Re: On ISSUE-6 (invalid values in @datatypes cause plain literals to be generated)

On May 17, 2010, at 12:21 , Toby Inkster wrote:

> On Mon, 17 May 2010 12:04:40 +0200
> Ivan Herman <ivan@w3.org> wrote:
> 
>> - @datatype="***goofy***" is not an NCNAME ie, it is not a term; it
>> does not lead to a proper CURIE either because (prefix must be an
>> NCNAME); ie, it has to be processed as a URI, yielding the URI:
>> 
>> http://baseuri/***goofy***
> 
> I was under the impression that the TERMorCURIEorURI attributes
> required full URIs, not just relative URI references, though this
> doesn't seem to be mentioned in the current core draft.
> 

Well this was up and down during discussions leading up to RDFa1.1. But, at the moment, the management of TERMorCURIEorURI says (in 7.4)

TERMorCURIEorURI
  • If the value is an NCName, then it is evaluated as a term according to General 
    Use of Terms in Attributes. Note that this step may mean that the value is to be ignored.
  • Otherwise, the value is evaluated as a CURIE. If it is a valid CURIE, the resulting 
    URI is used; otherwise, the value will be processed as a URI.

and processing URI simply allows relative URI-s to be used. I agree an example using relative URI-s would be good but, I believe, spec-wise this is fairly clear.

> There seem to be very few use cases where relative references in
> @property, @rel, @rev, @datatype or @typeof would come in handy.
> 

The fact of being able to use relative URI-s is some sort of a by-product of our great unification of terms, curies, and uri-s... So, in mind mind, the question does not arise in these terms but rather: what is the specific problem that forces us to artificially disallow relative URI-s? 

Note that the frequent usage for attribute values is something like

@datatype="blabla"

which will NOT be a relative URI, but will be interpreted (if possible) as a term. In practice, the situation for relative URI-s arises either in the slightly pathological example raised by Dan or when one fairly explicitly uses something like

@datatype="./something"

or

@datatype="#something"

and I do not see any problem having those... 

Ivan



> Can we clarify this either way in the next draft?
> 
> -- 
> Toby A Inkster
> <mailto:mail@tobyinkster.co.uk>
> <http://tobyinkster.co.uk>
> 


----
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