Re: [xml-dev] Who can implement W3C XML Schema ?

[Eric van der Vlist <vdv@dyomedea.com>]

Hi Rick,

Rick Jelliffe wrote:

> Surely the third <any> element in the instance is invalid?  It has three tokens in it.
> 
> 
>>I may also still have missed something in the rec and my example may be 
>>invalid, but the point would remain that if you take 4 W3C XML Schema 
>>processors on a simple test case, you have very often 4 different answers.
>>
> 
> Two of the products Eric are using are betas (MSXML and Xerces) and 
> the XML Spy is not the most recent (I don't know if it fixes the problem,
> and I don't know if the Turbo XML is the most recent.) 


I am using the latest versions available for each of the products 
including XML Spy 4.1. The distinction between beta and production seems 
very blur and the beta I have tested are often more "right" than the others!

> All it shows is that
> developers are leaving this till last or too late, rather than that there is some 
> crisis of complexity.


I hope you are proven right.

I think a clear message should be sent to the developpers rushing to use 
W3C XML Schema thinking that the technology is stable and building 
architecture on top of schemas which they have only tested against some 
few existing implementations: their schemas are likely to be bogus, they 
will probably not validate against other schema processors and will have 
to be updated when the tools will become more conform to the rec.

My personal conclusion is that they'd better stay away from W3C XML 
Schema until the tools are stabalized.

This is far worse than what we've had to endure with XML technologies 
till now (conform XML 1.0 parsers, Namespaces aware parsers and XSLT 
processors have become stable in much less time) and I don't think the 
users are expecting this.


> 
> As for the complexity in the area of datatypes, I have not heard anyone say that
> the facet-based approach is not the most elegant way to treat the problem.


This is just an example. I have similar experience with all the features 
I have tested up to now, in both schema part 1 and part 2.


> What is the alternative:  Only simple types? No specification of 
> type restrictions on an instance element?  Using little languages
> rather than facets (Schematron's approach, b.t.w., and powerful but
> messy)?  


I don't know. I was just trying to ring a bell!

Thanks for your answer.

Eric


> I certainly could live without being able to restrict types
> on an instance, but being able to decorate an instance with
> the same type as in the schema allows architectural processing
> (like Francis Norton's recent TypeTagger tool) which is
> certainly useful.
> 
> Cheers
> Rick Jelliffe

-- 
Rendez-vous à Paris pour le Forum XML.
                    http://www.technoforum.fr/Pages/forumXML01/index.html
------------------------------------------------------------------------
Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
http://xsltunit.org      http://4xt.org           http://examplotron.org
------------------------------------------------------------------------