On Fri, 9 Nov 2018 14:51:59 -0800, Charles Mills wrote:
>Gosh some of the error message documentation -- really the software design
>leading up to the error message -- is just plain awful!
>
>https://www.ibm.com/support/knowledgecenter/SSLTBW_2.3.0/com.ibm.zos.v2r3.bpxa800/fsum7131.htm
>
>FSUM7131 Out of space or reached the end of the archive file. If you want
>to go on, type device or file name when ready.
>Explanation
>There is no more room to write the archive or extract members, or the end of
>the archive file has been reached. It is also possible that the archive is
>corrupt.
>
>Was it really necessary to fold three different error conditions into one
>error message?
>
I hate when they do that!
>... And by the way, "type device or file name" is inappropriate
>in a batch job.
>
This looks like a message from tar. In olden days, the recovery was to
mount a different tape. OMVS tar no longer supports tapes. The developer
hadn't a clue how to adapt the message.
I once got a message from TSO ALLOCATE when I specified PATHMODE but not PATH:
... ENTER STATUS or SYSOUT
Each of the 2 suggested options was rejected as mutex with PATHMODE.
PATH (which should have worked) was rejected because it wasn't one of
the 2 requested options.
I had to argue strenuously with Support, which felt the behavior was OK,
and told me just to cancel the request ("cancel" wasn't an option) and re-enter
with all needed options specified.
When they caved, the behavior became:
... MISSING MUTUALLY INCLUSIVE OPTION.
(It didn't say which). A hard error with no option to recover.
>There, I feel so much better now.
>
Why?
They don't do enough fuzz testing.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
