Re: [ebxml-iic] RE: Concurrent thread use cases using <If><Then><Else> syntax

["Monica J. Martin" <Monica.Martin@Sun.COM>]

Jacques Durand wrote:

> Mike:
> comments attached <JD6>.
> I think we have resolved most issues.
>  
> A few comments on the latest BPSS POC cases scripts:
>  
> #1: I don't see that we need an "if...then..." in that test case:, we 
> just needs to fork an
> a thread (for handling "exceptions" ) asynchronously.
>  

mm1: The only comment I would make, and it really is a comment recently 
found by ebBP: The nesting concepts (such as those for if-then-else) may 
have limited function when handling complex (multi-party) interactions 
for the process and presumably any testing that supports it. We are just 
formalizing these ideas so I should have more details soon.
Thanks.

> #2: threads for M2 and M3 could just be started using a "split", and 
> then and-Joined for
> proceeding further on M4. If one wants either thread to "fail" the 
> test case in case of
> problem (timeout, bad message....) then we could use an explicit 
> "fail" statement in them.
>  
> #3: the processing of M6 is forgotten (both threads are same). And 
> here, it is a case where we could
> keep the reception of M2 in the main thread (after all, it is common 
> to both subcases !), make the precondition
> test to accept both "approval" or "rejection" (i.e. anyhting else 
> would make the case not applicable),
> and in the Assertion, we test for "approval": if True, we start Thread 
> 1, if not we start Thread 2.
> Could we enhance the matchResult tomake it a standalone element that 
> does this:
> <matchResult>
> <Result value="true">< ...start "approval" thread ...></Result>
> <Result value="false">< ...start "rejection" thread ...></Result>
> </matchResult>
> The same notation could be used to "fail" or "pass" a test case based 
> on some Assertion result.
> Opinion?
>  
> Jacques.
>  
>
>