On Thu, Jul 3, 2008 at 2:30 AM, Alan Ruttenberg <alanruttenberg@gmail.com>
wrote:
The latter. The normalization depends on the schema. There are lots of
> schemas and there will be more. So doing so would mean that developers would
> have to update their RDF software each time a new schema was approved.
you mean "scheme" (or "URI scheme")
If we determine that two IRs are the same if asking the WEB about them
> always (would) return the same representation, then because you can't ask
> the web about these other than by http (that's my assumption) and http
> treats them the same, they could be inferred to be the same (by reading the
> http protocol).
*If*... yes. But we haven't been able to get anywhere close to such a strong
statement about IRs. This is a classic extensional view, something that
doesn't follow from anything that's generally agreed about IRs (because
there is nothing agreed about IRs except that they don't have physical
properties). An intentional view (or is it intensional?) could also be
adopted, where two IRs could be distinguished by, say, the process that
generated their representations, even if the representations (and all other
HTTP responses) were completely identical.