Re: [ebxml-iic] call Monday 11am PT

[Monica Martin <monica.martin@sun.com>]

Michael Kass wrote:

> Jacques,
> Attached are my comments for today's conf call. In short, here are the 
> highlights:
> *Test Requirement #2* - How to write an abstract test to represent 
> testing all global syntax and/* semantics,*/
> as specified in ebMS. I don't think that one abstract test can do 
> that, but that the combination of ALL abstract
> tests can reasonably conclude conformance to syntax and semantics of 
> ebMS specification

mm1: This is a reasonable approach and consistent with other testing 
work of which I have been involved. This begs the question
whether we should start to consider the packaging for test scenarios 
more fully in the next version of the specification. I know we begun to 
do so
in the context of CPP/A, but this requirement indirectly indicates and 
others like it may provide some insight into further packaging for 
efficiency (taking a modular approach to the testing process).

> *Test Requirement #56* - I do not believe that we disagree about what 
> the specification is saying…
> I believe it is saying (rather poorly, and Pete and I agreed) if there 
> are no errors, there is no errorlist…
> NOT there are no “empty” ErrorLists present in a message…. Comments?

mm1: Agreed - may be worthy of a comment back to ebMS team regarding 
some more definitive language.

> I've also added some typo/grammar changes to the Abstract Test Suite 
> section that you wrote.