Lorna and all,
I agree with your comments re interoperability if we impose
order. The only solutions that I know of to solve this problem are:
1. Create a schema that does not enforce ordering, but does not
use uniqueness constraints and Xpath and then use some other mechanism
to validate metadata instances after they have been validated to the XML
schema e.g. Schematron.
2. Use another schema language, such as RELAX NG, that allows us
to model the constraints of the standard without resorting to uniqueness
constraints. This has the problem that there is less support for RELAX
NG than W3C Schema in software toolkits i.e. Castor. We can still
validate metadata instances using a RELAX NG parser.
3. Use different techniques in a W3C schema to directly model
the constraints, though this will produce a schema that will be
extremely hard to read and pretty hard to write in the first place.
4. Accept that we have to follow the standard exactly and wait
until the software support matures to support the required features.
Does anybody know of other alterniatives?
Regards,
Ben
---------------------------------
Dr Ben Ryan
HLSI Software Development Manager
University of Huddersfield
Tel: 01484 473587
E-mail: [log in to unmask]
---------------------------------
-----Original Message-----
From: Lorna M. Campbell [mailto:[log in to unmask]]
Sent: 10 May 2004 20:56
To: [log in to unmask]
Subject: Re: Implementing IEEE LOM Metadata
Hi there,
Thanks for all your work here guys! It may have been a wet weekend but
surely there are more fun things to do than playing with schema?!
As usual I'm very cautious about commenting on issues relating to
bindings so please feel free to ignore my input if it's unhelpful or if
I'm way off track....
The issues of ordering is pretty contentious and I suspect that there is
no easy way to resolve this. I gather from Ben's comments and from
other implementers I've spoken to that it is simplest to create a
binding that enforces order. However this constrains the LOM conceptual
data schema which does not enforce ordering. By enforcing ordering you
are effectively creating a profile of the original data schema.
Having said all that, we clearly need a workable solution but I'm unsure
whether this is the right way to progress. Ben, Phil and others may
remember the first CETIS CodeBash where we discovered that differences
in the way systems handled metadata frequently resulted in lack of
interoperability. For example, systems that expected ordered metadata
sometimes could not import packages that did not enforce ordering
despite the fact that these packages were effectively "valid".
If we are going to create a basic binding that will be of maximum value
to all UK LOM Core implementers it needs to be as generic as possible.
I would feel a bit uneasy if we constrained developers who build
profiles based on the UK LOM Core by effectively requiring them to
enfore ordering.
I'd appreciate hearing the thoughts of those who know a lot more about
XML bindings than me!
All the best
Lorna
All,
I have looked at the existing IEEE XML Binding and have come to
the conclusion that if we want a really simple schema that is easy to
read and easy to process using castor/xerces etc that I would have to
impose order on the models or develop another way of checking the use of
elements that are only allowed zero or one times.
I would go for the approach of using order but this does mean
that if we validate externally produced LOM files they may fail because
the IEEE LOM does not specify ordering.
Does anybody have any opinions on this before I go ahead and
finalise the first draft?
Nothing like a wet Saturday for writing schemas :)
Regards,
Ben
---------------------------------
Dr Ben Ryan
--
Lorna M. Campbell
Assistant Director
Centre for Educational Technology Interoperability Standards (CETIS)
Centre for Academic Practice, University of Strathclyde
+44 (0)141 548 3072
http://www.cetis.ac.uk/
---
This transmission is confidential and may be legally privileged. If you receive it in error, please notify us immediately by e-mail and remove it from your system. If the content of this e-mail does not relate to the business of the University of Huddersfield, then we do not endorse it and will accept no liability.
|