[xmlschemata] Re: Headaches w/ XDOTs

Just to add some more reasons for headaches...

Michael Fitzgerald gave the other day some pointers for related issues
on xml-dev ([2] is probably the most relevant to your work):

[1] http://www.thaiopensource.com/relaxng/design.html#section:10
[2] http://www.cs.ucla.edu/~mani/xmlresearch/datamodels/subtyping.pdf
[3] http://lists.oasis-open.org/archives/relax-ng/200206/msg00022.html

You could also have a look at Sun's MSV [4] which includes a Relax NG
engine validating WXS schemas read as Relax NG patterns. I see it as a
proof that WXS features (or at least 90% of them) including type
assignment can be built on top of Relax NG...

[4] http://wwws.sun.com/software/xml/developers/multischema/

I would like to work in this direction with my xvif sooner or later as
well. The idea I have (but I haven't tested nor even formalized it yet)
is to adapt the derivative algorithm so that after a derivation the
"emptied" pattern is kept and marked as emptied instead of being thrown
away. Keeping these emptied patterns is probably the only way to
preserve some meta data (or PSVI to use a name more famous).

Hope this helps.

Eric

-- 
See you in San Diego.
                               http://conferences.oreillynet.com/os2002/
------------------------------------------------------------------------
Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
(W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
------------------------------------------------------------------------