[topicmapmail] UDDI, semantics and XTM
Kal Ahmed
kal@techquila.com
Mon, 09 Feb 2004 21:19:08 +0000
Hi Max,
I think that Conal has already said all that I was going to add on this
matter. The important thing is that you do not need to duplicate the
information you hold about the entities - just the identity. Essentially
you are creating a set of binding points in your topic map to which you
can bind UDDI entities. If you think of a "UDDI domain" and a "Topic Map
domain", then these binding point topics are the way to cross the chasm
from one domain to the other (being rooted at the Topic Map domain
side). You would need to do the same with RDF (rdf:about). There is no
need for the binding points to copy anything more than the identity from
the UDDI domain into the Topic Map domain unless you really want to.
Having said all that, there is no reason why an application could not
present a topic map interface to a UDDI repository (mapping on the fly)
- or indeed why a UDDI repository could not use a topic map as its
store. After all UDDI only constrains the interchange syntax (and as I
see in UDDI v2 at least, the lengths of certain values) - UDDI does not
constrain the implementation to be an RDBMS or an XML database.
I have read through some of the UDDI v2 specs and AFAICS the basic
structure of the v2 registry is pretty straightforward and could easily
be represented by a topic map.
Hmmm - topic map enabled web services anyone ? :-)
Cheers,
Kal
On Mon, 2004-02-09 at 09:24, Max Voskob wrote:
> Kal,
>
> What about objects that are instances of those entities?
>
> ENtity "BUsiness"
> Object: Microsoft
> Object: "IBM"
>
> Are you suggesting that I create a topic for IBM and a topic for Microsoft?
> In this case, I create a replica of my instance of UDDI registry as a Topic
> Map. In it not the intention.
>
> If we follow this scenario, every object will have to have its own topic. It
> will simply duplicate the workload.
> The taxonomy may be not under control of the UDDI node, so it will be
> impossible to create a new topic or assocication on top of what was provided
> by the taxonomy owner.
>
> Please, correct me if I'm wrong.
>
> Cheers,
> Max
>
>
>
> ----- Original Message -----
> From: "Kal Ahmed" <kal@techquila.com>
> To: "Max Voskob" <vobla73@paradise.net.nz>; <topicmapmail@infoloom.com>
> Sent: Monday, February 09, 2004 1:07 AM
> Subject: Re: [topicmapmail] UDDI, semantics and XTM
>
>
> > Max,
> >
> > Your UDDI entities (services, endpoints etc) are all topics from the
> > topic map point of view - as long as you want to say something about
> > them (like their relationship to one another), then you should make a
> > topic in the topic map that stands for the UDDI entity.
> >
> >
> > On Fri, 2004-02-06 at 21:58, Max Voskob wrote:
> > > Con,
> > >
> > > One issue is that it is not clear how we model the link between a
> taxonomy
> > > and UDDI entities.
> > >
> >
> > Can you not use the identifiers that UDDI assigns to each entity ? If
> > you are using URNs for entity identifiers, then that would seem to be a
> > subject address.
> >
> > > Let's say that I have an entity A and taxonomy T.
> > > If it is a matter of linking A to one of the topics from T then one URI
> > > would identify the taxonomy and another the topic.
> > > What if I want to show that entity A is in a relationship with entities
> B, C
> > > and D?
> >
> > If you have topics standing in for entities A, B, C and D, then its a
> > simple association between those topics.
> >
> > > The first thing that comes to my mind is that I create an association,
> but
> > > how do I link the entities to roles and association memebers?
> > > Assume that I create a topic for each entity. In this case I begin to
> point
> > > from the taxonomy to the entities which is not acceptable - the taxonomy
> > > becomes tightly coupled with a particular instance of UDDI server and
> its
> > > entities.
> > >
> >
> > It depends on what you mean by "pointing" - there is no addressing being
> > done here, just a use of common identifiers. Also, remember that the
> > taxonomy part and the entity description part can be separate topic
> > maps. Equally, any entity relationships can be yet another topic map.
> >
> > > How can we use a taxonomy to describe relationships between UDDI
> entities,
> > > for example?
> > >
> > > Not sure I expressed myself clearly.
> > >
> >
> > I'm not sure if I got to the bottom of your concerns, perhaps it would
> > help if you could give some concrete examples - perhaps point us at some
> > UDDI instances and explain what kind of information you would see being
> > in the realm of the topic map (if you have some examples of that too
> > that would be helpful).
> >
> > Cheers,
> >
> > Kal
> >
> > > Cheers,
> > > Max
> > >
> > >
> > >
> > > ----- Original Message -----
> > > From: "Conal Tuohy" <conalt@paradise.net.nz>
> > > To: <topicmapmail@infoloom.com>
> > > Sent: Friday, 6 February 2004 17:03
> > > Subject: RE: [topicmapmail] UDDI, semantics and XTM
> > >
> > >
> > > > At first glance I'd say that XTM looks eminently suited for your
> purpose.
> > > >
> > > > Was there some specific reason you thought it might not be very
> suitable?
> > > >
> > > > Con
> > > >
> > > > > -----Original Message-----
> > > > > From: topicmapmail-admin@infoloom.com
> > > > > [mailto:topicmapmail-admin@infoloom.com]On Behalf Of Max Voskob
> > > > > Sent: Friday, 6 February 2004 16:47
> > > > > To: topicmapmail@infoloom.com
> > > > > Subject: Re: [topicmapmail] UDDI, semantics and XTM
> > > > >
> > > > >
> > > > > Murray,
> > > > >
> > > > > Fair point!
> > > > >
> > > > > I'm not saying that you have to do the work for me. It is
> > > > > just that I'm in
> > > > > my square and might miss something important.
> > > > >
> > > > > I thought XTM community might be interested.
> > > > >
> > > > > Cheers,
> > > > > Max
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----- Original Message -----
> > > > > From: "Murray Altheim" <m.altheim@open.ac.uk>
> > > > > To: "Max Voskob" <vobla73@paradise.net.nz>
> > > > > Cc: <topicmapmail@infoloom.com>
> > > > > Sent: Friday, 6 February 2004 16:24
> > > > > Subject: Re: [topicmapmail] UDDI, semantics and XTM
> > > > >
> > > > >
> > > > > > Max Voskob wrote:
> > > > > > > Hi all,
> > > > > > >
> > > > > > > I would like to invite you review these UDDI documents
> > > > > and comment on
> > > > > the
> > > > > > > suitability of XTM for our requirements.
> > > > > > >
> > > > > > > At this stage, it looks like XTM is not very suitable for
> > > > > we are trying
> > > > > to
> > > > > > > achieve, but it is my only opinion and I can be easily wrong.
> > > > > > >
> > > > > > > I would very appreciate your feedback posted to this list or to
> me
> > > > > directly.
> > > > > > >
> > > > > > > The documents can be found here:
> > > > > > >
> > > > > > >
> > > > > http://www.oasis-open.org/committees/download.php/5322/uddi-sp
> > > > ec-tc-taxonomies-discussion-20040205.doc
> > > > > >
> > > > > >
> > > >
> > >
> http://www.oasis-open.org/committees/download.php/5323/uddi-spec-tc-req011-1
> > > > 4-value-set-20040205.doc
> > > > > >
> > > > > >
> > > >
> > >
> http://www.oasis-open.org/committees/download.php/5337/uddi-spec-tc-req11-14
> > > > -semantics-20040205.doc
> > > > >
> > > > > Max,
> > > > >
> > > > > Rather than making us do all the work for you, could you at least
> > > > > summarize the Microsoft Word documents you've posted? And why you
> > > > > think XTM is not very suitable despite being still enough a
> > > > > candidate that you're asking the Topic Map community for input?
> > > > > What about XTM makes you think it is a candidate? And what about
> > > > > it makes you think it's not suitable?
> > > > >
> > > > > Murray
> > > > >
> > > > >
> ......................................................................
> > > > > Murray Altheim
> http://kmi.open.ac.uk/people/murray/
> > > > > Knowledge Media Institute
> > > > > The Open University, Milton Keynes, Bucks, MK7 6AA, UK
> .
> > > > >
> > > > > LEO (July 23-Aug. 22): After English, astrology is my second
> > > > > language. Like a language, it's both logical and messy; it's
> > > > > useful in making sense of the world, yet full of crazy-making
> > > > > ambiguities. At its best, astrology is a playful study of the
> > > > > metaphorical link between the human psyche and the sun, moon,
> > > > > and planets. It's not a science. It's an elegant system of
> > > > > symbols, an art form with a special capacity to feed the soul
> > > > > and educate the imagination. When regarded as a precise method
> > > > > for predicting the future or when used to pander to the ego's
> > > > > obsessions, it becomes a deserving target for satire. So there
> > > > > you have it, Leo. I've clarified the essential views that
> > > > > underlie all I do in this horoscope column, and which therefore
> > > > > color the relationship between you and me. Now I challenge you
> > > > > to do what I just did: Get together with the people you care
> > > > > about and articulate the fundamental assumptions that form the
> > > > > basis of your connection.
> > > > > -- Rob Brezny's horoscope for 4 Feb 2004
> > > > > http://www.freewillastrology.com/
> > > > >
> > > > >
> > > >
> > > > _______________________________________________
> > > > topicmapmail mailing list
> > > > topicmapmail@infoloom.com
> > > > http://www.infoloom.com/mailman/listinfo/topicmapmail
> > > >
> > > > _______________________________________________
> > > > topicmapmail mailing list
> > > > topicmapmail@infoloom.com
> > > > http://www.infoloom.com/mailman/listinfo/topicmapmail
> > > >
> > >
> > > _______________________________________________
> > > topicmapmail mailing list
> > > topicmapmail@infoloom.com
> > > http://www.infoloom.com/mailman/listinfo/topicmapmail
> > --
> > Kal Ahmed <kal@techquila.com>
> > techquila
> >
> >
>
> _______________________________________________
> topicmapmail mailing list
> topicmapmail@infoloom.com
> http://www.infoloom.com/mailman/listinfo/topicmapmail
--
Kal Ahmed <kal@techquila.com>
techquila