Re: [ebsoa] SOA and Shared Semantics

["David RR Webber" <david@drrw.info>]

Duane,

I like the way BPSS approaches this.  Essentially the BPSS
model is exactly equivalent to a Prolog instance - with
assertions, predicates, clauses and facts - so it can
be proven to be deterministic and complete.

These design priniciples will certainly help us as we go
forward here.

Cheers, DW.

----- Original Message ----- 
From: "Duane Nickull" <dnickull@adobe.com>
To: "Chiusano Joseph" <chiusano_joseph@bah.com>
Cc: "ebSOA" <ebsoa@lists.oasis-open.org>
Sent: Wednesday, July 07, 2004 11:41 AM
Subject: Re: [ebsoa] SOA and Shared Semantics


> Patterns and blueprints!
>
> We may decide to define how business payload metadata is built by using
> a context + assembly + core component type methodology (not mentioning
> them by name).  There are patterns and blueprints we can use to describe
> this.
>
> One caveat:
>
> How can we normatively constrain a design time activity that clearly has
> no test at runtime?  An XML parser does not care that you made your XML
> instance by modelling or random.  If someone does not use core
> components, context and modelling, why would we (as either an
> application or human) care.  That is the real question.
>
> Duane
>
> Chiusano Joseph wrote:
>
> >Any thoughts on the question below?
> >
> >Joe
> >
> >Chiusano Joseph wrote:
> >
> >
> >>Thanks David.
> >>
> >>How about in more general terms, stepping aside for a moment from
> >>specific emerging initiatives - I'm thinking more generally, such as:
> >>what role should ontologies play? Why are they important in this
> >>context? How about metadata registries? etc.
> >>
> >>Thanks,
> >>Joe
> >>
> >>David RR Webber wrote:
> >>
> >>
> >>>Joe,
> >>>
> >>>At this juncture I would like to point out
> >>>that the BCM/EPR team is preparing a set of
> >>>requirements for submission to ebSOA.
> >>>
> >>>This stems from the work and meetings in
> >>>Norway over the past two weeks.
> >>>
> >>>And indeed semantic information is a
> >>>significant part of this - not only
> >>>for common understanding - but also
> >>>when used to create ontologies and
> >>>topic maps for discovery and reuse
> >>>of components.
> >>>
> >>>The essence of the BCM/EPR needs can
> >>>be distilled in part from the PPT
> >>>presented in Norway - available
> >>>from http://www.eprforum.org
> >>>
> >>>However - we are working on formalizing
> >>>the exact mechanisms and base requirements
> >>>so that an ebSOA based set of components
> >>>can support an BCM/EPR deployment.
> >>>
> >>>It's obviously good to be able to focus
> >>>on a real deployment model - to be able
> >>>to hone exact requirements, and
> >>>combination of requirements - rather than
> >>>things being noted in a piecemeal fashion.
> >>>
> >>>Hope this helps.
> >>>
> >>>Thanks, DW
> >>>
> >>>Quoting Chiusano Joseph <chiusano_joseph@bah.com>:
> >>>
> >>>
> >>>
> >>>>Is this a topic that we're planning on including in our spec? (I think
> >>>>we should) We can emphasize the role of ontologies in helping ensure
> >>>>"shared semantic understanding" between participants in an SOA.
> >>>>
> >>>>Joe
> >>>>--
> >>>>Kind Regards,
> >>>>Joseph Chiusano
> >>>>Associate
> >>>>Booz | Allen | Hamilton
> >>>>
> >>>>
> >>>>
> >>>http://drrw.net
> >>>
> >>>
> >>--
> >>Kind Regards,
> >>Joseph Chiusano
> >>Associate
> >>Booz | Allen | Hamilton
> >>
> >>
> >
> >
> >
>
> -- 
> Senior Standards Strategist
> Adobe Systems, Inc.
> http://www.adobe.com
>
>
>
>