[
Lists Home |
Date Index |
Thread Index
]
At 01:50 PM 12/4/2002 -0500, W. E. Perry wrote:
>Jonathan Robie wrote:
>
> > But all data has a lexical representation, whether or not it also has a
> data
> > type.
>
> From my point of view, this is exactly backwards, and in that difference of
>perspective is the explanation for where you and I have failed to persuade one
>another.
I'm really not at the point that I have understood you and disagree. I
simply don't understand what you are saying.
>The lexical form is not the representation of data. Data--meaning, as I
>think you do, semantically rich information including, in your specific
>understanding, datatypes--is endowed with specific semantics by either the
>actuality or the possibility that it is, or will be, used in a specific
>way by a
>specific process. You look at my lexical form in light of the purchase order
>processing or maintenance scheduling or address reconciliation process
>which you
>expect to perform against it and you see in my text--reflected from your
>intent of
>how it be used--the datatypes and other properties which you require. This
>leads
>you to believe that my text is accurately understood as the serialization or
>lexical representation of your abstract data structure. But it is that only
>imaginable from the perspective of your process. On its own terms, it has no
>abstract progenitor, no fluid preverbal Gestalt. And on its own terms it
>is vastly
>more useful that the particular instantiation of your abstract form, precisely
>because it may be locally instantiated, and prove locally useful as many other
>data forms than you narrowly envision.
Could you please give a concrete example, with angle brackets and code,
that illustrates the problem? I agree that processing may impose semantics
on data that was not envisioned by the creator of the data, but you seem to
be saying more - that providing data types in XML data interferes with
unforeseen reuse.
Please illustrate this with a concrete example.
Jonathan
|
