David, I do not want to argue about the history.
SOA (Web Services, ebXML, and many other SOA related concepts, standards, etc.)
has not suddenly emerged from nowhere.
Suresh, if I understand correctly your point, you
think that the business process modeling, grouping and taxonomy of
processes are important SOA enablers. I completely agree with you. It is
very important but not the only aspect of the entire SOA picture. What I am
also interested in is the SOA enterprise architecture and real alignment between
the business and technology. No marketing hype but real stuff. These are my
questions. Not only for you but for everybody on this thread. I am just
using these questions to continue with discussion. Do not worry I
will provide my answers later and I hope that you will keep adding new questions
as well. The more questions and the more discussions trying to form answers
the better. I am sure we will not change the world but at least we can start to
communicate more often and exchange ideas.
1. How do we integrate business
processes we model as services and technology? Here I mean a real
automatic integration without army of developers working on the "SOA
Enterprise Architecture" to implement processes just modeled or to
implement changes just introduced. To me that kind of architecture is not SOA.
That is old school which we shortly talked about in our history
introduction.
2. How do we make SOA agile? Again automatic
implementation of changes without coding and both static and
dynamic.
3. How do we model and support business process
semantics? Not just message exchanges and message routing.
4. How do we use standards and support standard
convergence?
5. What is our SOA Reference
Model?
6. What is our complete SOA reference
architecture?
7. Do we have an SOA methodology and what the SOA
methodology is all about?
8. What are our SOA best practices?
9. What are the most critical SOA missing points
and failures so far and how to fix them or not repeat them?
10. What is the overall OASIS plan with regards to
SOA, SOA TCs related work and their relationship?
11. How can we start to work together and use each
others "products"?
12. How can we (TCs and OASIS) convince more
leading vendors to be involved in the SOA specs development and start using SOA
specs we are working on?
I think that we should keep the entire original
content of the messages for the completeness of the thread.
Goran
----- Original Message -----
Sent: Tuesday, November 15, 2005 11:06
PM
Subject: RE: [ebsoa] RE: [fwsi] RE:
[soa-rm] RE: [ebsoa] The real SOA challenge?
Hi,
I am new to the group. Please bear with me if I do
not understand how this works.
Agree 100% with David Webber, the key is in this point
that David makes "So much for history - no need to dwell there - the challenge now is to
bring real business enabled XML-driven processes
and SOA to fruition."
Which means we need to
clearly address:
1. SOA and processes integration by broad domains
(Finance, Telco, Transport, Defence, Government etc.) We could use models such
as the eTOM as a starting point.
2. Establish a standard grouping of processes within the
business domains
3. Then look at which of these can be enabled using
XML.
4. In Australia the DOMEDI group did similar work in the
Transport domain (thouogh for a different purpose)
5. Finally a taxonomy of the business, processes,
enablers and XML transactions
Regards. Suresh
________________________________________________
Dr. Suresh Hungenahally|
Capgemini | Melbourne
Senior Manager
477 Collin St. Melbourne 3000
Australia
T. + 61 3 9613 3343 |
M. + 61 402 408 816| F. + 61 3 9613 3333.
Join the Collaborative Business
Experience
________________________________________________________
Goran,
While I like what you say vis SOA here - I have to take you to task on
history ; -)
The ebXML foundation came out of the work of the XML/edi Group aligned
with the work of CEFACT on ooEDI.
It completely breaks the mold on EDI. Yes - things like ebMS are
founded on EDI+ communications - but the core of the ebXMl stack -
registry-centric XML-driven collaboration - and the notion of the "Fusion of
Five" is designed to sweep away the old EDI practices. Unfortunately XML
as implemented today is nothing more than a slightly better EDI. The
whole XML revolution is yet to truely take hold.
As for web services - this was a shameless grab by IBM and Microsoft -
attempting to be first to market ahead of ebXML with concepts that in essence
are nothing more than "real-time EDI" - with all the same strengths and
weaknesses of the
original real-time EDI implementations - just using XML and the
internet instead of EDI and private networks.
So much for history - no need to dwell there - the challenge now is to
bring real business enabled XML-driven processes and SOA to fruition.
Cheers, DW
----- Original Message -----
For example, Web Services are initiated with an idea of
services introduced by Microsoft 6-7 years ago and this idea has been formed
from distributed computing and component-based computing ideas and "learn on
mistakes" lessons from many organizations in 80s and 90s; ebXML is mostly
based on EDI ideas and "learn on mistakes" lessons,
etc.
This message contains
information that may be privileged or confidential and is the property
of the Capgemini Group. It is intended only for the person to whom it is
addressed. If you are not the intended recipient, you are not authorized
to read, print, retain, copy, disseminate, distribute, or use this
message or any part thereof. If you receive this message in error,
please notify the sender immediately and delete all copies of this
message.
|
|