Re: [soa-rm-ra] An epiphany

[Francis McCabe <frankmccabe@mac.com>]

Hi Danny,

 I realize that I have some explaining to do...

 I don't see Counts-as as being connected to semantic engagement; at least not at the moment.

 Like I said, this is really about the dichotomy/relationship between private internal action and publicly shared/understood interaction. In my view, the messaging system is, at this level of abstraction, simply a medium in which you can have such a public interaction. It is necessary to have such a medium because we are fundamentally talking about peer-to-peer models crossing ownership boundaries.

 On top of this, is the strong intuition that one can have multiple layers of ecosystem -- systems of systems of systems etc. 

 One of the diagrams that we have drawn internally is like a spider plant:

Spider Plant.pdf

Entities may be interacting with each other within a single ecosystem or across multiple ecosystems. (This is, in part, a reaction to people trying to foist ESBs on us.)

On the whole, the spider plant is only visible to someone considering the ecosystem view. To code, it is fundamentally not visible; on the other hand, the code *does* see a 'stack' of protocols that it must go through in order to get its needs met.

Frank

On Jan 15, 2009, at 8:26 PM, Danny Thornton wrote:

Our strongest semantic tie in the RA is section 3.1.2 Communication as Joint Action as defined in Figure 7.  We definitely need a semantic tie in the RA.  I think Section 3.1.3, Using Communication for Service Action, will make most people scratch their head and wonder what it is for.

I am guessing the epiphany may have had something to do with Figure 8 when defining capability action and message action.  The "Counts as" definition states:

"'Counts as' is a relationship between two systems in which one action, event or concept in one system can be understood as another action, event or concept in another system."

This seems somewhat equivalent to "Semantic Engagement" in Figure 7, minus Content where Content is the message for the purpose of the RA.  I see message as being figurative for whatever is exchanged during a communicative action.

Is Sociology the answer to SOA?

Danny

-------- Original Message --------
Subject: Re: [soa-rm-ra] An epiphany
From: Francis McCabe <frankmccabe@mac.com>
Date: Thu, January 15, 2009 1:08 pm
To: SOA RA <soa-rm-ra@lists.oasis-open.org>

The basic intuition is something like this:

If you are a piece of code, there is nothing to distinguish sending a message from any other action -- other than the name of the method you are calling. Negotiating multiple levels of APIs is just more code. This leads to a view that says that there is nothing special about sending messages -- from the point of view of the code; However, you *do* have to have the right semantic engagement otherwise it *does not work*™.

If you are at the level of the ecosystem then you are not at all concerned with code. You are however concerned with what actions are being performed -- in an effectively public arena. Once you also adopt the concept of systems of systems of systems™ then your focus becomes how players in a system should interact in a way that respects the public semantics of what is going on. It is all about the public semantics of action and not anything really about messages. An individual system, or system of systems may be using MQ series, JMS, or Tibco as a medium of interchange; but as soon as you go to the third or fourth level all that message traffic loses its significance once again.

I know I am not being coherent. And it is alright if the group decides not to follow me down this rabbit hole. I just know that, for me, it is the right mental model.

Frank

On Jan 15, 2009, at 9:43 AM, Danny Thornton wrote:

We are expanding on the concepts in the SOA-RM.  To cut the size of the S0A-RA in half, what specific sections do you think could be cut? 

Danny

-------- Original Message --------
Subject: [soa-rm-ra] An epiphany
From: Francis McCabe <frankmccabe@mac.com>
Date: Thu, January 15, 2009 8:56 am
To: SOA RA <soa-rm-ra@lists.oasis-open.org>

You guys may well not like this, but I believe that I have found a way
to cut the size of the RA in half....

I am not currently in a 'fully formed' state of this idea; but it goes
something like this:

There are primarily two perspectives involved in SOA ecosystems: the
perspective of the code and the perspective of the ecosystem. In the
former, the principal ideas are semantic engagement and counts-as. In
the latter, the principal idea is public semantics. (I know that this
is very cryptic, but it does unfold ...)

On top of this is one further important concept of system cardinality:
we have been talking explicitly about 1 system (the Enterprise SOA)
and many systems (the Internet SOA). In fact, this is not fully
realistic; this structure is itself recursive: we need to be able to
talk about systems of systems of systems; etc. In effect, we need to
be able to characterize how systems of systems are put together.

Some implications that I can see are:
1. no commitment to message exchange.
2. no explicit commitment to description.
3. instead is the concept of private action being interpreted as
public joint action at a particular semantic engagement.
4. to meet a given need may involve 'tunneling' through multiple
layers of such interpretation.

If this were to be fully worked out, it would result in something that
would be better identified as SOA meta-architecture.

It may also be too much.

Noodles of the day :)
Frank

<bottom.letterhead>

smime.p7s