Re: namespace usage as assertions

At 01:04 PM 6/21/00 -0500, Dan Connolly wrote:
>There's a distinction between the resource that a URI identifies
>and entities that may represent that schema; I don't seem to
>be able to get that point across, no matter how many times
>I cite the URI and HTTP specs. I'll try to paraphrase again,
>slightly differently:
>
>         a namespace name is a URI
>         a URI identifies/points to a resource
>         a resource may be represented by/described by an entity (such as a
>schema
>                 document)
>
>What is "not a goal" is "that it [a namespace name] be directly
>usable for retrieval of a schema". That's quite different
>from saying that it's not a goal that a namespace name
>identifies/points to a resource.

Yes, I do understand what you're saying; I'd put it like this, to make it 
more concrete:

- A URI refers to a resource
- A resource is a bucket
- An entity is what fills the bucket and can be poured out on demand, and
   what fills the bucket may change on a regular or irregular basis

But my point is that nothing in the spec suggests that pointing to a 
resource/bucket *was* a goal; the goal was to be able to use a string that 
had the characteristics of uniqueness and persistence.  I have finally come 
around to an opinion expressed by some others on this list (and off it): we 
made a mistake in choosing URIs/URI references for namespace names.  Your 
explanation above, while accurate in describing the situation today, 
doesn't change my mind about whether we should have made it that way.

         Eve
--
Eve Maler                                    +1 781 442 3190
Sun Microsystems XML Technology Center    elm @ east.sun.com

Received on Wednesday, 21 June 2000 15:24:10 UTC