Re: The 'resource' identified by a namespace name URI should be the namespace

Michael Mealling scripsit:

> Ok. I see the problem here. "Resource" as we (I) have been using the
> term is a _logical_ thing, not a specific set of bits and bytes.

Just so.

> In many cases the electronic representation of that logical thing
> can have many different representations (html, text, gif, png, etc).

Or it may vary over time, or it may not even exist (see the "brick:"
scheme).

> In order to compare two Resources (the logical thing) you can ONLY
> compare their identifiers and thus, since all you have to compare
> is the identifier, if one does not equal the other then the logical
> things they identify are different.

Not necessarily.  If the URIs are equal, the resources are the
same; but if the URIs are not equal, the resources may or may not]
be the same.  In the case of the "data:" URI scheme, distinct URIs
implies distinct resources, but that isn't true in general.

> Right. What our problem seems to be is a terminology conflict.
> In my terminology universe the namespace is the resource but it
> is a logical thing that is only known or handled by its URI

Well and good, but what is the URI of a namespace?  The Namespace Rec
doesn't say.  It says that the *name* of a namespace is (i.e.
has the form of) a URI *reference*.  Not the same thing at all.
The name of a resource is just one of its properties, not necessarily
or typically the same as the URI at all.

The book I was reading tonight has the URI of "urn:isbn:0671578081",
but its *name* is the string "Komarr".  (Yes, I'm a Miles Vorkosigan fan.)

My proposal says how to discover the URI of a namespace; prepend
its name with the string "data:,".  This works whether the namespace
*name* has the form of an absolute URI, an absolute URI with a
fragment-id, a relative URI reference, or whatever.

-- 
John Cowan                                   cowan@ccil.org
	Yes, I know the message date is bogus.  I can't help it.
		--me, on far too many occasions

Received on Thursday, 1 June 2000 20:55:22 UTC