[dsdl-discuss] Interesting feedback from .NET implementor of Schematron

Daniel Cazzulino, who implemented Schematron in .NET, has given some
interesting feedback on Part 3. He likes the direction, especially
with let. But he does not think the new features are elegant enough,
and there are some minor changes that would help.

<sch:let>
----------

In particular he wants to have let expressions available on
and scoped to rules, patterns, phases and the schema.
(Their context would be the document element).

What struck me about this idea is that
  1) It would provide a different approach to the problem we found
in my suggested draft Part 8. It strikes me that if part 8 has the
problem, other bindings could also have it.

It also makes it neater to have external parameters (the highest scope)
again, which I had dropped reluctantly because I had not had
a specific request for it.

I think this is worth doing, and will not cause any violance to
existing semantics.

Daniel is not keen that there are no formal parameters

<sch:include>
-------------

He also suggests dropping abstract rules in favour of XInclude.
Also dropping the sch:include in favour of XInclude. That way
one component in a schema can reference another to reduce redclaration.
He prefers @inherits to @is-a.

I am not convinved by these, but open to last minute suggestions.

@language
---------

He thinks these should be pairs, and perhaps lists that allow
you to say "this will work with both XPath1 and XPath2". I don't
think so, in the absense of any new info. Daniel also wants to
makes sure that EXLST is included in the reserved lists.

So my intent is that the new version will have this neater
<let> idea.

Cheers
Rick Jelliffe

--
DSDL members discussion list
To unsubscribe, please send a message with the
command  "unsubscribe" to dsdl-discuss-request@dsdl.org
(mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)