[rdfweb-dev] FOAF auto-discovery <link> attributes

[Dan Brickley]

* Sami Haahtinen <ressu at ressukka.net> [2003-09-18 21:25+0300]
> On Wed, 2003-09-17 at 21:20, Kevin Reid wrote:
> > Sami Haahtinen <ressu at ressukka.net> wrote:
> > > At first thought, i would have agreed also. But if you think of it,
> > > rel="alternative" is used to define alterenative versions of the _same_
> > > page, for example RSS, layoutless and so on. foaf doesn't really fit
> > > this description. On the other hand, when you think of it some more what
> > > foaf really does is, that it provides extra info about the page like:
> > > author, author relations and so on. 
> > 
> > Taking Dan Brickley as an example:
> > 
> > <http://rdfweb.org/people/danbri/> contains:
> >   ... Email: <a href="mailto:danbri at w3.org">danbri at w3.org</a> (main
> > mailbox); ...
> > 
> > <http://rdfweb.org/people/danbri/rdfweb/danbri-foaf.rdf> contains:
> >   ... <mbox rdf:resource="mailto:danbri at w3.org"/> ...
> > 
> > This is approximately the same information, in an alternate form.
> 
> While the above is correct, the foaf file contains just a fraction of
> the data that the main page has. although the foaf file has some of the
> data (like RSS alternatives contain headlines and links) the same foaf
> file applies to more than one page.
> 
> > Also, danbri-foaf.rdf does not contain any information about
> > <http://rdfweb.org/people/danbri/>, other than that it is Dan Brickley's
> > foaf:homepage. Therefore it is not (primarily) metadata with respect to
> > the HTML document.
> 
> I agree, it's not really metadata and it's not an alternative format..
> so what is it?

Both concepts are underdetermined, so we picked one and got on with out
lives. The notion of 'metadata' is notoriously hard to formalise. It's 
just data. The main thing is we have a reasonably deterministic way of 
looking in the HEAD of an HTML document for hints about nearby RDF, and 
for hinting as to when it might include FOAF markup.  

cheers,

Dan