Re: [xml-dev] parsing markup with Perl

["Simon St.Laurent" <simonstl@simonstl.com>]

People who create and maintain cars rarely need to be metallurgists.

However, they rarely argue that metal parts are not metal, and typically 
wonder how reliable the results will be in plastic.

I get that the immanence of text gets in the way of the transcendence 
you'd like to encourage through models.  It's a tough theological style 
problem.

Please recognize that the nature of bytes matters deeply, however, 
before you try to build on top of and away from those bytes.

Thanks,
Simon

On 2/10/14 1:10 AM, Hans-Juergen Rennau wrote:
> A few facts.
> 1) It is impossible to write an XQuery which requires that input XML is
> provided as text.
> 2) It is impossible to write an XQuery which is affected by any aspect
> or detail of the text representation of XML input, should there be such
> a representation.
> 3) It is impossible to write an XQuery which requires that output XML is
> serialized into text.
> 4) Same points for XPath and XSLT stylesheets.
> 5) It is impossible to write an XSD which requires that the described
> documents are represented by text.
> 6) It is impossible to write an XSD which constrains in any way any
> details of a possible text representation.
> Facing all these impossibilities for a year or ten, on a daily basis,
> one may lose interest in the text representation of XML and concentrate
> on the information content.
> Remains the question if the perspective of a person dealing with
> (modeling, processing, producing) XML primarily via XSD, XPath, XQuery,
> XSLT is of any relevance when speaking about XML.
> Hans-Juergen Rennau
>
> *Von:* Simon St.Laurent <simonstl@simonstl.com>
> *Betreff:* Re: [xml-dev] parsing markup with Perl
> "
> Deepest wrongness may also be profound.
> "Markup is not just text" is true.  Markup is not text" is short and wrong.
> "
>
> .
>
> "
> None of which [debate about XML] even begin to happen without the text
> foundation.  Sorry.
>    Please come back to earth.
> "
>
> Thanks,
> --
> Simon St.Laurent
> http://simonstl.com/

--
Simon St.Laurent
http://simonstl.com/