Re: LC-185 comments from XML Core WG

> It is not clear to me whether the creation of the PSV is seen as a process
> that augments the existing infoset or as a process of creating a completely
> new infoset.  

It is described as augmenting an existing infoset.  In most cases it
adds new properties.  The only case I can see where an existing
property is changed is that an element's [attributes] property may
have additional members for schema-defaulted attributes.  (However, an
element with a default value - necessarily a string - does not have
character children added.)

A new property is used for the schema-normalized value (and this seems
to mean that a attribute defaulted by a schema won't have a standard
[normalized value] property - which might be considered odd).

Another oddity of defaulted attributes and elements is that they have
a [specified] property with the value "schema", whereas in the pre-schema
infoset this property is boolean.  Furthermore undefaulted elements
have a [specified]  value of "instance".  (Note that the items with
these properties either didn't exist or didn't have [specified]
properties in the pre-schema infoset.)

-- Richard