Re: [soa-rm] The riot act (toned down version)

[Ken Laskey <klaskey@mitre.org>]

Duane,

I am on travel and have been out of the email loop for a week, so forgive 
me if I've missed something.

I think there is general agreement that
(1) our primary purpose is the RM and we agree with the RM definition
(2) also producing a RA will be useful because providing only an abstract 
RM may not be sufficient to define our conclusions and how we expect others 
to move forward.

We agree a RM is abstract and has a small number of unifying concepts.  Our 
struggle is to identify exactly what comprises that minimum set and what 
interactions need to be included.  The question is how to provide clarity 
and some consensus when there is not immediate obvious consensus at that 
level. This does not mean consensus is not possible, just that it is not 
automatic and we need to identify ways to look at the problem that will 
help us find that consensus.  I tell my children that an engineer rephrases 
a problem until it fits a pattern for which there is a possible solution 
and then makes sure the solution fits the original statement of the 
problem.  It is not valid to point to a convenient solution and insist it 
fits the problem when it does not.  It is also not acceptable to just give up.

How can we rephrase the problem to make it workable while making sure it is 
the problem we have agreed to solve?

Ken

At 12:40 PM 6/7/2005, Duane Nickull wrote:
>TC:
>
>I would like to defer on the current questions for a moment and reflect on 
>a few things.  Things are starting to heat up in the TC and if we are 
>going to make progress, we have to agree on something.
>
>My observations is that we are all after the same goals (or hope we 
>are).  The high level goal is to define SOA.  In our charter, we have 
>agreed to do this by first building a Reference Model, followed 
>possibly  by a RA.  This idea of two separate documents is in our 
>charter.  The current proposal for this is therefore conceptually in 
>alignment, just the names and details require sorting.
>
>We also have a definition of Reference Model in our charter and some Q&A 
>further explaining it.  We all agreed to this when we signed up.
>"Reference Model - A reference model is an abstract framework for 
>understanding significant relationships among the entities of some 
>environment, and for the development of consistent standards or 
>specifications supporting that environment. A reference model is based on 
>a small number of unifying concepts and may be used as a basis for 
>education and explaining standards to a non-specialist. [1] A reference 
>model is not directly tied to any standards, technologies or other 
>concrete implementation details, but it does seek to provide a common 
>semantics that can be used unambiguously across and between different 
>implementations."
>
>Here is the problem:
>
>A TC needs some common ground it can stand on in order to operate. This TC 
>needs to have our charter and referenced materials as common ground to 
>start.    At this point, I hope that no one challenges this definition of 
>a RM.  If we do not have that as common ground, we have some hard times ahead.
>
>We also have defined Architecture in our charter:
>
>"Architecture - Software architecture for a system is the structure or 
>structures of the system, which consist of elements and their externally 
>visible properties, and the relationships among them.[2]"
>
>This appears to be in alignment with what several on this TC are calling 
>for WRT the second document (Which I am also in favor of since it clearly 
>has great value).  The RA (second document) is where things like Service 
>Consumers are going to be present, along with messages, security etc.
>
>SOA RM TC recap:
>
>What is going well:
>
>We have tons of very good talent in this TC and a huge number of active 
>participants with lots of ideas.
>What is not going well:
>
>We have tons of very good talent in this TC and a huge number of active 
>participants with lots of ideas.
>In other words our greatest strength is also our greatest weakness.
>
>So how do we move forward?
>
>1. We have to have our charter as common ground.  If there is a serious 
>desire to challenge the tenets of what constitutes a Reference Model, that 
>argument does not belong in this TC.  That can be done elsewhere.
>It is out of order for this TC to do this and we need to rely on common 
>ground to move forward.
>2. Let's keep focused on the RM.
>
>3. Let's start the RA soon.
>
>Duane
>
>

--
      ---------------------------------------------------------------------------------
   /   Ken 
Laskey                                                                \
  |    MITRE Corporation, M/S H305    phone:  703-983-7934   |
  |    7515 Colshire Drive                    fax:      703-983-1379   |
   \   McLean VA 22102-7508                                              /
     ----------------------------------------------------------------------------------