NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/bit/listserv/ibmmain/2920

Text File | 1992-12-30 | 5.5 KB | 101 lines

Comments: Gated by [e-mail]
Path: sparky!uunet!stanford.edu!bcm!convex!darwin.sura.net!paladin.american.edu!auvm!UTMARTN.BITNET!##09
Message-ID: <30DEC92.11249007.0009.MUSIC@UTMARTN>
Newsgroups: bit.listserv.ibm-main
Date:         Wed, 30 Dec 1992 10:24:56 CST
Sender:       IBM Mainframe Discussion list <[e-mail]>
From:         Larry Holder <##[e-mail]>
Subject:      Reply to the new year
In-Reply-To:  In reply to your message of TUE 29 DEC 1992 08:31:00 CST
Lines: 89

> Once again I am getting ready to teach our (applied) operating systems course.
> I draw this straw every couple years. And this time I am more confused than
> ever.
>
> The trade press makes it sound as if everyone is abandoning mainframes for uni
> boxes, networked pc's,  and the like. I ask myself what I should really
> emphasize in my course. I think one answer is that the basics haven't changed
> that much and I still must deliver them. But how to package them - what
> system(s) to use for examples and exercises - those are the questions.
>

The trade press is always going to emphasize the new and exciting stuff.
They will point out innovations and trends that do need to be looked at.
However, in practicality, mainframes are not dead, nor is their
traditional use going to disappear overnight.  Using our shop as an
example, we are beginning to push ahead toward some distributed
processing, but there are no forseeable plans to chuck the mainframe
applications.
I am not saying that our careers will not be radically changed as time
marches on; I am only saying that there will not be an overnight
overhaul.  Take the old "Cobol is dead" forecast.  Why didn't it die?
Because too much time and money had been invested in it, and to drop
it and convert everything to something else was not feasible.
Yet how many articles about Cobol have you read in the trades lately?
Other than once in a great while for Cobol-II, don't hold your breath.
(I once had an article rejected by a nationally-read mainframe magazine
because the Cobol subject matter had been deemed "old stuff", not
taking into account that folks who have not been in the business as
many years might find the material new and valuable to them).

I agree with your point that the essentials are still there in most
any system you pick as your guinea pig.  Queues, priorities, SVC's,
etc etc, are the concepts you are trying to get across.
Whether you demonstrate them using MS-DOS on a PC or VM on a mainframe
really gets down to this: what operating system can the students get
the most access to, and have the least detrimental affect upon your
administrative (production) computing?

Consider other courses we have: Assembler is still taught on the
IBM mainframe.  It costs us a small partition in VSE for their jobs.
No big deal.  Could it be taught on a PC just as well?  Probably, but
the professor's expertise is with the mainframe flavor, so why change.
Or our CICS application programming course.  It has its own online
partition, so as not to affect any real-use CICS.  Would we like to
get it off the mainframe?  Yes, and Micro-Focus has the products.  What
is the holdup?  Tight budgets, so there goes CICS on the PC (for now).

Back to the operating system teaching.  Seems like Peter Norton wrote
a really good book titled something like "Programmer's Guide to the
IBM PC" and had all kinds of neat try-this sections, and covered a
lot of internals stuff.  I may be getting two of his books combined;
seems like he also wrote "Inside the IBM PC".  Anyhow, find them both.
See if there's enough internals and enough try-this's to satisfy the
hands-on requirements.  The great thing about messing with a PC is,
if you crash it, so what.  Turn it back on and try again; no one else
is affected.  Even if you teach using MS-DOS, it does not prevent you
from delving into mainframe internals via lecture.

> If you can spare a few minutes I'd really like to know what is happening in
> your shop. Is an MVS or VM mainframe as important as ever? Are you likely to
> abandon your 3090's for Risc/6000's or Suns in the near term? (Near term might
> be 2 - 5 years.) Are your mainframes coexisting well with networks? What
> operating systems do you want new hires to know about? etc?

We have a RISC/6000, among other small machines, but we are working them
into cooperative and distributed processing schemes at present.

By the way, thinking of size, our 9221-150 has such a small footprint
in our machine room, compared to the 43xx box and 3880 dasd we moved
out, that I half-joked about putting a pool table in the empty space.
I think hardware technology is going to (or should I say already has to
some extent) blur the lines dividing micro/mini/mainframe, in terms of
size, speed, etc.  Perhaps my VSE/ESA operating system will never die;
in a decade it may run on a box sitting on part of my desk.

> I will really appreciate the information. If I get eneough replies the summary
> might be valuable to everyone.

The above replies are totally my own hot air and all disclaimers
apply.  As I am in the computer center and only serve as an adjunct
faculty member, I can get you the name of the instructor most likely
to be your counterpart here, if you'd like.  Hope some of this helped.

Larry Holder,  Senior Programmer/Analyst         Bitnet: ##09@UTMARTN
The University of Tennessee at Martin            Phone:  [phone]
Martin, Tennessee  38238  U.S.A.                 Fax:           -7841

VM/ESA 1.0, VSE/ESA 1.1, CICS/VSE 2.1, COBOL-II, DL/I, SQL/DS, QMF
PSW (Personal Status Words): "Saved By Grace" (Romans 1:16 & 8:38)