OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   RE: [xml-dev] W3C Schema: Resistance is Futile, says Don Box

[ Lists Home | Date Index | Thread Index ]

Francis Norton wrote:
>XSD types are great for Web Service (program-to-program) use cases in
>opinion - you get the choice of specifying language-level types (eg 
>xs:byte) where appropriate, or application-specific (eg xs:long, with 
>maxExclusive and minInclusive) where more appropriate. And for 
>human-to-human use cases, you can just leave them all as strings if 
>that's the functionality you want. So what's the problem?

Hmm when we start talking XSD types with Web Services I get really wary,
I'm getting burned out here with my own personal dislikes of XSDL I
certainly don't want to get started on a thread where everyone starts
arguing WSDL. *shudder* 
I think a lot of my problem has to do with what I consider my personal
responsibility as an "xml dialect maker", and a lot of this personal
responsibility is tied up with personal views about development etc.
that I could not clearly enunciate here, maybe not in many years of

But I would say that the personal responsibility I feel comes partly
from the marketing of Xml Schema founded dialects where someone talks
all about how data will be validated etc. and thus you are never going
to end up with bad data; obviously never bad data is big promise but
someone having made it and people naively expecting it I feel it
incumbent upon me to do my best to provide it. And this is what bugs me
with this huge specification is that for all the talk of validation and
datatypes rolling your own datatypes is so hard once you get past the
trivial examples of restricting numerical ranges, or string content to a
sequence of xsd:enumeration values, that people just leave the hard
stuff as a string. And that leaves me obsessively trying to restrict
that string to something passable with a long-ass regular expression for
days on end.  :(

And anyhow I guess now I'm starting to repeat myself way too much cause
I made some of these observations when I first entered this thread, and
probably the only conclusion is to, as someone suggested some days back,
get out the good word about RELAX-NG about which I have a sad anecdote
to relate:
Recently my boss received an email, passed on from a business partner
who was passing it on from a friend of his in New Zealand who wanted to
switch all their old dtds to Xml Schema and wanted some pointers. I
pointed to the beginning of the Xml Schema considered harmful thread,
pointed to a bunch of other Schema resources, went into depth as to what
the benefits of RELAX-NG and Schematron were to my mind as opposed to
XSDL and so forth. Well for reasons that I think are partly political I
believe most of my RELAX-NG and Schematron stuff was cut from the
finalized communication, in the interest supposedly of not overwhelming
the poor guy trying to come to grips with Xml Schema, who I suppose will
now be overwhelmed by attempting to come to grips with Xml Schema. 


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS