On Fri, 13 Feb 2009 14:18:08 -0800, Schwarz, Barry A wrote:

>To the nearest power of 10, how many different ways can you make a
>mistake.  How many of these are you willing to pay for the developer to
>handle?  How many other of your desired updates are you willing to put
>on hold while he does?
>
I don't agree, given the complexity of z/OS data management,
that an I/O error on an input data set is unlikely to the
orders of magnitude you suggest.  And for "the developer to
handle" the condition is merely to employ the intrinsic OS
facilities to provide a lucid report.  For comparison, injecting
similar errors:

In HLASM:

1                                                                               
                                PAGE    1
 HIGH LEVEL ASSEMBLER - INTERRUPT AND DIAGNOSTIC DUMP                           
             HLASM R5.0  2009/02/13 15.31
0** ASMA999U Assembly terminated - SYNAD Exit taken - Permanent I/O error on 
SYSIN    data set
          ,SPPGASM ,ASSEMBLE,JES ,I,SYSIN   ,GET   ,WRONG LEN 
RECRD,**************,QSAM


And in IEBGENER:

1DATA SET UTILITY - GENERATE                                                    
                   PAGE 0001
-IEB352I WARNING: ONE OR MORE OF THE OUTPUT DCB PARMS COPIED FROM INPUT

 IEB351I I/O ERROR ,SPPGGENR,TST001  ,OMVS,*,SYSUT1  ,READ  ,WRONG LEN 
RECRD,00000000000000,BSAM S

Whereas:

>-----Original Message-----
>From: Paul Gilmartin
>Sent: Friday, February 13, 2009 1:57 PM
>
><unsnip>
> 14.25.44 JOB06043  IEC020I 001-4,I70APP  ,APPLY   ,SMPCNTL ,JES
> 14.25.44 JOB06043  IEC020I EROPT IS 'ABE' OR NOT SPECIFIED
><snip>
>
... tells me only that something went wrong and the programmer
chose not to report the error conscientiously.

>This tells me hardly anything about what was wrong.  Would
>a SYNADAF have returned a buffer with more information?

Apparently.  I suspect the HLASM and IEBGENER messages
embed SYNADAF results.

>Can a customer reasonably expect that a developer would
>have coded the SYNADAF and displayed the reply in SMPOUT,
>job log, or elsewhere.  I don't think the right answer is
>that if I make a mistake I should be prepared for an
>inexplicable outcome.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to