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

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   When validation is appropriate

[ Lists Home | Date Index | Thread Index ]

Just to follow up on the thread last week about validation, there is an excellent 
paper online from the Extreme Markup conference "A Decade of DTDs and 
SGML in Scholarly Publishing: What Have We learned?" at 
http://www.idealliance.org/papers/extreme02/html/2002/Rosenblum01/EML2002Rosenblum01.html

It puts validation more in the context of organization behaviour and control:

"As many publishers have discovered, no matter how complete the documentation, 
strict conformance to the DTD designer's interpretation can only be ensured with 
software-based quality control tools. However, SGML producers will not use quality 
control tools and fix reported problems without negative feedback loops. Publishers 
must actively monitor and enforce quality standards, even after providing tools to SGML 
producers."

The question that springs to my mind is "what kinds of constraints are these
quality control tools implementing?"  -- are they the kinds of things that
Schematron is good at, or the kind of things that WXS is good at
(or both or neither)?

I have repeatedly raised the point over the last few years that because there
are no systematic studies into this kind of thing, we really must take all
schema standards with a pinch of salt to a certain extent: they all provide 
consistent and technically excellent tools, but there is no reason to expect 
that they meet the kinds of tricky requirements that people face in real life.  

(Of course, I think the rational repsonse to this is that we need 
a general framework for validation within which little languages
can be plugged in, to allow innovation and natural selection. Hence
Schematron, ISO DSDL and XAR.  We need to provide a path by 
which individual  proprietary solutions can, as they prove themselves, 
migrate to the public domain.)

So I would be very interested in hearing (directly or through XML-DEV)
from XML-DEVers (especially lurkers) what kinds of non-DTD 
constraint-checking they have had to implement, whether to help
data entry, QC, QA or unit testing.

Cheers
Rick Jelliffe




 

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

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