Re: [docbook-tc] Are the Schematron assertions normative in 5.0?

[Norman Walsh <ndw@nwalsh.com>]

/ Jirka Kosek <jirka@kosek.cz> was heard to say:
| Norman Walsh wrote:
|
|> I could go either way, but we should decide.
|
| AFAIK Schematron should be formally standardized by ISO soon (well,
| yes *soon* in ISO's point of view could be one year). ISO Schematron
| uses different namespace than Rick's original version. It might be
| better to make Schematron assertions normative after Schematron gets
| ISO standard status. But I don't have very strong opinion about
| this.

Good to know about ISO.

With DTDs, it was always understood that the DTD couldn't capture all
the semantics of a language, that's why the documentation is normative.
In RELAX NG, we can get a lot closer. The Schematron assertions
actually test for things that we say are language constraints (a glossterm
linkend must point to a glossentry; the top-level element must have 
a version attribute, etc.).

The only downside I see to making the assertions normative is that there
aren't very many validators that actually perform those checks. So you validate
with jing using the schema that has the annotations, but jing doesn't test
the annotations, so do you get the illusion that your document is more valid
than it really is? And is that a problem?


                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com>      | Puritanism--The haunting fear that
http://www.oasis-open.org/docbook/ | someone, somewhere may be
Chair, DocBook Technical Committee | happy.--H.L. Mencken

PGP signature