XML Schema:Datatypes

Hi W3 XML Schema group,

While reading your Schema:Datatypes specification, I 
am puzzled by the potential for removing the Number datatype.
I also do not understand why you do not further break Number
down into at least whole and floating point numbers.  However,
the more serious problem is possibly omitting numbers from
the spec all together.  This truly baffles me: I would tend to 
believe that the simple dichotomy of text (string) versus number
(like floating point value for money) would be the most obvious
and critical requirement of datatypes in an XML document.

Now, I have some idea of the problem the editor's note mentioned
in the Number section that there are many extant lexical representations
for Numbers.  However, I do not see the issue of numerous existing 
representations being an adequate justification for omitting a central
datatype -- that's why standards are powerful because they limit the 
field from what is possible to what is generally useful.

As for numerous lexical representations -- I don't see the problem with
using the same lexical representation that Java uses for numeric literals.
While I'm sure there is more complexity to the issue than that, I find it
hard to believe it is not possible to create consensus on a lexical 
representation
for the concept of whole numbers and floating point numbers.  If that 
lexical representation requires further information, I would suggest adding
attributes.  In the end, I am saying that I believe that numeric datatypes
are absolutely necessary.

If I am totally off base on my thoughts here, I would ask the document
editor to please make more of the issues and problems known in the next
version of the draft (if there is another draft version).

Looking forward to the next working draft,

 - Mike
-----------------------------------------------
Michael C. Daconta
Author of Java 2 and JavaScript for C/C++ Programmers
Author of C++ Pointers and Dynamic Memory Management
Sun Certified Java Programmer and Developer
http://www.gosynergy.com

Received on Friday, 21 May 1999 01:09:20 UTC