Re: PSVI

[Rick Jelliffe <ricko@allette.com.au>]

From: Henry S. Thompson <ht@cogsci.ed.ac.uk>
>Too late to use this as a banner to rally the troops behind, we've
>already done it :-)!  This is _precisely_ how the XML Schema
>architecture is specified:
>
> 1) Start with an XML+Namespaces compliant parser output (need _not_
>    be validating, but _may_ be);
> 2) Specify what schemas are to be involved _outside_ the instance;
> 3) _Augment_ the Infoset.
 
Perhaps the last could be split too--
 3) determine the type of the data and augment the infoset with type info
 4) validate the data and augment the infoset with outcomes

Cheers
Rick Jelliffe