[drmaa-wg] Testcase ST_EMPTY_SESSION_CONTROL

[Peter Troeger]

I can agree to your argumentation, but then I suggest a tracker item for 
spec improvement. The original Condor DRMAA author and you did not have 
the same understanding in this point, so it must be formulated more 
explicit in the spec.

Regards,
Peter.

Andreas Haas schrieb:

>On Sun, 16 Oct 2005, Peter Troeger wrote:
>
>  
>
>>Hi,
>>
>>the current test suite contains a test case
>>(ST_EMPTY_SESSION_CONTROL) where all drmaa_control() variations are
>>performed on an empty session. The original test case from Sun
>>expects DRMAA_ERRNO_SUCCESS in all cases, but the spec says nothing
>>about the behavior in this situation. Some of this calls are not
>>successful with the current Condor DRMAA library. I would like to
>>remove this test case, since this seems to be implementation-
>>specific. Comments ?
>>    
>>
>
>Hi Peter,
>
>I thought ST_EMPTY_SESSION_CONTROL would be a natural and reasonable
>return value for drmaa_control(DRMAA_JOB_IDS_SESSION_ALL) being
>performed on an empty session. The purpose of
>drmaa_control(DRMAA_JOB_IDS_SESSION_ALL) is to perform the control
>operation on all jobs within the session. In case of the empty
>session it is clear that no error can occur, thus it succeeds :-)
>
>But anyways. For me it is obvious this case will occurs not too
>frequently so I'm fine if you skip it from the compliance testing
>procedure.
>
>Best regards,
>Andreas
>
>  
>