Re: [sca-bpel] Issue 61: proposal

[Mike Edwards <mike_edwards@uk.ibm.com>]

Folks,

I'm finding it a bit hard to understand
the reluctance to use the term "SCA runtime".

The SCA Assembly specification is absolutely
stuffed with references to the "SCA runtime" and there
are quite a number of normative statements
that use that term explicitly.  Indeed, there is even a whole
section devoted to "SCA Runtime
Considerations", and the "SCA runtime" is a conformance
target.

So why be coy about using that term
in the SCA BPEL specification?

From the discussion on the call, it
seemed as if Danny's main issue was actually over the point in time
when the SCA configuration data is taken
and used to determine the values of variables in the BPEL
process.  The SCA Assembly spec
makes it quite clear that a whole series of alternative runtime
designs are catered for, from everything
being statically determined ahead of time to a completely
dynamic system where the configuration
can be changed on the fly.  Using the term "SCA runtime"
accommodates all of these design points
and does not have implications about the point in time at which
things are done.


Having preached my sermon, here is my
proposal for the text in question:

"The WS-BPEL specification requires
that the variable initialization logic in scopes that contain (or whose
children contain) a start
activity use only idempotent functions
in the from-spec. In SCA, WS-BPEL process variables which are associated
with either 
SCA properties or with SCA multi-valued
references have their values set by the SCA runtime based on the configuration
of the 
SCA domain composite for the component
which uses the process as its implementation.  Idempotency applies
for a given 
configuration of the SCA domain composite
- but if the configuration of the SCA domain composite changes, then the
values
of properties or of references can change.
The SCA Assembly specification allows for a variety of SCA runtime designs,

including ones that permit the configuration
of the SCA domain-level composite to change dynamically. See the SCA Assembly
specification [ASSEMBLY], specifically
the sections on "Domain-Level Composite" (10.7),  "Dynamic
Behaviour of Wires in the SCA Domain"
(10.8) and "Dynamic Behaviour of
Component Property Values" (10.9) for details. For such runtimes,
the values of properties and of
muti-valued references can change with
a change in the configuration of the Domain-level composite."


Yours,  Mike.

Strategist - Emerging Technologies, SCA & SDO.
Co Chair OASIS SCA Assembly TC.
IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
Phone & FAX: +44-1962-818014    Mobile: +44-7802-467431  
Email:  mike_edwards@uk.ibm.com

From:	Danny van der Rijn <dannyv@tibco.com>
To:	Anish Karmarkar <Anish.Karmarkar@oracle.com>
Cc:	OASIS BPEL <sca-bpel@lists.oasis-open.org>
Date:	14/01/2010 17:55
Subject:	Re: [sca-bpel] Issue 61: proposal

---

As I was discussing on the call, I'd like to suggest changing
the following sentence:

"SCA BPEL In
an SCA WS-BPEL Runtime, variables associated
with SCA properties and
multi-valued reference references
are set by the runtime
based on the configuration
of the containing composite. "

I have to check, but on last week's call, Mike was saying that Assembly
has a defined term (along the lines of "Domain Configuration"?),
which I think should be explicitly referenced instead of  the phrase
"based on the configuration of the containing composite." 
I think that the current wording is too loose, and confusing.

On 1/14/2010 12:15 AM, Anish Karmarkar wrote:

At the end of section 3.1 Properties we add the following text:

The WS-BPEL specification requires that the variable initialization
logic in scopes that contain (or whose children contain) a start
activity use only idempotent functions in the from-spec. In an SCA
WS-BPEL Runtime, variables associated with SCA properties and
multi-valued reference are set by the runtime based on the configuration
of the containing composite. Therefore, the idempotency is guaranteed
only for a given SCA domain-level composite. The SCA Assembly
specification allows for a variety of implementation including ones that
allow dynamicity for the SCA domain-level composite. See Sections
10.7.1, 10.7.2, 10.8 and 10.9. For such runtimes, the properties and
muti-valued references may change with a change in the domain-level
composite.

Comments?

-Anish
--

---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail.  Follow this link to all your TCs in OASIS at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php

---

Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU