Re: [xsl] James Clark on Schema

["J.Pietschmann" <j3322ptm@xxxxxxxx>]

Wendell Piez wrote:
> Ah! Worlds in Collision!
>
> On one side are those who see data types as essential features of any 
> information set.
>
> On the other, are those who see data types as useful (even important and 
> sometimes essential) for local processing, but for interchange, as Just 
> Another Label.

Well, I met the two views in this form: Given <foo>12345</foo>
1. It is a number! I can add 1 to it!
2. Don't care. I get it as a string from may API anyway.

A neat expression of this collision was the redefinition of
the german ZIP codes in, uh, 1992? The new codes had leading
zeros, thereby thouroughly confusing programs which was build
on the assumption that ZIP codes were and always will be
numbers, and that numbers don't have leading zeros if printed
on letter envelopes...

Interesting question: does the example above support the
view that data types are actually important or the view that
the only data type that really matters is "string" and
everything else is optional?

J.Pietschmann


XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list