Re: Java and the AS/400

Chris,

Darn!  I just figured out the next big technology improvement, but can't
figure out how to capitalize on it!!!  If someone could just figure out how
to make the calories that you and I burn replying to each other affect
something other than the brain, wrists, and fingers, we'd be _RICH_ ;-D!!!

In a message dated 97-11-15 10:44:44 EST, you write:

> Now for the big issue, what does Java mean to the AS/400?
>  
>  I have often wanted to give a seminar called "What the Hell is Java
>  and why should I care?" for AS/400 managers.
>  
>  Java is often maligned on the midrange circuit. Usually (no insult),
>  it is because midrange managers don't bother to look into Java any
>  deeper than they look into any other PC based technology. Java get's
>  labeled "Language of the month" or "World Wide Web fluff."

In defense of "the midrange circuit", I think that many spent the same time
and money that I did trying to learn "C" (being the "language of the month")
because the AS/400 finally supported it and none of us wanted to fall into
the "COBOL purgatory" that many of our MVS colleagues did.  Guess what?  None
of our managers _WANTED_ "C" applications on the AS/400!  "C" ended up
serving those whose management demanded they move to UNIX, and being a total
bust on the /400.

I think that many regard JAVA with reservation in light of their "C"
experience, rather than "World Wide Web fluff" or "Language of the Month".
 In truth, "C" has been around for quite some time, and only gained popular
(or, "Language of the Month") status when GUI became the catchphrase of the
day, and Midranger's everywhere looked for the best way to implement GUI.
 "C" was it.  Unfortunately, proprietary packages and screen scrapers ended
up taking over the Midrange GUI market, so all of our "C" education was for
naught.  I also know of many organizations that, being very conservative and
making the best possible choice, chose OS/2 as their client OS.
 Unfortunately, nobody bought their OS/2 products so they're averse to
investing those kinds of dollars again -- only to get burned.  "PowerBuilder"
was big at one point, but I'm back to seeing more AS/400 than "PowerBuilder"
ads in my area.

I think that the midrange market _WANTS_ JAVA to work, but it scares the hell
out of them given their past experiences...

>  The AS/400 is currently the finest business machine available in the
>  world. It has a host of feature rich applications and a large
>  installed base of faithful users. 
>  
>  But the days of the AS/400 are numbered. Why? Because it cannot
>  evolve to give consumers what they are buying. 
>  
>  We all know that they cornerstone of the AS/400 is reliability. I use
>  the pacemaker analogy all the time. When you buy a pacemaker, do you
>  want the shiniest, or the most reliable?
>  
>  But what is happening is that the market is evolving into buyers who
>  are not aware that a higher level of reliability exists. At the same
>  time, this group of buyers is funding Microsoft's development of an
>  operating system that will someday be able to compete with OS/400 on
>  it's own ground. 

Hmmmm.  Hard to argue.  Today's "computer literate" person has, indeed,
become far too comfortable with the "three fingered salute" and having to
toast a half-day's work because the OS crashed.  Part of the problem though,
has been businesses' willingness to be a co-conspirator in this useless waste
of time.  At most of my client sites, if the /400 were to go down for even 30
seconds during the day there would be _ABSOLUTE HELL_ to pay.  Yet these same
sites seem to have no problem with the network going down for 30 minutes to
THREE HOURS a day (as long as the factory floor isn't affected)!  Do these
people not realize that office personnel _AVERAGE_ $5/hour higher than their
factory counterparts in pay?  Why is this acceptable, while factory outages
aren't?  The factory cannot run without their support personnel any more than
they can without their shop floor devices!

>  IBM has tried to push back by trying to convince application
>  developers to create AS/400 based applications with GUI front ends.
>  But this is a losing battle. 
>  
>  First, midrange developers are very slow to invest in new
>  technologies. They tend to ignore them and hope they will go away.
>  Failing, the technologies sooner or later do fade away, becoming
>  yesterday's "Flavor of the Month" and the AS/400 sits with
>  no new look. 

I wonder why?  My former employer had $70K worth of hardware and software
lying around the office not being used because IBM convinced them to invest
in "CallPath", they also had a telephone switch that was _FAR_ more than they
needed for the same reason.  IBM "suckered" them in, and then gave no leads
or support for the application after they had made the capital investment.  I
disagree that new technologies are ignored, and successful technologies
_NEVER_ become "flavor of the month".  I think that the majority of us are
just waiting to see if JAVA becomes a successful technology.  Midrange
developers are slow to adopt new technology because it is _OUR_ systems that
run 90% of the world's business -- a trust that is _NOT_ to be taken lightly
(and the business world is notoriously conservative).  "C" failed, GUI is
failing, "CASE" (while still providing benefits) hasn't lived up to its
promise, just what _ARE_ we supposed to "jump on" next?

We "Midrangers" don't want to end up like our COBOL brethren on the ES/9000
but, short of taking advantage of new AS/400 technology, nothing else has
"shown us the money".  The 36,8/E were bad, native OS/400 applications were
good.  "C" was bad, ILE RPG was good.  CallPath and FAX/400 were bad, TCP/IP
connectivity has been good.  CISC was good, and now RISC is good.  UNIX is
still bad.  AS/400 consultants and contractors are _NOT_ better paid than
their DEC and HP counterparts because programming IBM systems is rocket
science, they are better paid because they have better overall business
knowledge than their DEC and HP counterparts.  The VAX and the HP/9K still
want to be "all things to all people", the AS/400 is _STRICTLY_ a business
machine.  If you want that engineering application, choose a UNIX box like
the RS/6K.  If you want to run a business, the AS/400 and the ES/9000 are
really the _ONLY_ viable choices in the market today.

>  Second, the applications which do get developed actually defeat the
>  purpose for which they are intended. Rather than bridge the gap from
>  green screen to multi-platform and give AS/400 apps more appeal, they
>  make the AS/400 application as unreliable as the PCs used to connect
>  to them. In order to give the AS/400 the "new look" we have to take
>  away it's primary feature.

Yes, but this speaks to the "mix and match" nature of PC hardware and
software.  I'd say that the successful GUI AS/400 developer will make their
applications work as plug-ins to a major browser, and make the browser
developer handle the hardware differences.

>  New application developers have a choice. They can build a new AS/400
>  based application and hope to somehow wedge into the market now
>  controlled by a very ingrained set of AS/400 Business Partners, or
>  develop for the lower end "shrink wrap" market. 

But that market of AS/400 BP's is easier to "break into" than that "shrink
wrap" market, IMHO.

>  The shrink wrap market is less demanding, since buyers in that market
>  are accustomed to applications which crash. A developer can ship a
>  product much earlier in the development process. Why wait until it's
>  done if you can have a revenue stream at the 90% mark?
>  
>  So new applications are going to other markets. 
>  
>  What does Java change in all this?
>  
>  Java is _the_ answer to all this. While no solution is perfect, Java
>  covers a lot of ground. One of it's chief winning points is that it
>  is not an IBM product. As a result, it is not susceptible to being
>  destroyed in the press as simply another of IBM's attempts to unify
>  platform development. 
>  
>  Instead, IBM can take a technology that is being accepted in the
>  marketplace, and utilize it fully for their own purposes. 
>  
>  Java is an object oriented language. That gives it a big boost in
>  development speed. This is something that midrange developers have
>  needed for a long time. Most of them are more afraid of change than
>  willing to embrace it, so they won't move to OO development unless
>  forced. That's okay, because Java opens the door for bringing
>  developers in who are otherwise PC developers. So, midrange coders
>  can either adapt or be replaced.

Agreed to some degree, but I think that most midrange developers with which I
am familiar will "jump on" this once it's proven.  I am, however, familiar
with a plethora of persons that fit your "don't want to change" mode.  The
skillset in this market will certainly be interesting after the Y2K "crisis"
is over...

>  The Java virtual machine is a unique opportunity for IBM and midrange
>  developers. If IBM can see to it that there is a Java workstation
>  that is as solid as a green screen terminal, then midrange developers
>  have the ability to create C/S applications which allow their
>  customers decide whether they want Win95 PC front ends, or
>  workstations. 
>  
>  It also allows developers to decided to go after all markets, shrink
>  wrap and bigger. An application developer that today wants to start
>  developing a new app can start out developing the app in a shrink
>  wrap format. If they should sell enough copies to grow enough, it
>  might be that they would want to customize that app to deal with an
>  enterprise and expand. 
>  
>  Java still needs growth. The Java workstation still needs
>  enhancement. IBM still needs to entice enterprise developers to move
>  applications to Java. 
>  
>  So, you see IBM spending $200,000,000 per year on developing Java
>  tools and products. The Java workstation is one of those products
>  getting a lot of attention. VisualAge for Java, another. Plus, the
>  San Francisco project add incentive. 

The sad thing is, $200M is a small portion of the annual budget.  It's like
AIDS vs. Breast Cancer research -- why spend an exponential amount on AIDS
research (a disease caused by personal choice) over breast cancer (which is
nobody's fault and affects _FAR_ more people)?  I'd rather see IBM spend a
_LOT_ more money on JAVA than I would see them spend _ANY_ more money on ILE
RPG...

>  Darn, here it is late at night again. 
>  
>  Any comments?

See above ;-).

Ducking!!!!

Dean Asmussen
Enterprise Systems Consulting, Inc.
Fuquay-Varina, NC  USA
E-Mail:  DAsmussen@aol.com

"Income tax returns are the most imaginative fiction being written today." --
Herman Wouk
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to "JAVA400-L@midrange.com".
| To unsubscribe from this list send email to MAJORDOMO@midrange.com
|    and specify 'unsubscribe JAVA400-L' in the body of your message.
| Questions should be directed to the list owner/operator: david@midrange.com
+---