[
Lists Home |
Date Index |
Thread Index
]
Hello All,
I've been working on DITA (Darwin Information Typing Architecture) XML
Schemas internally at IBM for a number years and with DITA TC since it
inception. Although, DTDs are still very much the norm in the
documentation world, DITA has had an XML Schema version for a few years.
The XML editor Syntext Serna has made use of those schemas for couple of
releases.
I've had my share of frustrations trying to implement the DITA
architecture using XML Schemas. Most issues will be addressed in the
XML Schema 1.1 specification.
The main issue that remains unsolved from an architectural standpoint is
the inability to rename/alias elements in a specific content model while
making use of the inheritance model. Basically, being able to use the
something like the substitutioGroup mechanism in a local context. With
respect to inheritance, DITA cannot make full use of the XML Schema
inheritance model. DITA is still fully represented in XML Schema
without using the inheritance model. It means that developers who wish
to create new DITA specializations have to manually validate its
inheritance model which is prone to error.
Out-of-the-box schema support for XML Schema in a publishing pipeline is
another reason that in the documentation world use of schemas is very
slow in its adoption. Hopefully, that will change when freely available
applications that support XSLT 2.0 becomes more mainstream. Web browsers
still do not support viewing/transforming schema validated XML document
because they still uses XSLT 1.0 processors underneath the covers.
User's must use CSS stylesheets to render their schema validated documents.
Since the both DTDs and XML Schemas are normative for the DITA
specification we are currently working on supporting both DTDs and XML
Schemas validated documents in the DITA-OT (DITA Open Toolkit at
SourceForge) publishing pipeline. It will work with XSLT 1.0
processors too.
Kind regards,
Eric
|
