Newsgroups: comp.lang.apl
Path: watmath!watserv2.uwaterloo.ca!watserv1!FZC@CU.NIH.GOV
From: "Dick Holt" <FZC@CU.NIH.GOV>
Subject:  Kaizen Strategy for APL Education
Message-ID: <Bwu956.7z1@watserv1.uwaterloo.ca>
Sender: root@watserv1.uwaterloo.ca
Reply-To: TDARCOS@MCIMAIL.COM
Organization: University of Waterloo
Date: Wed, 28 Oct 1992 15:44:26 GMT

Kaizen Strategy for APL Education          (about 1500 words)

A Survival Loop for APL

If APL is to survive, more people must be able to quickly learn
to use it in their work or business.  Recently, two strong new
APLs have become available that are fast, cheap or free, and that
have big workspaces.  To take full advantage of these new APLs,
the APL community needs to rethink its strategies, especially for
APL in education.

An emerging theme of these new APLs is that they are continuously
and incrementally improved.  The Japanese term for this process
is "kaizen".  Kaizen explains, in part, the hammering that Japan
is delivering to the auto industry world-wide.

Kaizen develops product innovation at the same time that it
generates cash flow to pay for further innovation.  Kaizen
strategy is a quick feedback loop from innovation, to market, and
back to innovation again.  This is perhaps the one loop that
APLers should learn to love.

The kaizen loop is apparent in IBM's steady and cumulative
improvement of TryAPL2 -- six enhancements in three years.  Also,
the once-orphaned Sharp APL/PC has been jump-started by Iverson's
feature-packed APLs for both 386s and lesser machines.  And, the
interactive on-screen APL lessons from the Capital PC User Group
(CPCUG) have been incrementally improved in multiple ways since
1991, in two generations beyond Z.V. Jizba's pioneering effort.

Schools or the Workplace?

Kaizen strategy throws down the gauntlet to the APL community:
people who are serious about the future of APL must concentrate
their efforts in the workplace, not in the schools.  It's the
workplace that provides the feedback to spur technological
innovation, and it's the workplace that provides the cash flow to
pay for it.  In contrast, most US schools are non-productive
bureaucracies that resist innovation, are technologically
backwards, and whose students place at or near the bottom in
international comparisons of math and science skills.  Even if
you believe that schools are effective places to learn, there's
ample empirical evidence that most schools have low or zero
interest in using or teaching APL.

In the global information economy where APL competes, learning is
a lifelong activity.  Thus, most people today learn far more on
the job than they do in school -- they have no choice.  Product
life cycles are short, and technology changes fast.  Unless the
time between learning APL and applying it at work is short,
technology advances rapidly, and the learner is quickly out of
date.


Salmon, the Lottery, and Wishful Thinking

George Bernard Shaw once compared photographers to spawning
salmon: they take a million pictures in the hopes that one will
turn out well.  Promoting APL in schools is no different.  Like
spawning salmon, this strategy is tantamount to the wishful
thinking of a lottery player -- but for a smaller prize, at
dreadful odds, at some unknown time in the future.  What's worse,
those who hope to revive APL by promoting it in schools have no
way to measure the effectiveness of their strategy.  In contrast,
the effectiveness of a kaizen strategy is quickly, and
mercilessly, measured by sales.

Academicians may demur, citing the enthusiasm that students show
when introduced to APL.  But this is not so much a tribute to APL
as it is to the innate curiosity and energy of youth -- who can
be equally enthusiastic about quickly learning Urdu, tort law, or
the game of Go.  However, APL is subject to rapid technological
improvement and change, whereas Go isn't.

In its 1991 APL classes, the CPCUG found that only one of
its 14 learners was a student.  All others were professionals
seeking to improve their productivity at work.  Six of the 14 did
not have English as their native tongue.  This typifies today's
dynamic global market within which APL must compete.

New, Big, Fast, Cheap Material

Here's a source list of APL material, new or improved within the
past year or so:

Iverson Software: APLI386 or APLIPC, available for $US60, from:
                  Iverson Software Inc.
                  33 Major Street, Suite 466
                  Toronto, Ontario
                  Canada M5S 2K9  416-925-6096,  Fax 416-488-7559

Iverson APLs have boxed arrays, color, graphics, a full screen
editor, a fast 386 version that works in Windows, F1 help,
printer support, and imperfect but usable documentation.  APLI386
requires a 386 machine, (huge workspaces, 32 bit speed).  APLIPC
is functionally identical to APLI386, but doesn't require a 386.

IBM TryAPL2:      Free demo disk of APL2/PC, available from:
                  IBM APL Development
                  M46/D12-278B, Santa Teresa Lab
                  Box 49023, San Jose CA 95161-9023

TryAPL2 has nested arrays, color, graphics, a number of useful
workspaces, on-screen documentation, and a LJ printable user
manual.

Iverson's APLI386 and APLIPC, and to a slightly lesser extent,
TryAPL2, can be used to do real, on-the-job, work in APL.

CPCUG Lessons: Interactive on-screen APL lessons, developed by
the APLSIG of the CPCUG, are available in generic APL for the
interpreters above (and for other APLs).  CPCUG lessons are
available for $US25 postpaid world-wide by mail order from:

                  HRH Systems
                  Box 4496
                  Silver Spring, MD, 20914 USA.

CPCUG Lessons are also downloadable free (download request may be
needed) from the BBS\APL, 301-384-3672, 1200/2400/9600b, N-8-1,
24 hours (if this article is published after about mid-February
1993, the BBS\APL phone number will have been changed).

Kaizen: Continuous Incremental Improvement:

From an educational point of view, these two new APLs aren't
perfect.  For example, TryAPL2 clears the state indicator upon
errors, error trapping is undocumented, it has no file system,
and it has only the obsolete del editor.  Iverson Software
products aren't even designed to teach APL.  They're designed to
do real work.  Paradoxically, this may make Iverson material
well-suited to expanding the use of APL.  MicroSoft's QuickBASIC
isn't designed to teach BASIC.  We'll see.

From an educational point of view, the CPCUG lessons aren't
perfect either.  Because they're generic, they don't exercise
many of the powerful features of specific APLs, like nested
arrays, file systems, error trapping, or screen management.

It doesn't matter that these materials are imperfect.  What
matters is that there be in place a "virtuous circle" of kaizen,
a process whereby APL products get (1) quickly to market, (2)
tested by users, and (3) relentlessly improved as a result of
user feedback.  Anyone who has recently bought a software update
will recognize that this is a dominant strategy, almost to a
fault, in today's software industry.

For kaizen strategy to succeed in APL, the APL education infra-
structure must abandon low-performance, commercially non-viable,
APLs.  They must instead emphasize near-term market-driven high
performance APLs that can do real work on the job or in business.

And, like the rest of world commerce and industry, the APL
infrastructure must invest more heavily in electronic media --
bulletin boards, CIS, Internet, and the like -- mainly because
these are thousands of times faster and more flexible than print.
Electronic media provide user feedback and market intelligence
that can't be obtained in any other way.  They provide an
international infrastructure that print can never match.  Tele-
communications provide a way to distribute, learn, and use APL
that simply and quickly bypasses the increasingly obsolete school
system and print medium.

Kaizen strategy implies a shift of mindset for ACM/SIGAPL, for
the APL trade press, for APL conference organizers, and for APL
educators everywhere.  It implies a shift away from long-term
pedagogy, and toward near-term market pull.  It implies a shift
away from schools, and toward the marketplace.  It implies a
shift toward market-driven APLs that simultaneously meet
commercial demand and, by their innovation and by their
value-added, change the very nature of that demand.

So What?

It would be easy to dismiss kaizen as short-term incrementalism.
What can't be dismissed is its glaring economic success in world
markets.  What can't be dismissed is its strong ability to
deliver a steady stream of technologically superior products to
users who want and need high quality/performance and who are
willing to pay for it, because it produces strong results for
them in their affairs.  This is the only feature of APL that
really counts.

In short, kaizen strategy challenges the APL infrastructure to
focus its educational efforts on new APL material that:

     - is fast, with big workspaces,
     - has common (or nearly so) keyboard/interface/features,
     - can be learned in the workplace, not at schools,
     - has commercially-motivated vendor support, and
     - has customer-driven technological innovation.

The purpose of this article is to provoke discussion of the
future course of APL, especially in education.  I've exaggerated
some points to spur debate about APL's place, if any, in the
fast-changing markets of the '90s.

These are not new ideas.  James said: "Faith, if it hath not
works, is dead".  The shift toward kaizen strategy is now well
underway in some parts of the APL community.  How do you think it
can be accelerated?



