First post and all that so I apologise in advance if I'm talking rubbish that has been discounted before but...

In the OASIS SOA Reference Model group we did (I think) a decent job of defining the core concepts of SOA.  One of the big goals there (and for me the biggest challenge around SOA) was to be both business and IT agnostic, which is one of the challenges in most modelling approaches right now. I share the disbelief that there hasn't really been any work around what should be the core area of SOA, namely modelling the services.  By this I mean the "blobs" of the services not their interfaces, message patterns or orchestrations which are really about the technical implementation but the actual principles and drivers of the services themselves.

Most of the time people are modelling at the technology level (IME) rather than considering service to be a true architectural approach that impacts the whole IT (and business) lifecycle. 

Nothing is preventing us creating good SOAs, the problem is that the tools positively encorage BAD SOAs to be created because they do not enable us to model a Service Architecture independent of its implementation.  This was in part the goal of releasing the SOA notation/methodology stuff into OASIS last year to try and stimulate the debate at a higher level of abstraction, and hopefully at some stage it will begin to happen.

Steve



On 23/06/06, Robin <[EMAIL PROTECTED]> wrote:

Hi Dennis,
I think the reason why there is no consensus about a detailed model is
that there are many models or architecture views to depict a SOA.
I was last week discussing about SOA patterns with Gregor Hohpe.
We have the same problem at that level, there are message-exchange
patterns but also orchestration patterns, service implementation
patterns, service operational patterns, service interface patterns,...

When modeling a SOA we will model the services implementations, model
the services interfaces, model the processes orchestrating the
services,...

All these models have a different semantic today but is it really
preventing us to create good SOAs?

Robin
http://blogs.ittoolbox.com/eai/applications

--- In [email protected], Dennis Djenfer
<[EMAIL PROTECTED]> wrote:
>
> Thanks Stefan,
>
> It's always encouraging to know that others are struggling with the
same
> issue. It's remarkable though, that there has been written so many
books
> and articles about service-orientation and yet there is no consensus
> about a detailed model of service oriented concepts.
>
> // Dennis
>

.


__._,_.___


SPONSORED LINKS
Computer software Computer aided design software Computer job
Soa Service-oriented architecture


YAHOO! GROUPS LINKS




__,_._,___

Reply via email to