W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > October to December 2000

RE: XML Schemas: the wrong name

From: Bullard, Claude L (Len) <clbullar@ingr.com>
Date: Wed, 25 Oct 2000 10:30:07 -0500
Message-ID: <2C61CCE8A870D211A523080009B94E430503A9D6@HQ5>
To: "Steven R. Newcomb" <srn@coolheads.com>, elharo@metalab.unc.edu
Cc: xml-dev@xml.org, www-xml-schema-comments@w3.org
I agree that interchange of data objects and 
ready-to-run information presumes a lot of 
a priori agreements.  On the other hand, 
it is not hard to use data binding and 
to generate gets and sets virtual interfaces 
from the XML data object.

Isn't that a transformability aspect?  In effect, 
with the schema, one can create a trivial API 
by transforming the data object into the interface 
specification.  That is straightforward data binding.
What is not transformed is the actual implementation 
which is the black box handler.  It is the transform 
of lexical description based on the names.  

Ready to run suggests to me that all of the rules 
for the conditions under which a transform is applied 
are known; that is, the binding rules.

In layered systems, procedural rules bind layers operationallly, 
so in a real sense, these procedures call the 
correct transform based on the event and it is 
assumed that using workflow modeling and testing, 
a behaviorally correct system emerges.  If it is 
behaviorally correct (knowledge is as true as it 
is predictive), then the system functions and the
designer need not be concerned with anything more 
abstract than that.  It comes down to identifying 
similarity based on invariants and aligning the 
transforms into a goal-directed sequence.

What are the other kinds of information?  Is it the 
layering?  Perhaps you could detail the concept of 
ready to run information.

Len Bullard
Intergraph Public Safety

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h

-----Original Message-----
From: Steven R. Newcomb [mailto:srn@coolheads.com]

There is this common wisdom out there that the structure of interchanged
information should also be, in effect, the API to that same
information.  But, in fact, it's only true for a simple subset of the
kinds of information that need to be interchanged, and to which APIs
must be provided.  The fact that the word "schema" is now used to mean
"model for interchange objects" as well as "model for ready-to-run
information" further hides the important distinctions between the two
Received on Wednesday, 25 October 2000 11:30:13 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:49:53 UTC