M/F terminology tai-kwan-do, ad nauseum

["JACK P. SHAW" <jsh -at- SOFTWARE-AG -dot- DE>]

The VM/CMS/operating system/TP monitor/whatall controversy
is exactly what I was hoping to avoid when I answered Carl
Gotlieb's request for mainframe definitions--I should have
known better...

Carl asked for a "generic term for the so-called 'operating
systems' that run under MVS and VM...", and so I took the
analogous approach by defining CMS (the Conversational Monitor
System component of IBM's Virtual Machine host operating system
("host"--a key word here...) as an interactive (read, "con-
versational") monitor when it acts as "a two-way communication
system between system users and CMS (IBM VM/ESA General Info.,
p. 20)".

I oversimplified by saying "...there is the possibility to run
other operating systems such as MVS and VSE in CMS...", and
referring to CMS as a "platform for other operating systems".
Technically, VM is the platform for "guest" operating systems
such as MVS and VSE; CMS is, indeed, not required. However, anyone
I've asked who has run such guest systems on VM said they were hard-
pressed to do so without CMS to do such things as command lists, user
ID definitions, etc. And, CMS is part and parcel of VM.

In fact, the classic "traffic cop" functions (resource and job
management, input/output, etc.) are performed by the Control
Program (CP) component of VM. But I don't think that's what Carl
wanted to know. Here is where I have trouble with Melvin Classen's
comment that CMS is a "true" operating system. Historically and
functionally, a "true" mainframe operating system is a resource
manager first, and an interactive user interface a distant second.
To my mind, CMS itself is much more the latter, very little the former.

As a dyed-in-the-wool VM/CMS bigot myself, I understand someone taking
issue with the tendency to define VM/CMS as anything less than
a full-function mainframe platform. IMHO, if IBM had given it
the hype it was due in the past, they wouldn't be down there where
they are today. But everyone has the same problem with pidgeonholing
VM/CMS that Melvin Classen does when he says:

  "...Admittedly, CMS "cuts some corners"; it is not possible
   to run CMS on the "bare" hardware, unlike MVS/ESA and DOS/VSE...".

True, but then you can't do much with CMS anyway without VM and
its constant companion, CP. Again, in conflict with Melvin's comment,
it's in fact CP and not CMS that creates the "virtual reality" effect
to let you think you've got your own printer, disks, card reader(!!)
and the like. To belabor, an IBM definition:

      "Control Program (CP). A component of VM/ESA that
       manages the resources of a single computer so multiple
       computing systems appear to exist. Each of these...
       virtual machines is the functional equivalent of an
       IBM System/...", etc.

--from IBM VM/ESA General Information, p. 187.

As such, I deferred to my understanding of CMS as IBM defines it--
the tool to, "...Write...programs, create...info. files, run application
 programs, process jobs...,manage the CMS working environment...
(etc./IBM VM/ESA General Info., p. 20)". Hence, I classified
CMS as a "TP monitor"--itself an unwise choice of terms and not
recognized by ISO; "interactive monitor" is the "accepted" term
I show for the user interface to a mainframe operating system.

It would have made the issue moot if Carl could have just called
the whole shooting match, "VM/CMS operating system". But Carl's
customer wasn't buying that, as he said.

As I said, ad nauseum... . Sorry, Carl. Mainframe people are like
this... . Basta.


Jack Shaw/jsh -at- software-ag -dot- de
Software AG (who takes no responsibility for and little pleasure
from my opinions...)
Darmstadt, Germany