Re: Linking to @about in RDFa Lite?

Hi Sebastian,

On Oct 31, 2011, at 17:11 , Sebastian Heath wrote:

> Hi,
> 
> Looking at section 2.3 of the new RDFa Lite draft [1], it starts:
> 
> If you want people to link to things on your page, you can identify
> the thing using a hash and a name. For example:
> 
> <p vocab="http://schema.org/" about="#manu" typeof="Person">
> 
> 
> 
> Is that correct? I think it implies that the presence of @about is
> enough to get a browser to move to that part of the page.

I do not think that is true. @about does not necessarily refer to a clickable link; it just defines a URI (with the base value). A URI with the #manu is perfectly valid even if it does not correspond to an @id somewhere in the file. It is just not clickable directly. Well... it is clickable according to the way links are managed, it just would not lead to anything particular when the fragment is managed.

There is a huge discussion on the various RDF lists, and even on the TAG now, on how to use and interpret 'hash URI-s'. But RDFa is oblivious to this discussion, it just generates URI-s as the user sets them.


> 
> Should the html be
> 
> <p vocab="http://schema.org/" id="manu" about="#manu" typeof="Person">
> 
> 
> I'm far from filing a bug because I may misunderstand the markup.
> 
> 
> Broadening the conversation, having to put both @id and @about on
> elements to define RDFa patterns that are "about themselves" has
> always struck me as odd. But perhaps it's an inevitable cost of RDFa
> being a graph-based, distributed format. Or perhaps we're in the
> rabbit hole of information and non-information resources.
> 
> I guess I'm more broadly raising the topic of a convention for
> marking @id attributes as visible to an RDFa (lite?) processor. Has
> there been any thinking on that? I understand it might mess up
> backwards compatibility. But I could really (really!) use it to
> simplify my markup.

Actually, I was not on the RDFa 1.0 group when this was discussed; but I believe the separation of @id and @about, for the reasons I said above, was a reasonable decision. We could, of course, discuss the possibility for @id to be _accepted_ by an RDFa processor as an @about (if @about is not around, that is), just like it accepts @href in the absence of @resource, but that has not really come up in various user feedbacks and requests so far...

Ivan


> 
> -Sebastian
> 
> 
> 
> [1] http://www.w3.org/2010/02/rdfa/drafts/2011/ED-rdfa-lite-20111030/#about
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf