RE: [xri] Designating a normative schema file

> > Drummond Reed wrote on 2009-08-18:
> > Second, RE the normative schema files discussed in (1) below: OASIS
> rules
> > are that we MUST declare only one schema file normative, i.e., if we
> publish
> > both RelaxNG and XSD, we must designate one as normative.
> 
> Scott Cantor replied:
> 
> That matches our expectation, but as I told Will today, the current schema
> isn't valid yet. I'm willing to produce a non-normative XSD, but I'll wait
> until there's a final version of the original.
> 
> I know from experience that it's hard to produce valid schemas. If there
> isn't the necessary expertise to do so here, then I strongly advise
> letting
> me produce a normative XSD, because I *will* get it right.

I completely agree that valid schema are not easy. We checked our RelaxNG
schemas for XRI Resolution 2.0 with several validators before calling them
done. I suggest we do the same here.

Once we have a validated RelaxNG schema, the XSD should (in this particular
case) be pretty straightforward.

=Drummond