--- In [email protected], "Steve Jones" 
<[EMAIL PROTECTED]> wrote:
> 
> I'd be a bit concerned if an application, plain or otherwise, wasn't
> part of a business service.  Part of the problem is that people 
> don't raise themselves above the technology because "they 
> understand its just going to the Phoenix Application".  I've had 
> occasions where when you looked at it from the business perspective 
> and found that actually Phoenix shouldn't be the destination at all 
> but it had become the accepted wisdom of IT that it should.

We agree then that architectures should be service-oriented. But 
sometimes they are not.

> At a architecture level it isn't important whether its event or not,
> but at a design level its true that events aren't everywhere.

I disagree. At the architecture level, events can be first class 
citizens, just as services are. There are plenty of business events 
to be identified and specified at the L1 level.

> And on the later I'd disagree.  I've never seen a decent business
> architecture described from an event basis.  If its T-SOA v EDA 
> then I agree, but if its B-SOA then I've just never seen an 
> equivalent B-EDA.

I've never seen an architecture described only from an event basis, 
just as I've never seen an architecture described only from a service 
basis. There are usually other principles presented, though they are 
usually not explicitly called out because everyone is wrapped up 
around the SOA mania.  

> Again I think this is the difference between implementation and
> architecture.  Service Architecture as a principle lends itself, 
> from experience, to good EDA implementations.

I guess we simply disagree to the level of were event-driven 
principles can be applied. Event-driven is not an implementation 
choice, IMO.

> Bloody hell my ground state really is grumpy...

Completely understood. Your posts are no grumpier than usual. :)

-Rob

Reply via email to