[
Lists Home |
Date Index |
Thread Index
]
I totally agree. That's what I meant by "looking for a proper design
pattern". XML does lack a proper, extensible processing model.
So in the emergency, and without guidelines, people immediatly dropped the
extensibility part when writing their code. Well, at least I did so :).
I mean, I can handle schema evolution. There is nothing that can't be solved
by (re)writing code... But I'd rather write some code that would be less
expensive to maintain when the schema changes, or code that can still
operate in a coherent way when presented with a document following the new
schema. Note that rejecting a document for fear that something wrong could
be performed is coherent, too (think about
http://www.ge.com/services/nuclear/plantcontrol.wsdl).
I'm wondering if an extensible processing model can be built a posteriori on
top of a data model which miss typing (or schema, or whatever anyone wants
to call it) information, both at the document and element level. Unless
element names alone (without their context) are sufficient to process XML
documents in a meaningful and useful way, of course, but I don't think they
are.
Regards,
Nicolas
>-----Message d'origine-----
>De : Eric van der Vlist [mailto:vdv@dyomedea.com]
>Envoye : mardi 5 mars 2002 15:00
>A : Nicolas LEHUEN
>Cc : 'Simon St.Laurent'; 'xml-dev@lists.xml.org'
>Objet : Re: [xml-dev] XML doesn't deserve its "X".
>
>
>Nicolas LEHUEN wrote:
>
>> That's because extensibility is not something that can be achieved by
>> looking at the data only. I'm sorry to ressurect my old rant
>"meaning is not
>> in data alone, it's both in data AND code", but alas, it's
>still true here.
>
>I wouldn't say that, but rather that the minimum amount of processing
>model hasn't been defined after XML 1.0 and, to stick to my example,
>that Namespaces in XML has not been more directive.
>
>I don't want to bind full applications to XML documents (that would be
>the end of the "late binding" which is the power of XML) but I think
>that nobody has wanted to take the risk of defining any
>processing model
>for XML after XML 1.0 and that we trying to maintain the scafolding
>which as been built on top of XML 1.0 with the few tools
>defined by the
>original recommendation!
>
>Eric
>--
>See you in Seattle.
>
>http://knowledgetechnologies.net/
>---------------------------------------------------------------
>---------
>Eric van der Vlist http://xmlfr.org
>http://dyomedea.com
>http://xsltunit.org http://4xt.org
>
>http://examplotron.org
>---------------------------------------------------------------
>---------
>
|
