Re: [cam] Context for SOAs

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

Joe,

Of course the SOA story is still emerging.  The best solution
for context is IMHO - the BCM Choice Points.  Nothing
else comes close.

What CAM is doing with context is more tactical - and
certainly the work that is happening right now with BPSS,
CPA, ebMS and CAM with context is setting new ground
for integration driven by context.

Again nothing on the WS* side matches this potential to
harmonize context across the delivery systems from
business process definition down to information exchanges
messaging and application integration.

So I see that the WS* world will again learn from ebXML
here on context.  This lack of purpose built approach for
context I think shows a huge gap in WS*.  Not that
WS* could not start using CAM - but its a bigger leap
for them from where they are today.  If you look at
BPEL especially - the notion of configuring and driving
WSDL and BPEL processes by context is a very
tough one to envision even how to start making that
work.  That's why I mentioned CORBA - because
in many ways the current WS* state of art reminds
me of CORBA components and that whole
deployment model.  Its loosely coupled but very
tightly bound together, and changing anything
requires in-depth knowledge of the inner workings.
Its fine if you just accept the blackboxes "as is" -
but ultimately that hub-n-spoke model of SOA is
an old one - and known to fail to scale.

The BCM has laid this out in great detail.  If WS*
were beginning to adopt the techniques that the
BCM teaches to ensure interoperable agile
information systems, then I'd be seeing it
differently.  That may begin to happen
eventually - but one feels that in the SOA
space this is all a hoped for rather than
guaranteed set of outcomes for WS*.

What you need to build for your clients
is a check list of these criteria to measure.
The BCM spec's provide these.  Then they
can start asking the WS* providers if they
deliver SOA using these.  If the answer is
"no quite" - then one has to determine
from there if the short falls are critical for
the solution envisioned.

Hope that helps.

Cheers, DW.

----- Original Message ----- 
From: "Chiusano Joseph" <chiusano_joseph@bah.com>
To: "CAM" <cam@lists.oasis-open.org>
Sent: Wednesday, February 04, 2004 9:58 PM
Subject: Re: [cam] Context for SOAs


> Thanks to all for your responses. I should have clarified that I wasn't
> looking for an ebXML-specific response, but rather an approach that can
> be leveraged for Web Services in general. Not all of my clients will be
> using ebXML (and potentially none will*).
>
> Joe
>
> *This is not a statement regarding my support of ebXML (which is high),
> but rather a reality that I have to work under.
>
> Chiusano Joseph wrote:
> >
> > I just attended an excellent session on SOAs given by IONA (Eric
> > Newcomer), Gartner, and [one of my company's competitors]. The idea of
> > context for services came to my mind.
> >
> > More specifically, let's say that we have multiple service consumers in
> > an SOA, but one particular consumer has a need to have its own
> > context-driven version of a message or operation in a WSDL document that
> > is being discovered. The OASIS WS-CAF TC does not cover this aspect of
> > context (I verified with Eric Newcomer) - is this something we're
> > considering covering in CAM?
> >
> > Thanks,
> > Joe
> >
> > To unsubscribe from this mailing list (and be removed from the roster of
the OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/cam/members/leave_workgroup.php.
>
> To unsubscribe from this mailing list (and be removed from the roster of
the OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/cam/members/leave_workgroup.php.
>
>