Re: The 'provenance' discussion...

On Aug 25, 2011, at 18:07 , Gregg Kellogg wrote:

> On Aug 25, 2011, at 8:33 AM, Ivan Herman wrote:
> 
>> I am afraid we were discussing ghosts...
>> 
>> At least in RDFLib it is absolutely no problem to generate a triple that, when serialized in turtle, is <>. On the other hand, for any RDFa document we have the base, which is well defined in the core spec and which is always a URI. The only problem is that in our _internal_ communications we always used <> incorrectly as referring to what is, essentially, <base>.
>> 
>> Put in another way, it should not be a problem to generate a triple into the default^H^H^H^H^H^H^Houtput graph of the form
>> 
>> <> foaf:page <base> .
> 
> I've always interpreted <> as the same as <div about="">, that is <> is a relative IRI which is expanded by the document's base IRI. This is also the case in Turtle, at least in my implementation. Of course, if the document has no base IRI, then it is the same as URIRef("").
> 

Actually... it is the same but it means something different in different context!

When you are in HTML, the <div about=""> expands into the value of <base>. Unless stated explicitly, base is the URI of the HTML document of origin.

When you are in the serialized RDF, then <> expands into the URI of the... RDF Document's URI! Which is different then the <base> value that is extracted from the RDFa source. It is... the name of the 'named' graph than contains the triples that are in that specific file.

Ie, these two things are automatically different; in other words, there are no contradictions, in fact...

Ivan




> If there is some way to refer to the document's source IRI in the presence of a base declaration, I'm not aware of it, unless it is done out of scope of base, which can be done for XML or SVG, but not for HTML, as xml:base is not used.
> 
>> (or any other predicate that we may want to use there).
>> 
>> Ivan
>> 
>> P.S. Technically, in the case of RDFLib, it is perfectly possible to create a URIRef of the form:
>> 
>> s = URIRef('')
>> 
>> ie, with an empty URI, which can be considered as an identification of the current graph...
>> 
>> ----
>> 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
>> 
>> 
>> 
>> 
>> 
>> 
> 


----
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

Received on Thursday, 25 August 2011 16:32:00 UTC