- From: Norman Walsh <Norman.Walsh@Sun.COM>
- Date: Fri, 06 Jun 2003 11:17:32 -0400
- To: Jonathan Robie <jonathan.robie@datadirect-technologies.com>
- Cc: Jeni Tennison <jeni@jenitennison.com>, John Cowan <jcowan@reutershealth.com>, Mike Champion <mc@xegesis.org>, <public-qt-comments@w3.org>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
/ Jonathan Robie <jonathan.robie@datadirect-technologies.com> was heard to say:
| As a thought experiment, suppose you wanted to support a schema
| language that allowed just DTDs plus XML Schema simple types. If you
| write an XML processor for that language, you could have the processor
| create data model instances directly, or it could have some
| intermediate format, the map from the intermediate format to the data
| model instance.
It may be worth pointing out that we decided in Gaithersburg to make
Section 3.6 of the data model informative rather than normative. You
can build a data model any way you darn well please provided that you
don't voilate its constraints.
So, in fact, I don't think there's anything preventing an implementor
from making a processor that recognizes xsi:type and xsi:nil even in
the absence of a schema and builds a data model accordingly.
Be seeing you,
norm
- --
Norman.Walsh@Sun.COM | The shoe that fits one person pinches
XML Standards Architect | another; there is no recipe for living that
Web Tech. and Standards | suits all cases.--Jung
Sun Microsystems, Inc. |
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.7 <http://mailcrypt.sourceforge.net/>
iD8DBQE+4LCMOyltUcwYWjsRAkhSAJ4g9ggnJZp2jdndXUwjBvUsP+2AXgCgmIMC
kDTho/7DO7cbxTGhA5/5KVc=
=EFs/
-----END PGP SIGNATURE-----
Received on Friday, 6 June 2003 11:18:21 UTC