[topicmapmail] Fragmented XTM for web metadata, and some ontology?

[Murray Altheim]

Kal Ahmed wrote:
> On Sun, 2003-06-29 at 05:59, Robert Barta wrote:
> 
>>On Fri, Jun 27, 2003 at 05:36:39PM +0100, Kal Ahmed wrote:
>>
>>>>So I think that we shoul NOT depricate resourceData.  Instead, we should
>>>>strengthen it so it can specify the type of the literal data it holds.
>>>
>>>Absolutely. Though I go one step further. I don't want a reinvention of
>>>data typing. I just want to be able to include XML from namespaces other
>>>than the XTM one inside the resourceData element. Then I can use
>>>MyFavouriteSchemaLanguage to do not only datatyping but to express other
>>>constraints on the content. Then all I want after that is the ability in
>>>TMCL to restrict occurrences of particular types to having resourceData
>>>content from particular namespaces. Then I would be happy.
>>
>>Kal,
>>
>>Duly noted. But then I would assume that someone might also want to
>>check constraints on the data INSIDE the XML fragment as well.
>>
>>This cannot/should not be done within TMCL but would have to be handed
>>off to the XML constraint language of the author's choice.
> 
> Yes, absolutely. IMHO that separation of concerns is an advantage with
> this proposal over the proposal to include datatypes as part of SAM/XTM.

I suppose this really depends on what one considers as a definition of
the term "datatype". XSD is a constraint language. We can look at the
"Datatypes in XTM" proposal in different lights: first, as a method of
*labeling* (typing) the interpretation of a literal value (e.g., as a
date, a time, a URI, a boolean, etc.); secondly as a method of attaching
a constraint. But it's possible to do both at the same time.

I've not attempted to provide a constraint language (as there are others
spending far more time and effort than I'd ever on this problem), I'm
concentrating on labeling/typing, which is a necessity in its own right.
If if one can label a literal value with an XSD primitive, by extension
there's no reason why one couldn't use the same approach to "label" it
with a custom datatype, where the PSI points to a bit of XML Schema markup
that actually provides that constraint, so *if* an application wished to,
it could retrieve that markup and actually perform the constraint checking.

Murray

...........................................................................
Murray Altheim                         http://kmi.open.ac.uk/people/murray/
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK                    .

        "There's a lot of intelligence out there that you don't
         know if it's true or not."  -- Anonymous US official
         http://news.bbc.co.uk/1/hi/world/middle_east/3014850.stm