Tue, 29 May 2007 11:58:33 +0100
Message-ID: <640dd5060705290358kac3895bw719194ae15ef8b74@mail.gmail.com>
To: public-rdf-in-xhtml-tf@w3.org

Hi Peter,

> Thanks for the detailed response! I now feel I know a lot more about
> RDFa.... the reason I missed @rev is that it is not in the Primer yet.

Oops...how very true.

> Some further responses:
> >
> > Unfortunately, with microformats you have to define the rules of
> > mixing for each additional vocabulary, so it quickly gets out of hand.
> >
> What I meant is that I can use hCard and hCalendar in a single page and
> the XSLT for the two should still extract the proper information. In
> this sense, they don't interfere with each other.

That's sort of true...but by using multiple parsers it becomes
difficult to tie together information from separate passes, unless an
RDF approach is taken to the data.

> (MFs can interfere
> when there are name clashes, and of course  I realize that RDFa is a
> significant step forward in this sense. Also, no more writing XSLT
> stylesheets per microformat... yuhuu!)


> >
> >> Note that from this code the bookmarklet still produces a number of
> >> nonsensical triples involving the non-RDFa classnames.
> >
> > As I say, I'm almost certain that behaviour is incorrect, but we'll
> > check.
> Great!

I think Ben has clarified this now, and provided a link to a more
recent version of his bookmarklet.

> [snip]
> > As you can see, it generates the same triples that you want, but at
> > the point of parsing foaf:name there is nothing to indicate 'we are
> > now in foaf:Person parsing mode' (and it's a key design goal of RDFa
> > that there doesn't need to be).
> I do think there should be a locality principle... It seems logical to
> me that related information (attributes of the same object) will need to
> appear close to each other on an HTML page and I think this locality
> should dictate the placement of the markup (content first, markup
> second).  In this view, the above is not a logical markup.

This is interesting. I don't think we can remove the ability to refer
to items that are not directly contained within the hierarchical
structure of HTML/XHTML, since that is a crucial way of avoiding
unnecessary repetition with RDFa. However, your original point was to
do with whether the mark-up that indicates the 'type' of an item could
appear on the same element as an item that indicated a predicate, and
I do agree that this might feel odd to a new author.

> >      <span
> >       about="...#peter_mika"
> >        property="foaf:name"
> >        rev="swrc:author" href="...#mika06jws"
> >        class="foaf:Person"
> >      >Peter Mika</span>
> >    </span>
> >
> This should work, I'll check. In this case, how would you add a second
> attribute of the same person?

I would split the subject from the predicates, which has the
side-effect of giving you the more 'logical' layout that you refer to:

   rev="swrc:author" href="...#mika06jws"
    <span property="foaf:title">Mr.</span>
    <span property="foaf:name">Peter Mika</span>
    <a rel="foaf:homepage" href="...">Home page</a>

> [snip example where I list all publications first, and all authors second]
> Again, I don't like this solution as much. It looses locality, e.g. if
> someone copy-pastes only the bit about the publication, he/she looses
> metadata about the author... further, I I would now have to hide somehow
> the author's list at the end (to prevent it from being displayed).

I agree--I was just trying to show different scenarios. So if you had
twenty publications on a page, and there were lots of authors in
common across the publications, then you could have links from each
publication to the many authors, and only have to show the author
details once. But as you say, there are disadvantages to that
approach, as well.

> > Looking at your original mark-up, I think that the only thing you were
> > missing was the use of @rev. And given that you were hampered by the
> > 'non-prefixed class name' problem, that's pretty good going!
> Thanks for all the help! Actually, development felt a bit like
> experimentation, running the bookmarklet and tweaking the code (an XSLT
> stylesheet) until I got the triples I wanted :) It's not too bad, but if
> I can give one more advice: try to think of various ways people will
> abuse RDFa markup... and test the tools against these test cases. I'm
> saying this because I've recently looked at some microformat data
> (vCard) and my quick statistics show that even after tidying only about
> 29% of manually authored hCard markup in HTML pages results in a correct
> vCard. And hCard is still easy compared to RDFa...

I agree. There are some interesting ideas for tools around, that may help this.

> All the best with finishing up the work,

And thanks for the input...it helps a lot.



  Mark Birbeck, formsPlayer

  mark.birbeck@x-port.net | +44 (0) 20 7689 9232
  http://www.formsPlayer.com | http://internet-apps.blogspot.com

  standards. innovation.
Received on Tuesday, 29 May 2007 10:58:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:15:05 GMT