On Wed, Dec 19, 2018 at 9:37 AM Rob Berendt <rob@xxxxxxxxx> wrote:
I understand. My point was to go ahead and suggest the RFE as it is not
out of the realm of possibilities for IBM to change defaults.
I totally agree on this. I would vote for it.
And while it's absolutely reasonable that from IBM's point of view,
system-supplied defaults should not be changed lightly, due to
compatibility reasons, they also have to examine the upside and
downside.
If IBM changes the default to *NONE, then the people who are depending
on *ALL will have a "failure" that consists of nothing happening. They
will have to reexamine what's going on and then recode their commands
to be more explicit, or change their own command defaults. I don't
know about IBM, but to me that seems like not a very harmful change.
Disruptive, yes. Annoying, sure. But not *harmful*.
Meanwhile, leaving the default as *ALL means that every so often your
situation ***IS*** going to arise, with people accidentally shutting
off critical services. Don't think for a second that you are somehow
"way above average" in your level of carelessness! ;) It's going to
happen to someone else.
It's just a totally boneheaded default, especially on a system that
prides itself on safety and reliability. It's an absolute footgun and
should be changed.
John Y.
midrange.com
