W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > January 2008

Re: Does @resource/@href complete hanging rels? (was Rethinking @src in the context of chaining rules)

From: Ben Adida <ben@adida.net>
Date: Tue, 08 Jan 2008 11:11:36 -0800
Message-ID: <4783CAE8.3020603@adida.net>
To: Manu Sporny <msporny@digitalbazaar.com>
CC: RDFa <public-rdf-in-xhtml-tf@w3.org>

Manu Sporny wrote:
> Ben, was the statement that we "don't
> want a bunch of triples being generated from everything off of a hanging
> rel." the only argument against @resource/@href completing hanging rels?

It's a big one. I'm surprised you think this is a good thing, since one
of the big arguments you (and others) have made is that we don't want
extra triples where authors don't expect them.

Try thinking about the use case with @href, not @resource.

What's particularly odd is that, under this proposal, it's okay to ask
HTML authors to change how they write their HTML to ensure they don't
get spurious triples. In other words, the default is "watch out, you
might get extra triples, take special care to write your markup so they
don't show up."

Whereas all I'm saying is, "Take special care to use @about when you
*want* triples to show up."

Just write this:

  <div about="#me" rel="foaf:knows">
     <div about="#ben" />
     <div about="#mark" />
  </div>

Asking authors to worry about RDFa because of triples they *don't* want
is poor design, in my opinion. Asking authors to worry about RDFa when
they're trying to generate triples, that's perfectly normal.

Again, with Mark's proposal, if you have a hanging @rel, then you can
*never* have a link inside that element without it completing the triple.

The other issues I've mentioned are:

- parsers are more complicated (extra rules in the current_subject
definition), and no new use cases are enabled.

- it's now more complicated to figure out what @instanceof applies to.
(Because Mark also suggests that @instanceof applies to the @resource
attribute in your example.)

- attributes shift pretty radically. Add a @rel on our examples, and
you've changed the meaning of @resource and added a whole new layer
(which was the complaint against my @instanceof proposal from a while
back). But if you use @about, which initially yields the same thing, and
then add the @rel, the change doesn't happen. Super confusing, in my
opinion.

-Ben
Received on Tuesday, 8 January 2008 19:11:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 19:11:44 GMT