Concur that this group should steer away
from "vendor specific" notation/methodology approaches.
As an example, things that are less vendor specific like
UML may be of interest, as it is a fairly common notation at the
design/detail design/build level and in a "top down SOA" may allow for easier
communication of the target to some key audiences -- may be a bit of a stretch
on UML though. One can get bogged down in notation discussions -- though it
is a sidebar that could be a bit important as the SOA blueprint progresses.
Would be of interest to be aware of parallel thoughts on "SOA &
DSL"
One impact of a top-down approach
to the technology of SOA is
-
may enable correlation back to the business --
the business-facing services one identifies need to have correlation
to business process/events either in 1-1 or aggregate
-
may help with the "bottom up" by providing, of sorts, a
specification to the 'DSIs of the world' as to what underlying general
technical capabilities/functions are required to support the business-facing
services in the SOA context - hopefully they can work towards and SOA
driven functional 'specification' that engenders flexibility
-
may offer opportunity to possibly take a top-down SOA and think
more of how one moves an existing non-SOA towards it rather than to fit it
into a non-SOA architecture
_______________________ Jeff Lamb, CCP Wells
Fargo Services, Enterprise Architecture 415.371.3106 Direct line Jeffrey.Lamb@WellsFargo.com
This message may contain
confidential and/or privileged information. If you are not the addressee
or authorized to receive this for the addressee, you must not use, copy,
disclose, or take any action based on this message or any information
herein. If you have received this message in error, please advise the
sender immediately by reply e-mail and delete this message. Thank you for
your cooperation.
I think NNL sits at a
higher and more abstract level than the SOA Blueprints and normally needs to be
combined with an AIM (Alcohol Ingestion Methodology) J
On SDM/DSI one thing
that is “interesting” is that it’s starting from the bottom and trying to work
upwards. Its going to be very interesting to see how it climbs up the
stack towards SOA from a product centric approach.
On the DSL element,
I’ll be looking at how to represent the SOA methodology using a DSL in the next
couple of months (anyone want to help?) but I’m not sure how this will drive
Service Orientation throughout DSI, its unclear whether it’s a generational or
decoration model, if it’s the former then I’d be very sceptical at it succeeding
in its aims. I agree entirely at looking at this in a vendor agonstic
fashion and MOST importantly driving from the top down and STARTING with
services rather than retro-fitting them into an architecture. On element
that surprises me with all vendors who claim to do SOA is that none have a tool
suite that starts with Services.
However… MOTION is sort
of related to this but at the same time not. It’s a Microsoft Consulting
Services offer that is only available if you employ them rather than being
something that Microsoft are rolling into the product stack.
From: marchadr@wellsfargo.com
[mailto:marchadr@wellsfargo.com] Sent: 02 November 2005 23:59 To: mmatsumura@infravio.com;
marchadr@wellsfargo.com; mike@mw2consulting.com; jharby@gmail.com;
soa-blueprints@lists.oasis-open.org Subject: RE: [soa-blueprints] Microsoft
Motion Modeling Methodology (M*4?)
What Steve doesn't have
napkins where he is?
I am interested in
Steve presenting NNL (Napkin Notation Language) as well.
-----Original
Message----- From: Miko
Matsumura [mailto:mmatsumura@infravio.com] Sent: Wednesday, November 02, 2005 3:26
PM To:
marchadr@wellsfargo.com; mike@mw2consulting.com; jharby@gmail.com;
soa-blueprints@lists.oasis-open.org Subject: RE: [soa-blueprints] Microsoft
Motion Modeling Methodology (M*4?)
Hi Mike and
company,
I'm looking forward
to discussing and resolving the "layers" approach which is implicit in the MW2
consulting blueprints (distinct from the OASIS blueprint effort, just sharing
the name) and how this can inform the path towards stratifying the appropriate
architectural models, patterns and practices.
I'm also looking
forward to seeing Steve demonstrate his notation and methodology on Coalogic,
although I understand he is travelling and may be unable to draw for us at the
moment.
Best,
Miko
From: marchadr@wellsfargo.com
[mailto:marchadr@wellsfargo.com] Sent: Wednesday, November 02, 2005 3:05
PM To:
mike@mw2consulting.com; marchadr@wellsfargo.com; jharby@gmail.com;
soa-blueprints@lists.oasis-open.org Subject: RE: [soa-blueprints] Microsoft
Motion Modeling Methodology (M*4?)
This is a very
interesting evolution in our direction.
When I get time next
week I will toss more things on the wiki helping to flush out the Coalogic
example in this way.
-----Original
Message----- From: Morris,
Michael [mailto:mike@mw2consulting.com] Sent: Wednesday, November 02, 2005
12:41 PM To:
marchadr@wellsfargo.com; jharby@gmail.com;
soa-blueprints@lists.oasis-open.org Subject: RE: [soa-blueprints] Microsoft
Motion Modeling Methodology (M*4?)
There are some very interesting approaches that
Microsoft is taking across the stack here. They are really pushing
their System Definition Model as a part of their DSI strategy to model
(hardware, applications, IT web services, etc) from the infrastructure
domain.
Granted, you have to use all Microsoft tools (eg.
Visual Studio 2005, Virtual Server 2005, etc) but they do provide a DSL
builder to allow others to create tools that generate models that can be
leveraged in other modeling environments.
I'd like to see if we may want to take a vendor
agnostic modeling approach with the SOA Blueprint:
* Start with the business problem and
define it
* lets model a business problem
utilizing a set of business tools - use CapGemini's contribution and put
this metadata in a model.
* Then can we find the right DSLs (they
don't have to be Microsoft) that produce artifacts for every "layer" of the
architecture that we can tie together in a reference solution model to solve
that particular business problem.
* We can utilize a semantic integration
approach to tie together multiple models
* Package this as a best practice
approach to the industry
just an idea on an approach...
Thoughts?
Mike
From: marchadr@wellsfargo.com
[mailto:marchadr@wellsfargo.com] Sent: Wednesday, November 02, 2005
11:36 AM To:
jharby@gmail.com; soa-blueprints@lists.oasis-open.org Subject: RE: [soa-blueprints] Microsoft
Motion Modeling Methodology (M*4?)
Here is more food
for thought:
-----Original
Message----- From: John
Harby [mailto:jharby@gmail.com] Sent: Wednesday, November 02, 2005
10:41 AM To:
soa-blueprints@lists.oasis-open.org Subject: Re: [soa-blueprints]
Microsoft Motion Modeling Methodology (M*4?)
Hasn't Rational XDE been covering that
one? [Marchant, Dan
R.] Yes but it doesn't have microsoft's plan for world domination in
mind. Starbucks and Microsoft should probably merge at some
point.
On 11/2/05, Duane Nickull <dnickull@adobe.com>
wrote:
This is old news. Visio has been doing
UML models since they bought it. The concept of linking that to a .NET
IDE is cool IMO.
Duane
-----Original Message----- From:
Porch Robert [mailto:
porch_robert@bah.com] Sent: Wednesday, November 02, 2005 9:18
AM To: marchadr@wellsfargo.com; mmatsumura@infravio.com; Duane
Nickull Cc: soa-blueprints@lists.oasis-open.org Subject:
RE: [soa-blueprints] Microsoft Motion Modeling
Methodology (M*4?)
I believe that you may be very close based on
what I have seen.
-----Original Message----- From: marchadr@wellsfargo.com
[mailto:marchadr@wellsfargo.com] Sent:
Wednesday, November 02, 2005 11:25 AM To: mmatsumura@infravio.com; dnickull@adobe.com Cc: soa-blueprints@lists.oasis-open.org
Subject: RE: [soa-blueprints] Microsoft Motion Modeling
Methodology (M*4?)
I believe this is the major tool initiative
they are working on for the end of the year or q1 next year
release. Think of RUP with a microsoft toolset and microsoft framework
applied to it.
That is basically what they are looking at doing
I believe.
But I can neither confirm or deny this
:)
-----Original Message----- From: Miko Matsumura [mailto:
mmatsumura@infravio.com] Sent: Tuesday, November 01, 2005 9:54
AM To: Duane Nickull Cc: soa-blueprints@lists.oasis-open.org Subject:
RE: [soa-blueprints] Microsoft Motion Modeling Methodology
(M*4?)
Anyone have any insights on this?
http://weblogs.asp.net/omagnusson/archive/2005/07/28/420856.aspx
Microsoft
Motion Business Modeling Methodology
At the Tech Ed 2005 Europe
Architecture pre-con Beat Schegler and Arvindra Sehmi talked about a
methodology framework that MCS have been working on called Motion which
is supposed to be:
"The Motion Methodology uses the concept of
Business Capabilities to model a business. A Capability describes the
what, not the how. A Capabilities Model abstracts structural
information (capabilities and connections) separately from dynamic
information (processes)"
Motion decomposes the whole business into
capabilities on many levels of granularity ( level 1-3), where level 1
represents core business capabilities (eg. Warehousing), level 2
represents capability groups (eg. Manage Products/Orders) and level 3
represent the business capabilities (order products, track etc. ). Each
business capability then has 80 attributes which describe it, such as
who owns it, input and outputs, best practices and
exceptions.
On top of this we then layer our business processes.
which manage and orchestrate messages going between these business
capabilities.
The Framework is supposed to come with a complete
set of deliverables templates and tasks and a unique "go in, go up, go
out" approach to modeling these business capabilities and processes,
FAQs, case studies and the while shebang.
When goggling around
for this, I've not been able to come up with anything so far. Are there
any news on this, when this will be released and to whom ? Will it be
publicly available like MSF or only to partners ?
I can see the
tooling for this going hand in hand with DSL tools in
the future. posted on Thursday, July 28, 2005 1:30
PM
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.
|
|