[dsdl-discuss] Reformulation of MNS for DSDL Part 4

I think MNS has several defects, but I like the general approach.

I have made a reformulation of it, called "Namespace Switchboard"
for want of a better title. I attach it here, and it is also at
http://www.topologi.com/resources/NamespaceSwitchboard.html

It is more powerul in some regards, and less powerful in others.
I think it is cleaner and simpler.

The defects I see in MNS (not all of which are addressed in Namespace
Switchboard, to reduce static) include:

* too complex: modes in particular look clunky
* "lax", "cover" etc conflate various concepts that would be better
  exposed orthoganally, to cover more cases
* focus of attention is wrong: rather than couching rules in
  terms of "validations" I think it is clearer to couch it in terms
   of "namespaces" which have various bindings, validation policies
   and scoping
* seems incompatible with WXS: in particular, in WXS you
   should only have one governing schema for a namespace,
   not different ones in different contexts
* no support for namespace versioning
* no support for dynamic schemas (i.e. phases)
* even in just passing parameters to a schema (which is
  an absolute requirement)
* it seems that MNS is based on overcoming specific RELAX NG shortfalls
rather than providing a more generally useful tool, in that it does not
seem to provide much that is useful to Schematron. (Maybe that is OK.)
* MNS as specified provides a layer which blocks off many of the
things I proposed for Schemachine, which I thought had been loosely
accepted by the WG at the last meeting.

Schematron was developed because of WXS's shortfalls in various
areas: using expressions in multiple namespaces (to cope with
namespace versioning) and providing phases (to cope with dynamic
schemas.) It seems that MNS just perpetuates WXS's mindset in
this regard.

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)