On 2015-12-08, at 10:05, J O Skip Robinson wrote:
> When you're the only kid in the toy store, you have free reign. Even z HMC
> uses the 'write-back' function for tuning updates. But z/OS is a complex
> shared environment. You can't allow random process-altering commands to
> update common control data sets. Recipe for chaos.
>
A professor in a class I took once countered such an argument with:
"Why do you assume it has to be done wrong!?"
Straw man. A wrong way would be for a sysadmin with a Windows-geek
orientation to:
o FTP a PARMLIB member to a desktop.
o Edit it with Notepad.
o FTP it back.
In fact, it's wrong only if two of them do it at once.
ISPF EDIT has some effective techniques for serializing updates to
PDS members, precluding two programmers' editing the same member
simultaneously. Certainly an interactive tool with a "Save as
Default" capability is less chaotic than handwritten notes to operators,
"When you IPL, be sure to issue all the following SET commands ..."
It seems to me that having a "... great many changes ... now made simply
by command ..." is equally a "[r]ecipe for chaos." Only slightly less
so if the commands are embedded in a script automatically run at IPL.
>
>
> On 2015-12-07 09:58, J O Skip Robinson wrote:
>> Gil's point raises an issue more critical than just the question at hand.
>> Once upon a time, 'reading JES2 parms' would have been a reasonable strategy
>> in general for determining how JES2 runs. Since the advent of pervasive
>> dynamic changes, however, the init deck as coded is no longer a reliable
>> window into JES2 processing. A great many changes are now made simply by
>> command. Old values are ignored on hot start and in many cases even on
>> all-system warm start. Only a cold start will reinstate coded parm values
>> that might actually be years out of date.
>>
>> There is today no substitute for a display command with full detail.
>>
> More modern systems, often on desktops, have similar dynamic change
> facilities. However they often have a "Save as Default" checkbox which does
> the equivalent of writing the changes back to the init deck and making them
> persistent.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
