Re: simple question on namespaces.

[Arjun Ray <aray@q2.net>]

On Fri, 29 Dec 2000, Tim Bray wrote:

> At 09:41 PM 28/12/00 -0500, Arjun Ray wrote:
> >On Thu, 28 Dec 2000, Tim Bray wrote:

>> In a public environment like the Net, where the point is to
>> agree on and share definitions, a "controlled vocabulary" without
>> a means to verify formal validity is magnificently useless.

> Now that's just silly.  There is no machine-processable definition
> of the semantics of HTML or SVG or PostScript or PDF or JPG or
> GIF 

"Controlled vocabulary" != "semantics"

(Just like "declaration subset" != "document type definition" :-))

Whether a "word" is in a "vocabulary" does not need a semantic
explication to be decidable, or does it?  (Even if it does, I quesion
whether this is true for all domains, i.e. a necessary generalized
view of the problem.) 

I've already quoted the examples involving ForgleBurp/Farglebarp et
al, but let me repeat one passage:

: Either the NS URI must *always* points a formal definition of the
: vocabulary (not schema) of the name space so that you (and your
  ^^^^^^^^^^^^^^^^^^^^^^^
: processors) can reliably examine that definition to validate that
: names you've encountered are really in the vocabulary

> or - you name it - but knowledge of how to deal with these data
> formats is self-evidently shared;  

Indeed.  A lot of the time we don't inquire into the exact details of
how the knowledge came to be shared; rather we proceed from the fact
that the knowledge *is* shared.  That is, it's a no-go without the
sharing.

> hence the requirement for MIME technology to identify them.

MIME types are much like notation declarations actually (and the usual
association of system-ids for notations with processors thends to
strengthen the likeness.)

> A similar need exists for chunks of XML, well in advance of us
> having technology to share definitions beyond the syntactic level.

Well, this is the type assetion problem that Eliot Kimber, among
others, have been ranting about for years.

  http://www.syntext.com/topics/sgml/kimber1.html

(More and more I'm convinced that the first three paragraphs of the
XML-Names spec harbor a whopping non-sequitur, never mind that there
was never a point to the first example in A.1 that I could ever
divine.)


Arjun