On 5/18/11 11:33 PM, glenn mcdonald wrote:
There's nothing wrong with SPARQL. Likewise, there's nothing wrong with RDFFine beliefs to have, I guess. I wish I had them, too. It would save me a lot of work.It isn't about separating the ID from the Name. It is all about separating Names from Addresses.We have different ideas about what "it" is. By "ID" I mean how the machine identifies the thing.
So machine oriented Identifiers. Otherwise known as machine generated synthetic identifiers e.g. uuids.
By "Name" I mean how the human identifies the thing.
Natural identifiers as exemplified by natural keys.
These are fundamental concepts inherent in machine-hosted data, no matter what its exposure or scope. "Address", to me, is a trivial tertiary thing derived from ID and machine context.
Yes, but to be clearer this is about Identifying locations e.g., Memory Addresses.
I support open data and web data and HTTP/REST-based representations of machine-context, and internationalization, and thus IRIs, but sometimes you'll care about those a lot, and sometimes you won't care at all. In my data-work I find that I sometimes care about Addresses, but I /always/ care about IDs and Names...
Absolutely! The focus is on the Name, but the big problem we have in the Linked Data realm boils down to not wanting to talk clearly about the expectation and behavior of Names.
World Wide Web users are used to Location Names (URLs) as Links and now we are moving them towards Links in the form of generic Names that resolve to Data Objects (Resources) at a Location.
Anyway, we agree (typically) until you point the finger at SPARQL and RDF which to me (like OWL) are all subject to conflation obscured narratives that ultimately comprise their respective virtues.
I continue to aim my concerns at the poor narratives rather than the messengers :-)
-- Regards, Kingsley Idehen President& CEO OpenLink Software Web:http://www.openlinksw.com Weblog:http://www.openlinksw.com/blog/~kidehen Twitter/Identi.ca: kidehen
