[
Lists Home |
Date Index |
Thread Index
]
- To: 'Mark Feblowitz' <mfeblowitz@frictionless.com>, "'ht@cogsci.ed.ac.uk'" <ht@cogsci.ed.ac.uk>, Dare Obasanjo <kpako@yahoo.com>
- Subject: RE: [xml-dev] Schema Namespace name, schemaLocation, and Schema V ersioning
- From: "Bullard, Claude L (Len)" <clbullar@ingr.com>
- Date: Thu, 18 Jul 2002 14:11:17 -0500
- Cc: David Carlisle <davidc@nag.co.uk>, jeni@jenitennison.com, xml-dev@lists.xml.org, "Duane Krahn (E-mail)" <duane.krahn@irista.com>, "Satish Ramanathan (E-mail)" <Satish.Ramanathan@mro.com>, "Andrew Warren (E-mail)" <awarren@openapplications.org>, "Kurt A Kanaskie (Kurt) (E-mail)" <kkanaskie@lucent.com>, "Michael Rowell (E-mail)" <mrowell@openapplications.org>
Arguably, none, but that is just one opinion.
Versioning is a rat's maze and troubles hypertext theorists and
practicioners because it points out why namespaces are so
difficult to swallow: the preferred reading must be established
a priori or it disrupts "blind interoperability". In short,
there is no such thing as guaranteed, non-negotiated, blind
interoperability. There is only a point in time or process
at which identity is assigned and meaning is agreed upon,
or simply, when to put on the blinders and forget about it.
This is sometimes called "symbol grounding", or the creation
of the control dictionary that will govern interpretation.
There is no way out of this trap and how much it costs depends
on the precision required across the number of agreements
established. That is PRECISELY why XML is a syntax specification
and nothing more. It is also why namespaces do not belong
in the core but should be the first thing to the right of the
+ (XML 1.0 + Namespaces + ...) where the additive statement is
the definition of the *system in effect* at the time of the
transaction (which layers are active and when and over what).
If for example, you take the counter opinion that namespaces
should in some sense, establish versioning, you can break pre-existing
processes if you do that in the namespace URI itself.
If you take the position that namespaces are really
not simply syntax disambiguators but are controls over the
interpretation of the element/attribute semantic, then you want
something like RDDL and the namespace value is an index to the
record of authority of interpretation, ie, the record to which
all members of an interpretive community assent (and remember,
any communicating entity can be a member of multiple interpretive
communties, and therefore, have to be individually responsible
for resolving conflicts of interpretation).
That said, you still have to pick a way to choose among what
the RDDL document options might provide. In this case, the
use of the namespace and a version attribute or some other
option criteria should be considered.
len
From: Mark Feblowitz [mailto:mfeblowitz@frictionless.com]
What role, if any, should the namespace name play in this scenario?
|