RE: [wsbpel] Groups - BPEL4WS 1.1 Issues Log May 23.doc uploaded

[David RR Webber - XML ebusiness <Gnosis_@compuserve.com>]

Ugo,

Another way to go at this is to state in the requirements that this
will be a layered specification.

Three layers:

1) Base - simple stuff - guaranteed no surprises / side effects.

2) Medium - normally good in well understood deployment models.

3) Advanced - recommended for internal use only.  New and exotic
    features.

Or similar classifications.

Then a conforming BPEL script @ level 1, running in level 1 conformant
processor should give you very high interoperability, and so on.

This is the approach we've taken in OASIS CAM already.

Thanks, DW.
========================================
Message text written by "Ugo Corda"
> 
Well, I guess it depends on how we define portability in the context of
BPEL. 

One definition of BPEL portability could be the following: to be able to
take a BPEL executable process that works in a particular environment, run
it in a different environment (different BPEL engine, different WSDL end
points), and obtain the same observable behavior in the new environment.

If we take this definition, it seems to me that BPEL portability should be
one of the main reasons for creating the standard, and should be one of the
major goals to be achieved by this TC.

Ugo
<