[soa-forum] RE: loose coupling

[g.thomas@xxxxxxx]

The SCA spec does a reasonable job of
describing loose coupling based on independence of service deployment (see
osoa.org). When services are composed (into b2b/g2g/a2a collaborations)
across disparate (in/out-sourced, org/geo boundary crossing) service hosts
(app-srv/containers), this relaxes technical and business coupling, leading
to market driven (CoI) composite apps, and tech/bus (insert your favorite
abused term here...) 'agility'. SCA by no means has a monopoly on this
idea, just wanted to point to something that does a reasonable job of trying
to distinguish tight/loose coupling in the context of an emerging SOA standard.

-g

"Metz Rebekah" <metz_rebekah@xxxxxxx>
Sent by: soa-forum-bounces@xxxxxxxxxxxxxx

08/17/2006 01:29 PM

Please respond to
"Service-Oriented Architecture CoP" <soa-forum@xxxxxxxxxxxxxx>

To	"Service-Oriented Architecture CoP" <soa-forum@xxxxxxxxxxxxxx>
cc
Subject	RE: [soa-forum] Question/Issue for the SOA CoP

---

From: soa-forum-bounces@xxxxxxxxxxxxxx
[mailto:soa-forum-bounces@xxxxxxxxxxxxxx] On Behalf Of David RR
Webber (XML)
Sent: Thursday, August 17, 2006 1:18 PM
To: Service-Oriented Architecture CoP
Subject: RE: [soa-forum] Question/Issue for the SOA CoP
 
[->] David,

 <snip>
and bravo!
 
I've also seen the need to return
to the critical topic of "loose coupling".  Everyone claims
they are doing loosely coupled implementations, but are they really?!?
 
What does loose coupling REALLY
mean - both in terms of transport layer integration, AND information exchange
integration?!
[->] I applaud your
recognition that what the phrase ‘loose coupling’ intends to convey is
not just integration at a technology level – it is at the business level!
 There are certainly many aspects of integration and I believe that
all of them have an impact on ‘loosely coupled.’  What is actually
loosely coupled?  If two systems both speak http, does that mean that
when those systems are used as tools that they are ‘loosely coupled?’;
are the humans that rely on those tools ‘loosely coupled?’ 
 
The XML/edi Group first coined
the term 'loosely coupled' back in 1998 - but since then it has taken on
a marketing gloss - without folks really articulating what and how their
toolsets actually support it.
[->] In a recent conversation
with a colleague, we also spoke of ‘coarse-grained’ in this same vein.
 When is something ‘fine-grained vs. coarse-grained is largely dependent
on perspective.  
 
Providing clear specifications
and standards in support of loose coupling remains just as tough now as
back in 1998.
 
Fortunately folks have at least
learned now what does NOT work with XML and schema - and are now returning
to the well to draw out what insights they need to leverage and understand
to resolve the operational issues.
 
Hints: key words - Role, Context,
self-adapting, fault tolerant, agile, simple XML....
[->] I also spent
several years in the field of grid computing research, just as it started
to migrate toward a web services based architecture.  I would venture
that there are many lessons in terms of context, adaptation, agility and
tolerance that could be learned from the grid community on these topics.
 
DW
[->] Rebekah
-------- Original Message --------
Subject: Re: [soa-forum] Question/Issue for the SOA CoP
From: "ajit kapoor" <ajitorsarah@xxxxxxxxxxxxx>
Date: Wed, August 16, 2006 6:18 pm
To: "Service-Oriented Architecture CoP" <soa-forum@xxxxxxxxxxxxxx>


Chris,
 
Thanks for sharing your work. I must say
I agree with you in so much as it addresses the necessary technology conditions
that must be met to have a SOA paradigm, but the sufficiency has largely
been ignored (may be based on assumption that people and process based
practices will follow).
 
I have been quite intrigued by the "SOA
Buzz" industry wide but am concerned by the lack of understanding
of the architecture (if viewed only from a technology viewpoint). The best
practices should be addressed fundamentally from the view point of standard
processes, people/organizational culture (a basic propensity to hinder
change)and then overlay the technical layer on it. Also even on the technology
front the standards are still in flux and the best efforts to integrate
higher layer of the SOA protocol stack must rely on the WS Interoperability
forum. Vendors are selling point solutions for the higher layer protocols
and we will end up creating EAI equivalent of the SOA integration problem.
Where will the values be in terms of reusability, etc. I believe the DoD
and Government in general as part of the GIG/NC initiative must take the
lead in coalescing multiple standards on all the fronts of people, process,
and technology so as to start talking about best practices building blocks.
 
I look forward to other comments and potential
recommendations. The result of this exercise should be, in my opinion,
a roadmap from current SOA platform to the desired state with People, process
and technologies roadmap defined in an evolutionary and standards way.
 
regards, 
ajit kapoor
Lockheed Martin
EIS/CTO
(retiring October 1, 2006)
----- Original Message ----- 
From: Chris
Gunderson 
To: 'Service-Oriented
Architecture CoP' 
Sent: Wednesday, August 16, 2006 4:32
PM
Subject: RE: [soa-forum] Question/Issue
for the SOA CoP
 
Brand, I’m working on a project
sponsored by DISA Joint Interoperability Test Command.  A key component
is the need to develop appropriate SOA implementation and validation and
verification guidance.  The idea is to provide a useful service to
GIG software developers that encourages re-use of best practices and code.
 
 
The attached is a work in progress.
 
 
So….1.  We have a mandate
to do this work.  2. We have no “not-invented-here” issues. 3. We
want to port as many industrial best practices as possible.  4.  We’re
on a fast track.
 
We welcome any and all help
from the SOA Forum.  The attached is very much a work in progress
with no pride of authorship. 
 
Best, Chris
 
Chris Gunderson
Research Associate Professor of Information
Science
Naval Postgraduate School
Principal Investigator, W2COG and Netcentric
Certification Office Initiatives
(O) 703 262 5332
(C) 831 224 5182

---

From: soa-forum-bounces@xxxxxxxxxxxxxx
[mailto:soa-forum-bounces@xxxxxxxxxxxxxx] On Behalf Of Niemann.Brand@xxxxxxxxxxxxxxx
Sent: Wednesday, August 16, 2006 3:08 PM
To: soa-forum@xxxxxxxxxxxxxx
Subject: [soa-forum] Question/Issue for the SOA CoP
 
I have been asked about the ownership/governance
issue and whether any best practices exist for SOA. We covered governance
in the First SOA Conference for E-Government, but I don't recall specifically
ownership being discussed. Any suggestions, references, etc. would be appreciated.
 
Thanks, Brand

---

_________________________________________________________________
Subscribe/Unsubscribe/Config: http://colab.cim3.net/mailman/listinfo/soa-forum/
Shared Files: http://colab.cim3.net/file/work/soa/
Community Portal: http://colab.cim3.net/
Community Wiki: http://colab.cim3.net/cgi-bin/wiki.pl?AnnouncementofSOACoP

---

_________________________________________________________________
Subscribe/Unsubscribe/Config: http://colab.cim3.net/mailman/listinfo/soa-forum/
Shared Files: http://colab.cim3.net/file/work/soa/
Community Portal: http://colab.cim3.net/
Community Wiki: http://colab.cim3.net/cgi-bin/wiki.pl?AnnouncementofSOACoP
 _________________________________________________________________
Subscribe/Unsubscribe/Config: http://colab.cim3.net/mailman/listinfo/soa-forum/
Shared Files: http://colab.cim3.net/file/work/soa/
Community Portal: http://colab.cim3.net/
Community Wiki: http://colab.cim3.net/cgi-bin/wiki.pl?AnnouncementofSOACoP

 _________________________________________________________________
Subscribe/Unsubscribe/Config: http://colab.cim3.net/mailman/listinfo/soa-forum/
Shared Files: http://colab.cim3.net/file/work/soa/
Community Portal: http://colab.cim3.net/
Community Wiki: http://colab.cim3.net/cgi-bin/wiki.pl?AnnouncementofSOACoP    (01)