In reviewing the discussion, I find some common agreements. (01)
Graham said: (02)
"
the
application of a service oriented approach to organisation and service
design of internal services is as valid to the application of S.O. to
the provisioning of external services. On slide 3, the phrase <business
layer> "A set of services that an enterprise wants to expose to
customers and clients" seems of exclude applying S.O. to the internal
design of organizations.
" (03)
he further asserts (04)
"I believe this to be too prescriptive. I
suggest a change." (05)
I believe that Charles is seeing a conceptual issue when there is an
orientation towards a consumer, where this orientation is from the supplier.
The question of "how" service provisions arise is missing. (06)
The core issue may be that service can be something done for one's self, or
even done for someone else without the other person knowing, or the supplier
knowing. (07)
This means, perhaps, that transparency over the space of transactions should
be an part of the reference model, and as part of various definitions of
service. How is this to be done? (08)
I was going to check the reference model to see how, or if, the issue of
transparency over the transaction space is described. However the link the
Ken gave is password protected. (09)
The solution that I favor is the harvesting of some source of textual
materials that is being posted as part of real time discussion between
suppliers and consumers, in addition to some type of monitoring of the
actual service exchanges that fit within the definition of the registries
governing these exchanges. Some filtering and human touching of the
"measurement" process is needed, but the goals seems clear and perhaps
should be prominent in in a reference model. (010)
I believe that such transparency is implicit in the notion of a blueprint or
template based decision loop involving participants. (011)
I fully realize that there are many issues that lie hidden, but as Rex
mentioned this issues are going to be part of the solution whether or not
the high level reference model make clear what these issues are. (012)
Rex said: (013)
"I can attest that the scope of our work in this
reference model is suitable for inter and intra enterprise
application of SOA principles" (014)
And of course, if one sees the transaction space from a general systems
point of view. But it is easy for this point of view to be lost, and (015)
Rex continued: (016)
"I am a bit
concerned that the marketing engines of the major players in this
arena are primed to spread a lot of smoke around, so anywhere we can
achieve clarity without having our crispness tied to a specific, let
alone proprietary technology, whether that is an endorsement of an
architectural style such as WSDL & SOAP-based web services versus
REST and AJAX or a focus within or outside the organizational
boundaries, I think we better serve ourselves and the marketplace' (017)
I believe that the natural tendency to shift the system viewpoint to
something designed to gain ROI is where we all lose, including those who are
trying to provide new benefits to the marketplace, and to thus earn income
for themselves and their companies. (018)
The ability to shift from internal service to external service and from
supplier to consumer is the key flexibility that the market needs. (019)
This should be addressed at the highest level of abstraction, the reference
model, and then re-expressed in all SOA standards. (020)
This is my opinion. (021)
I frankly agree that REST (Fielding's representational state transfer) and
things like Ajax represent new architectural concepts that allow the type of
complexity that systems theory suggests is required of a type service
orientation. (022)
_________________________________________________________________
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 (023)
|