W3C home > Mailing lists > Public > xml-uri@w3.org > May 2000

namespaces as[?] resources

From: Al Gilman <asgilman@iamdigex.net>
Date: Fri, 26 May 2000 13:43:59 -0500
Message-Id: <200005261734.NAA1128804@smtp1.mail.iamworld.net>
To: xml-dev@xml.org, xml-uri@w3.org
At 12:06 PM 2000-05-26 -0400, Paul W. Abrahams wrote:
>David Brownell wrote:
>
>> The "XML Namespaces" specification is quite clear that the
>> purpose of a namespace "Universal Resource IDENTIFIER" is
>> identification, not location.
>>
>> Were the purpose of those namespace URIs to be location
>> (as in:  dereference to get a schema) then the spec would
>> have used a "Universal Resource LOCATOR".
>>
>> URIs are very fit for the purpose of identification.  And
>> that's the task/purpose identified in the namespace spec.
>>
>> There's a clear line between identifying something ("my auto,
>> which has been stolen" -- it's got a Vehicle IDENTIFICATION
>> number [VIN] too!) and locating it.
>
>There's a difference between a VIN and a namespace name.  A VIN identifies a
>particular object such as your car.   A namespace name need not be associated
>with any object, real or virtual, at all.  The problem with URIs is that
>they're overloaded with an irrelevant significance, namely, identifying
>resources.   Any unique identifier would serve the purpose, and one with no
>connotations would serve it better.
>
>But I agree with your main point, which is that the purpose of namespace
names
>is not to locate anything.
>

[I am sorry for the tone of my last outburst.]

Identifying and locating have a life-cycle which is a genuine cycle.  Each
feeds on the other.

The actual problem here is that Tim sees an abstract thingie like a
namespace as appropriate to include within the range of possible values of
'resource' and you are assuming that a 'resource' is some particular object.

A view which may be close to what is driving Tim is that the fact that
<quote> a collection of [entity type and attribute] names have been set
aside, so that multiple instances of XML syntax may share processing by
identifying parts of their infosets with these names </quote>, is a
beneficial capability, a resource.  The identification of the [element type
and attribute] names as posessed of a relationship with certain classes or
methods of proper processing constitutes the resource.  A schema or other
dialect definition document or [open] collection [specified by a query] of
writings which collectively add knowledge about the proper processing of
the corpus of XML instances using these names is an articulation or
encoding of the resource; not the resource per se.

A namespace can't "not be a resource," if it is of any use at all.

Al
Received on Friday, 26 May 2000 14:05:06 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:32:43 UTC