Re: RDF URI References

Bjoern Hoehrmann wrote:
> * Nathan wrote:
>> However, if you take a look at the URI spec RFC 3986, you'll see that 
>> "an absolute URI with an optional fragment identifier" is the very 
>> definition of 'URI' and not 'URI Reference'.
>>
>>   URI         = scheme ":" hier-part [ "?" query ] [ "#" fragment ]
>>
>> see: http://tools.ietf.org/html/rfc3986#section-3
>>
>> So, why are we all saying URI Reference, and talking about URI 
>> References when in fact we mean 'URI, not URI-Reference'?
> 
> First, note that "RDF URI Reference" is a term on its own and is not
> equivalent to "URI Reference". As for "URI" versus "URI Reference",
> RFC 3986 was published in 2005, while the RDF specfication is from
> 2004 and references RFC 2396 where the terminology is somewhere be-
> tween different and unclear in this regard. RFC 3986 discusses some
> of the differences.

Ahh, so it's URI Reference from 2396:

  URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]

with the absoluteURI constraint.

Thanks for clarifying, along with Damien,

Best,

Nathan