Fwd: [xml-dev] status of OWL amidst other standards

[adasal <adam.saltiel@gmail.com>]

Apologies to the list, this email was supposed to have been to xml-dev@lists.xml.org and cc to Seth.
We have gone off line with issues not relevant to the list, I think.
But maybe others have thoughts about the following?
Are you familiar with the work at UMBC eBiquity Project: Context Broker Architecture (CoBrA
) ?

It seems that Harry Chen has moved from being a Doctoral researcher to
a little specialist start up, I imagine working in this area?

But I am interested in views of others about such developments, as I
tend to follwo what I can find on google and what is open source,
sometimes there are other (more complete) solutions around.

Solutions using OWL I would expect to either demonstrate thier ability
by showing fairly completely how it would be used in a specific area
or, as a tool, seemlesly slide the user along a path to arrive at a
usage.

The space seems to be surprisingly empty of solutions, but perhaps I just don't know?

---------- Forwarded message ----------
From: adasal <adam.saltiel@gmail.com>
Date: 06-Dec-2005 16:37
Subject: Re: [xml-dev] status of OWL amidst other standards
To: Seth Proctor <seth.proctor@sun.com>

Thank you both for your replies.
I think there is a slight misunderstanding. What I was asking was how
much the two approaches would be alternatives to each other rather than
specifically how to combine them.
It does seem, from your replies, that there is a large degree of
overlap and that XACML would provide a lot, but not all, of what might
be accomplished with OWL. This was very much a hypothetical question
about this sort of overlap, although it is related to a real world
solution that we *may* have to provide in the future. It is too early
to start assessing from other points of view, e.g. how much more
advanced XACML solutions might be compared to OWL. Of course that is
part of the answer to my question, I guess. OWL is broad and generic.
It could be used in this context. But, meanwhile specific solutions for
this context are being worked on in this case in XAMCL.
About the server, I can't give the pointer as it was an operation that
failed to get venture capital and so disappeared, I imagine. Just one
of those things!
Adam

On 06/12/05, Seth Proctor < seth.proctor@sun.com> wrote:

Hi Adam. Echoing Joseph's comments, XACML would be a good approach for your
problem. I don't know of anyone who has worked specifically with OWL and
XACML, but there has been some work done with ontologies and access systems.
In the XACML world, attributes are arbitrary identifiers, and can always
be resolved based on your context, so it would be pretty easy to write
support for exactly what you're describing. In the system I maintain [1] this
would be done via an AttributeFinderModule that is plugged into the XACML
PDP. I would be happy to walk you through how this works.

One question:

On Tue, Dec 06, 2005 at 02:14:26PM +0000, adasal wrote:
> [...]
> But I have a colleague who, intreaguingly, has told me that he worked on the
> first implementation of an XACML server.

Could you pass along a pointer? There are many systems that implement XACML,
and many servers that folks have worked on. I think I'm familiar with most
of them, but not what you've described...


seth


[1] http://sunxacml.sourceforge.net ... download the 1.2 release package and
    you'll find several examples of XACML policies, one of which looks
    similar to your use case.