Re: [sca-assembly] Continuing Test Assertion and TestCase review - section7-10

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

Dave,

Thanks again - it is great to have more
eyes looking at this.

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:	David Booz <booz@us.ibm.com>
To:	sca-assembly@lists.oasis-open.org
Date:	02/06/2009 15:08
Subject:	[sca-assembly] Continuing Test Assertion and TestCase review - section 7-10

---

Continuing with Section 7; I'm looking at SCA_Assembly_Test_Assertions_16.odt,
SCA_Assembly_TestCases_22.odt and the CD03 assembly spec.

31) ConstrainingType is not optional in terms of compliance? I thought
it was.

Well, that's an Assembly spec question.  The
Assembly spec has very few optional features.
If you want to change that, it means an Assembly
spec issue...

32) TA-7005 - Any ideas on what to do with this?

Yes, now that all the Contribution stuff
is working well in the testcases, I think the simple thing is to
create a separate Contribution for the testcase
that contains 2 constraint files containing conflicting
constraints with the same namespace/@name
combination.

Continuing with Section 8

32) ASM_8001_TestCase - why does this testcase contribution contain TestInvocation.wsdl
and the two java classes needed for the test? Aren't these supposed to
be moved to other contributions?

Yes, I suppose so.  TestComposite51
is invalid and so needs to be in a Contribution of its own.
I think that we need contributions ASM_8001_xxxx
with Java, BPEL etc

33) TA-8003 - Predicate is wrong. It should be "...as observable by
the service implementation."

Agreed

34) TA-8003 - Here's another one that probably could be tested by language
specific testcases. We've stumbled across a few of these now. It occurred
to me that the assembly TC should provide guidance to each language TC
stating that untestable assembly TAs should have a corresponding testcase(s)
OR a statement that the Assembly TA is also untestable by the language
TC. I'm trying to create some linkage/interlock between rules and semantics
which are stated in assembly but which are not re-stated in the language
specs, and therefore will probably fall through the cracks.

Yes, such formal mechanics for Assembly to
liaise with the language TCs is probably a good idea.

This testcase can probably be tested by both
Java and by BPEL implementations, where the concept of a
long running instance is available, through
the use of callbacks.  I'd have to defer to Bryan as to whether
the same can be done in C / C++

35) ASM_8009_TestCase only talks about covering the reference case, but
the underlying test also covers the service case.

It does.  I'll have to fix this.

36) ASM_8012_TestCase only talks about covering the reference case, but
TA-8014 covers services as well, so I think this test should be updated
to cover services (similar to ASM_8009_TestCase).

I think that ideally this requires a NEW
testcase to cover the service case. TODO


37) ASM80017 - claims to be covered by ASM-TA-7, but there is no such assertion.

Here is a missing TA and a missing TestCase.
 TODO.

Continuing with Section 9

38) ASM_9003_TestCase - interesting. It has a hardcoded URI (<binding.ws
uri="http://localhost:8080/TestComponent1"/>)
that needs to be addressable in order for the testcase to work. The composite
it's in will have to be factored somehow so that the correct URI can be
provided when running this test on a specific runtime.

I'm open to ideas on this one. Entity refs
is one possibility, allowing replacement without changing the Composite
file, but it is a bit cumbersome.

39) TA-10001 - typo, the source is ASM10001 (there's one too many zeroes
in the source row).

Fixed

Continuing with Section 10

No comments :-)




Dave Booz
STSM, BPM and SCA Architecture
Co-Chair OASIS SCA-Policy TC and SCA-J TC
"Distributed objects first, then world hunger"
Poughkeepsie, NY (845)-435-6093 or 8-295-6093
e-mail:booz@us.ibm.com

---

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