Re: [ebsoa] Re: [ebxml-bp] Re: [ebsoa] Re: [ebxml-bp] Closing thegap between MSI and BSI and move on

["David Webber \(XML\)" <david@drrw.info>]

Sacha,

> I would like to see a standardized BSH-MSH interface just to enable
> ebXML endusers to pick implementations. Or if we have today 30 MSH
> implementations and maybe 0 BSH you cannot choose from the 30 but you
> have to go with one which provides BSH functionalities (I made up the
> numbers).
> Regards Sacha
> 

I like this idea much better;  more concrete.  But we have sorta been
doing this with the Interoperability certification - at least at the ebMS
level with good success.

At the CPA level - this has not been so clear - since the interop' tests
did not focus on CPA functionality.  Consequently how much of CPA
is used depends very much on the implementer.  That we definately
need to improve on by adding more CPA features to interop' testing.

Now this extends to the BPSS level - to include things that the
business processing would want to have consistent across messaging
interfaces.

Just like with CPA - I think there is a minimal list we could come
up with as input to the interop' team - to consider for ebMS support.

Would this be more tangible than a conceptual level?

A caution here however - I spent 4 months last year building
three ebXML based prototypes and evaluating the functionality
they provided and their interoperability.  At the end of it we provided
feedback and source code that enhanced all of the three prototypes
because - while alot of the basics worked - a lot of things also 
worked a little differently.

What we learn from this I believe is that it is the marketplace that
drives better interoperability much more effectively than we can.

However - we can help by identifying key behaviours - so that
customers ask for and expect those - and vendors focus on
those first.  Everyones deployment models are different though so
not everyone has the same priorities.

Anyway - for now we need something useful in the V2 specification,
knowing that we need to do more for strict interoperability testing
as 2005 advances.

Thanks, DW