Re: [wsbpel] implicite links of the runtime engine (was: Implicit<sequence> macro)

[Assaf Arkin <arkin@intalio.com>]

Satish Thatte wrote:

>Popping up a level, the basic point is that minimality is in the eye of the beholder.  Why stop with eliminating sequence?  Links can be emulated by message based synchronization.  Concurrency (flow) can be emulated by a spawn feature, using message passing to emulate shared state.  In the end we have message passing including port reference passing, spawn, some notion of abstraction (declaration) and some form of recursion and not much else, i.e., the pi calculus.  Pi is provably able to emulate everything including XML data.  But I doubt it would satisfy most members of the TC as the specification we produce.  There is some element of judgment involved in deciding what is minimal enough for BPEL -- emulation is not a conclusive argument.
>
>  
>
Indeed some forms of "simplicity" actually introduces a lot of 
complexity ;-)

>I am sure none of you will disagree at this level, although the only argument I have heard for eliminating sequence so far is that it can be emulated with flow and links.  
>  
>
I'm going to start by disagreeing for a second and see what the 
consequences are.

Just as a reminder, my criteria is to "reduce the complexity of the 
specification while allowing interoperability/portability" (without 
which you don't really need a standard, do you?).

If you go all the way to pi-calculus you end up with a situtation where 
two systems may decide to use different WSDL operations or construct 
different messages for the same process definition since there is no 
interpretations of how names/actions should be mapped to WSDL. So to 
answer Edwin's question, you lose interoperability , you might as well 
have no spec.

Or, you can formalize the manner in which WSDL operations are used with 
a pi-calculus like definition, but then you need to add further rules to 
the specification. Off the top of my head I can think of two/three rules 
and I can ensure you they are more complex than the definition of 
sequence. So while turning the language from a high-level model to a 
low-level model, you would significantly increase the complexity of the 
specification.

So here's a very simple criteria for accepting/rejecting constructs that 
would accept the removal of <sequence> but would clarify exactly why you 
would not attempt to take <flow> out of existence.

Does this help?

arkin

> 
>
>Satish
>  
>