- From: Henrik Frystyk Nielsen <frystyk@microsoft.com>
- Date: Mon, 10 Jan 2000 08:12:54 -0800
- To: "Dan Connolly" <connolly@w3.org>, <uri@w3.org>
> We used "URI reference" because "absolute URI" excludes fragment
> identifiers,
> and we wanted
> http://example.net/#vocab
> to be a valid namespace identifier.
>
> But
> ../xyz/
> isn't a namespace identifier, until you combine it with a base
> absoluteURI.
I like to think of relative URIs as a simple model based compression
mechanism which doesn't affect the semantics of the identifier; changing
the relative or the base part has the same effect as changing the
resulting absolute URI. It should therefore always be possible to use a
relative URI wherever you can use an absolute URI
("absolute-URI-with-optional-fragment-id") in which case URI-reference
is the correct construction to use.
Section 5.1 in RFC 2396 gives some mechanisms for finding a base URI but
others may be devised. There is semantically no difference between a
relative URI that can not be resolved because the base can't be
determined and some URI you don't know how to resolve (for example
foo:...)
Henrik Frystyk Nielsen,
mailto:frystyk@microsoft.com
Received on Monday, 10 January 2000 11:14:14 UTC