- From: Bullard, Claude L (Len) <len.bullard@intergraph.com>
- Date: Tue, 3 May 2005 15:04:19 -0500
- To: "'ht@inf.ed.ac.uk'" <ht@inf.ed.ac.uk>, www-tag@w3.org
Intension is apriori. Identification is post hoc. (No measurement -> no identity). You can say something is logically true but to determine if it is semantically true, you have to inspect the assertions. You have to open the box and ask the cat. The answer is: that's fine. Lexically determined identity is ambiguous by design or it would be impossible to use discrete terms over a continuous space. Syntax is irrelevant to that fact. If a URI is opaque, it cannot be a name in the sense that inspection of a name should provide information about the resource. If it is a name, then the fact of multiple representations means it can be ambiguous even if it can be inspected. You have to open the box and ask the cat. There is no disagreement. There are different operations being applied. If someone makes is axiomatic that an http-prepended URI infers an operation, they mean it retrieves a represenation. If they make it axiomatic that the URI does not infer an operation, they mean it can never be a name, just a string. These are ORTHOGONAL systems that use the same syntax. You can't do BOTH simultaneously. You cannot solve the problem of Schroedinger's Cat without opening the box, and even if you open the box, you have to ask the cat. Resources get you exactly nothing. len From: www-tag-request@w3.org [mailto:www-tag-request@w3.org]On Behalf Of ht@inf.ed.ac.uk So wrt the second of my proposed dimensions [1], Differentiation, that there is disagreement is clear: some parties to this discussion (want to) use ordinary http: URIs as names for concepts, abstractions, properties, etc., i.e. things which are not information resources, and other parties think this is a bad idea, that ordinary http: URIs should be reserved for things you can retrieve. Fine, we all knew that. Now, the hard part -- _why_ does this disagreement arise?
Received on Tuesday, 3 May 2005 20:04:25 UTC