W3C home > Mailing lists > Public > public-rdfa-wg@w3.org > November 2011

Re: ISSUE: Using @id to set subject in RDFa

From: Jonathan Rees <jar@creativecommons.org>
Date: Sun, 20 Nov 2011 09:10:16 -0500
Message-ID: <CACHXnarJ1QXT-0qeT8TfHbpeGpbeUhSstp4QbMTbn4Lgk_tZcA@mail.gmail.com>
To: Toby Inkster <tai@g5n.co.uk>
Cc: Sebastian Heath <sebastian.heath@gmail.com>, RDFa WG <public-rdfa-wg@w3.org>
On Sat, Nov 19, 2011 at 3:55 PM, Toby Inkster <tai@g5n.co.uk> wrote:
> On Sat, 19 Nov 2011 13:02:24 -0500
> Sebastian Heath <sebastian.heath@gmail.com> wrote:
>> Using @id to set subject in RDFa
> I think that this would get us into a bunch of technical arguments with
> TAG.

Might we say "discussions"? I don't think all TAG members are
inflexible, or even all in agreement on this issue.

I for one understand both perspectives and would like to see rational
decision making and consensus building around the question. We've been
discussing this on www-tag and would welcome participation there
(although I recognize it doesn't always seem a welcoming place).

We can change the architecture, or we can say that RDF and the Web
have different architectures, or we can try to strengthen the TAG's
traditional point of view. I think any of these approaches would be
viable; it's a mere matter of standards-making (or not).

>        <section id="toby" typeof="foaf:Person">
>          <h1 property="foaf:name">Toby Inkster</h1>
>        </section>
> What does the URI <#toby> identify? A living person? An element in a
> DOM? A chapter within a document? Some combination of those? In RDF
> (and to a certain extent, web architecture in general) it's generally
> considered pretty important that a URI must identify a single concept.
> The following:
>        <section id="about_toby" about="#toby" typeof="foaf:Person">
>          <h1 property="foaf:name">Toby Inkster</h1>
>        </section>
> avoids such muddiness.
> Or you could use this:
>        <section about="#toby" typeof="foaf:Person">
>          <h1 property="foaf:name">Toby Inkster</h1>
>        </section>
> In combination with some Javascript that takes a URL fragment and uses
> it to focus and highlight some areas of the page.

This is my view too, although I think I like even better the practice
of having only one (maybe 2-3 very closely related) about="#x" in each
file, so that the Javascript isn't necessary. Personal opinion.

But it's important to make sure that people feel they have been heard
when they say they don't get why conflating the element and the about=
thing is a problem. This is very hard in email.

A concrete example to consider would be a page aggregating RDFa
content available from multiple sources under multiple licenses (e.g.
an Encylopedia of Life page). It needs to say that HTML or XML
elements are licensed in particular ways and give attribution. The
license subjects would be named by the id= URIs. But the elements will
be about things that are not those element (perhaps they are *other*
elements), and those things would have about= URIs. If you want to
change the party line, you have to explain how to redo this case.


> --
> Toby A Inkster
> <mailto:mail@tobyinkster.co.uk>
> <http://tobyinkster.co.uk>
Received on Sunday, 20 November 2011 14:10:56 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:05:27 UTC