Re: RDF Model vs Serialization (Language)

On Wed, 2007-03-14 at 16:04 +0000, Hammond, Tony wrote:
> Hi There:

Hi. Thanks for the careful review and feedback.

Can you confirm that it's the recent last call draft
that you were reading?
http://www.w3.org/TR/2007/WD-grddl-20070302/


(This reply is incomplete; I think I'll send it anyway
rather than waiting until it's complete. Please do stay
tuned for a more complete response...)

> I didn't see this covered any place in the mail archives but please
> excuse if it was. And please excuse the following if I'm just completely
> off the wall here.
> 
> But seems to me there is some confusion throughout the WD about RDF the
> data model and RDF/XML one serialization of (or language for) the RDF
> data model.
> 
> Section 1:
> 
> 1. "There are dialects of XHTML, XML and RDF that are used ..."
> 
> Shouldn't that be RDF/XML? Can one really have a dialect of a model? I
> can imagine having a dialect of RDF/XML, i.e. a profile of RDF rendered
> as RDF/XML, e.g. XMP.

It's a bit of a stretch; the RDF dialects here are not so much
syntactic profiles of RDF/XML, but patterns of usage of the RDF
abstract syntax; "schemas to ... ontologies" are suggestive
of RDFS and OWL.

Is it too much of a stretch to look at RDFS and OWL
as dialects of RDF?


> And also the other two are markup languages. I don't think the
> underlying data model - the XML InfoSet - is intended here.

Again, I hope you can read a bit generously, so that "dialect"
means "pattern of usage" in a broad sense.


> 2. "Here's the information contained in the XML fragments above, this
> time expressed as RDF:"
> 
> Well, no. "expressed as RDF/XML". (This one, I am sure is correct.)

The RDF/XML given in that example is notation for an RDF graph;
so it seems reasonable to say that this information is expressed
as RDF, i.e. expressed in the Resource Description Framework.

"RDF/XML" might also be correct, but it would suggest irrelevant
details.



> 3. "or one of the other RDF syntaxes"
> 
> I'm unsure about this. Can one strictly have a syntax for a model?

according to http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/ , yes:

"RDF has a recommended XML serialization form [RDF-SYNTAX], which can be
used to encode the data model for exchange of information among
applications."

>  Or is
> that a syntax for a serialization of the model?
> 
> 4. "GRDDL provides a relatively inexpensive mechanism for bootstrapping
> RDF content from uniform XML dialects,"
> 
> Again, isn't this sloppy? (But here I could be wrong. Maybe I'm just
> getting very paranoid.)

It's a bit sloppy/informal, but to replace RDF by RDF/XML would
be going the wrong direction; the more formal parts of the spec
are actually written in terms of RDF graphs, not RDF/XML documents
(except in one special case).

See also the output-formats issue
http://www.w3.org/TR/grddl/#issue-output-formats

[at this point, dinner time arrives. ]


> 5. "Content authors can nominate the transformations for producing RDF
> from their content"
> 
> Ditto.
> 
> 6. "the transformation will provide a faithful rendition in RDF"
> 
> Now here I really do think this means some serialization of RDF (well
> OK, it's likely to be RDF/XML)
> 
> 7. "that the transformation will provide a faithful RDF rendition"
> 
> Ditto.
> 
> 
> Section 2:
> 
> 1. "which are expected to transform the source document into RDF"
> 
> RDF/XML?
> 
> etc., etc.
> 
> Sorry to be a pedant. But to echo katemonkey here:
> 
> 	"Web standards help designers and developers create the pedantic
> web."
> 	http://thetenwordreview.com/reviews/programming/web+standards
> 
> :~)
> 
> Cheers,
> 
> Tony
> 
> 
> Tony Hammond

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Thursday, 15 March 2007 23:15:14 UTC