Re: A rose by any other name is just as thorny...

On Mar 28, 2010, at 03:25 , Manu Sporny wrote:

> On 03/26/2010 05:32 PM, Ivan Herman wrote:
>> So, let me try to reformulate that to see if I understand
> 
> This is mostly correct, minor nits below:
> 
>> - we have the generic @vocab and @profile mechanism
> 
> Yes.
> 
>> - there is no default set of keywords hardcoded in RDFa 1.1 Core
> 
> Yup.
> 
>> - there is no default @vocab or @profile for RDFa 1.1 Core
> 
> We /could/ have a default @vocab/@profile for RDFa 1.1 Core, but allow
> the language implementations to override it. Granted this makes little
> sense. If we don't do this, then we have to make sure that the languages
> that fold in RDFa MUST define @vocab/@profile
> 

Even if we have a default @profile, I think it would be absolutely minimal. My take is that, if we do define prefixes in a @profile, we could add there the prefixes for the W3C Recommendations, ie, prefixes for RDF, RDFS, OWL, XSD, SKOS, but I really do not see anything else meaningful there. And I do not see what keywords one would add there...

>> - there is no default set of keywords hardcoded in RDFa 1.1 XHTML
> 
> Yes, but there would be a default @vocab or a default @profile.
> 

Yes, probably. What I meant here is that there should not be some sort of a list of a keywords in the recommendation as we have now, but we should use the mechanism we develop elsewhere.

>> - there is no default @profile set for RDFa 1.1. XHTML
> 
> Maybe. We haven't decided whether we care about the junk triples that a
> default @vocab would create. If we care about junk triples, then we may
> want to opt for a default @profile instead.
> 

Yes, I see the discussion on the mail, and Toby's "Person" example is fairly compelling. Ie, we are not only talking about junk triples but also about case-sensitivity (or not). A default @vocab creates problems there, whereas a default @profile takes care of all that...

> The difference is that a default @profile would only create triples for
> keyword values that it knows exist. Granted, an RDFa processor could
> hard-code those keyword values as long as it keeps those values up-to-date.
> 
>> - there _is_ a default @vocab, conceptually set on the <html> element,
>> which is set to the value of "http://www.w3.org/1999/xhtml/vocab#".
> 
> Maybe. See above.

Yeah, probably not:-(

> 
>> The issue I
>> see is how various RDFa processors would know about those defaults. For
>> the (X)HTML case I would probably look at the suffix and/or the media
>> type and 'hardwire' the default that way; that could also work for other
>> XML dialects that happen to have a dedicated suffix and media type...
> 
> We could say:
> 
> 1. Look at the MIME type first if that is available.
> 2. If the MIME type is not available, look at the file extension.
> 3. If neither exist, assume RDFa Core's default @vocab/@profile.
> 

There is still a problem of where one would register the default @profile for a specific mime type. We could set up some sort of a registry for that at W3C, but it all becomes fairly complicated:-( Sigh...
 
Ivan


> -- manu
> 
> -- 
> Manu Sporny (skype: msporny, twitter: manusporny)
> President/CEO - Digital Bazaar, Inc.
> blog: PaySwarming Goes Open Source
> http://blog.digitalbazaar.com/2010/02/01/bitmunk-payswarming/
> 


----
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 Monday, 29 March 2010 10:15:28 UTC