Re: RDF statements' equivalence.

[Dan Pokorny]

> Let's have properties with Qnames a:p1 and b:p1, where
> xmlns:a="http://www.foo.com/ns#"
> xmlns:b="http://www.Foo.com/ns#"
> According to [NAMESPACES], those namespaces are not identical as they are
not exactly the same character-for-character, so the a:p1 and b:p1 come from
different namespace and are different.
> But the properties are represented by resources and they have URIs:
> http://www.foo.com/ns#p1 and http://www.Foo.com/ns#p1.
> According to [URI], those URIs are equivalent.

[NAMESPACES] makes a clear distinction between a URI refernce as used for a
namespace and a URI as used (functionally) as a URI reference:

"[Definition:] URI references which identify namespaces are considered
identical when they are exactly the same character-for-character. Note that
URI references which are not identical in this sense may in fact be
functionally equivalent. Examples include URI references which differ only
in case, or which are in external entities which have different effective
base URIs. "

According to this, there is no contradiction, and the two triples are not
equivalent.

However, there is a different problem lurking here.  According to [URI], a
"URI Reference"  by definition means what you have after the fragment
identifier has been removed.  Therefore, by definition,
http://www.foo.com/ns#p1 cannot be a URI reference!  In fact, the RDF M&S
refers to URI plus optional anchor id, not to URI reference.

So [NAMESPACES] uses URI references, and RDF uses URI+fragment identifier.
Hmm, what to do?

> Considering this, are the following (s,p,o) triples equivalent RDF
statements or
> not?
> (r, a:p1, v)
> (r, b:p1, v)
>
> References:
> [NAMESPACES] http://www.w3.org/TR/REC-xml-names
> [URI] http://www.ietf.org/rfc/rfc2396.txt
>

Cheers,

Tom P

Received on Thursday, 8 November 2001 09:57:59 UTC