[
Lists Home |
Date Index |
Thread Index
]
I think you are right that value space vs lexical
space is at the heart of some of the more contentious
threads.
But that is exactly where this discussion seems to run
off the rails. One group believes that syntax and structure
are the focus and limit of XML and that once past that boundary,
one is discussing an application language. Another group
believes it useful to define a value space, aka, primitive
types, that any XML application language can use. Another
group believes it best to define a means to add pluggable
types and not to define any set which all applications
must support.
Data is portable. Systems interoperate. How
much interoperability is wanted or affordable? What
options do we lose if we strengthen the coupling to
a primitive type set? I don't think I could work
with relational systems without value types, and we
do seem to share them among that application type
sensibly; but is that the only application type one
would have to share them with?
Today, #FIXED looks pretty awesome for its simplicity,
but of course, one has to read the schema/DTD to "know" it.
xsi:type looks pretty good, but it is still just a way
to annotate in accordance with a data dictionary. Bytes
must change state and that requires a value space.
I don't see any way out of this except to define
a means to plug in types with a non-normative primitive
set for free.
len
From: Don Box [mailto:dbox@microsoft.com]
In case it wasn't obvious, that sentence should have read "What good are
datatypes if one CANNOT work in terms of the value space..."
|
