Re: [sca-j] Re: Fwd: Issue 227: conformance section

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

Anish,

You're right in that indeed the SCA
runtime could throw an exception that makes no sense at all, but
there is no way that the test suite
can validate this in a portable fashion, so all that can be said is that
some form of exception is thrown.  
Even the timing of the exception is unknown since the SCA Assembly
spec makes it clear that for artifacts
in error, an exception can be thrown as early as the point when the
Contribution is given to the runtime
or as late as the point where the related component(s) are run.

Having said that, for the Tuscany runtime
bridge, we actually created a runtime-specific file which holds a
list of the detailed exceptions for
each testcase where an exception is thrown back to the JAXWS client.
This helped ensure that the correct
exception was being thrown - but of course that exception is not
standardized and is specific to the
Tuscany runtime.  However, Tuscany is marked as failing a testcase
if the exception is not the expected
one.  This has been very useful over the months in keeping Tuscany
on the "straight and narrow".


Yours, Mike

Dr Mike Edwards		Mail Point 146, Hursley Park
STSM		Winchester, Hants SO21 2JN
SCA & Services Standards		United Kingdom
Co-Chair OASIS SCA Assembly TC
IBM Software Group
Phone:	+44-1962 818014
Mobile:	+44-7802-467431 (274097)
e-mail:	mike_edwards@uk.ibm.com

From:	Anish Karmarkar <Anish.Karmarkar@oracle.com>
To:	sca-j@lists.oasis-open.org
Date:	24/02/2011 03:01
Subject:	Re: [sca-j] Re: Fwd: Issue 227: conformance section

---

Mike,

The reason I said "appropriate exception" was because in the
case where 
the exception is not specified, it still won't make sense to have a 
ArithmeticException or TypeNotPresentException or something of that 
nature. That would indicate that there is something else going on.
But I suppose we can't categorically say which exceptions are 
disallowed. So I would be ok with what you suggest but perhaps with some
explanatory text.

-Anish
--

On 2/22/2011 7:41 AM, Mike Edwards wrote:
>
> Folks,
>
> I would not separate the Positive and Negative tests in the way Anish
> proposes. I see no purpose in this.
>
> Each testcase has some expected output - whether some positive response
> or an exception - and the thing
> that matters is that the actual response matches the expectation.
>
> A further problem with the proposed wording is that "appropriate
> exceptions" is a misleading thing in that only in a subset of
> cases is the exception specific - in other cases a general exception
is
> all that the spec describes. Also, for many of the cases
> where a SPECIFIC exception is expected, it usually involves cases
where
> the Java API is being used - and in these cases
> the testcase catches the exception in the code under test and validates
> that it is the correct exception - reporting back to the
> test client a POSITIVE response typically along the lines "TEST_xxxx
> expected FooBar exception received" ;-)
>
> So I'd prefer a simpler set of words:
>
> "An implementation that claims to conform to this specification
MUST be
> able to run all applicable Tests in
> Section 2 TestCases, producing the Expected Output.
>
> Note that conforming to this specification is considered a necessary
> though not a sufficient condition to conform to the [POJO] specification."
>
>
> Yours, Mike
> ------------------------------------------------------------------------
> Dr Mike Edwards            
     Mail Point 146, Hursley Park    
             
> STSM                
 Winchester, Hants SO21 2JN
> SCA & Services Standards          
       United Kingdom
> Co-Chair OASIS SCA Assembly TC          
               
        
> IBM Software Group            
               
      
> Phone:                
 +44-1962 818014          
               
        
> Mobile:                
 +44-7802-467431 (274097)          
               
        
> e-mail:                
 mike_edwards@uk.ibm.com          
               
        
>
>
>
>
>
> From:                
 Anish Karmarkar <Anish.Karmarkar@oracle.com>
> To:                
 David Booz <booz@us.ibm.com>, Bryan Aupperle <aupperle@us.ibm.com>,
> Mike Edwards/UK/IBM@IBMGB, Ashok Malhotra <ashok.malhotra@oracle.com>,
> Martin Chapman <martin.chapman@oracle.com>
> Date:                
 21/02/2011 16:15
> Subject:              
   Fwd: Issue 227: conformance section
>
>
> ------------------------------------------------------------------------
>
>
>
> OASIS email seems to be done. Here is my email from yesterday.
>
> -------- Original Message --------
> Subject: Issue 227: conformance section
> Date: Sun, 20 Feb 2011 22:34:51 -0800
> From: Anish Karmarkar <Anish.Karmarkar@oracle.com>
> To: OASIS Java <sca-j@lists.oasis-open.org>
>
> Here is a proposal for replacing the conformance section of POJO CI
> testcase document to resolve issue 227[1]:
>
> ----
> Normative code artifacts related to this specification are considered
to
> be authoritative and take precedence over specification text.
>
> An implementation that claims to conform to this specification MUST
meet
> the following conditions:
> 1. The implementation is able to run all applicable Positive Tests
in
> Section 2 TestCases producing the Expected Output.
> 2. The implementation is able to run all applicable Negative Tests
in
> Section 2 TestCases producing appropriate exceptions.
>
> Note that conforming to this specification is considered a necessary
> though not a sufficient condition to conform to the [POJO] specification.
>
> [The non-normative ref section will have to be updated to include
the
> POJO CI spec reference].
>
> ----
>
> If we wanted to go further along the lines of what we discussed at
the
> last call. Section 11.3 of the POJO spec should be updated to append
the
> following:
>
> "7. The implementation MUST meet all the conformance requirements
> defined by the SCA-J POJO Component Implementation v1.1 TestCases
> [POJO-TC]."
>
> [The normative ref section will have to be updated to include the
POJO
> TC reference]
>
> Comments?
>
> -Anish
> --
>
> [1]
> http://docs.oasis-open.org/opencsa/sca-j/sca-j-pojo-ci-1.1-testcases-1.0-csprd01.pdf
>
>
>
>
>
> ------------------------------------------------------------------------
>
> /
> /
>
> /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/
>
>
>
>
>
>

---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail.  Follow this link to all your TCs in OASIS at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php

---

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