MacHack 1994/MacHack 1994.toast/MacHack™ 1987-1994/MacHack™ '92/Talk & Papers ’92/Tom Pittman (Pass the Torch)/Tom Pittman

Text File | 1992-05-05 | 12.0 KB | 193 lines | [TEXT/QED1]

Pass the Torch: Enabling the Next Generation of Hackers
by Tom Pittman

I used to think of programming as "beating back the advance of entropy" by
injecting information and order into an otherwise entropic system. Then I
discovered a different sense in which the complexity I added was itself an
expression of entropy: as we pour energy and information into the technology
that drives modern society, it thereby becomes more chaotic and less
predictable.

As I write this, I am involved in the software redesign of a product, the
original requirements of which proved so complex that the lead programmer had
to eliminate half the features just to get it working. Most of those feature
requirements were added back in when I took it on, and I also find the
complexity overwhelming. It is interesting to observe that the same creeping
feature creature that bit off Jim's head is now poised, jaws agape, over mine -
never mind that I have a lot more experience and professional skills than he
ever did.

On my first Mac, after upgrading it to 512K, I ran a lean 70K System file on a
RAM disk and experienced an amazing leap in productivity over all previous
computer systems in my career, large or small. Today I do even better with a
rather corpulent 600K System file (not counting fonts), but I dread the day I
must switch to a 1.6+ megabyte System 7. "Pink," the developing system software
Apple contributed to its IBM alliance, reportedly has already over a million
lines of code and 10,000 objects.

My first public software product was a Tiny Basic interpreter that fit in a 2K
ROM, and I wrote it in less than a month. Now I sell a 600K HyperTalk compiler
that with far better tools took some three years to get right.

The first Apple computer was built in a garage from a handful of logic gates on
a 10" board back when most microprocessors sold for $400. Today's high-end
Apple computer is still built on about the same size board with the same number
of chips (and the CPU still costs something like $400), but we are talking
about large-scale custom integrated circuits surface-mounted on four-layer
boards with wiring traces so tiny they cannot be done by hand.

My father's first car, a Ford Model A, was simple enough in construction that
he could do all his own maintenance with $20 worth of tools. I similarly
maintained my first vehicle with tools that I already had. The car I bought
this year, however, is so complex that the shop manual alone costs over $40; a
computer many times the price of the car is required for a simple tune-up.

Part of my college tuition was paid from photography profits. I was no Ansel
Adams, but I had a well-equipped darkroom capable of good color wedding
photographs. Today all that is replaced in the industry by high-tech color
print machines and 24-bit color editing software on Macintosh computers too
expensive for my budget.

Everywhere you look, western culture is replacing simple methods with complex
ones, manual efforts with computer technology. And everywhere the
sophistication required to master this technology is zooming out of sight. Who
is going to do it?

In America, the world leader in technological innovation for 100 years, advance
comes as a curious mixture of big business professionalism and amateur
tinkering ("hacking"). The aircraft industry was started as a hobby in a
bicycle shop, but now it is the exclusive domain of a half-dozen huge aerospace
companies. Computers were invented in research labs funded by large government
grants, but they did not begin to remake modern technology until
microprocessor-based computer systems became affordable to hackers. The
spreadsheet, probably the single most significant software innovation of all
time, was invented by a student on an Apple II. A culture of computer phreaks
growing up on personal computers carried innovative ideas into industry, where
it's hard to find an electronic gadget any more that is not controlled by a
microprocessor.

Software engineering technology and new programming tools are increasing by
leaps and bounds, but the required complexity in new computer-based products
goes up exponentially. Frederick Brooks wrote in IEEE Computer magazine a few
years back that there are no silver bullets to solve the software crisis. What
he meant is that we cannot expect any magical cures to the problem. And he is
right. Engineering in general, and software engineering in particular, can only
extend known technology in predictable ways. But the personal computer industry
was built by hackers, not engineers. Only science fiction writers - and readers
- dared to predict what has come about in two decades. To the rest of the world
it was a big surprise, the magical silver bullet. True magic is not
predictable.

There is a fundamental difference between engineering and hacking. An engineer
tries to reduce risks, to build on and improve existing technology to
accomplish known goals in controlled steps. The hacker, like the proverbial
fool jumping in where angels fear to tread, ignoring probable failure and
potential consequences, optimistically sets out to do the impossible - and
often succeeds. The engineer is like the expert skiers I watched in the winter
Olympics a few years back: in full control of every muscle, every turn
perfectly executed. The hacker is like the graceless fellow who stormed down
the hill, skipping and sliding on every turn: you were sure he would spill out
like so many before him who looked equally out of control. This guy somehow did
not take a fall on that particular trip down the slope - and he won the gold
medal. The engineer is a paid professional; the hacker works for love on his
own resources. The engineer is trained; the hacker is inspired.

How, then, do we inspire the hackers and breed technological winners? If we
knew that, we could apply sound engineering techniques to develop them, and
they would be engineers, not hackers. A hacker is an artist, and like the
artists squandering controversial government grants in the news today, most of
their output is worthless. But you cannot know a priori which hackers are going
to produce and which will fail. Indeed, most will fail to deliver any positive
benefit to society. Some may even fall into destructive potholes producing
obscenity and computer viruses. But we still encourage artists in general, and
we should still encourage hacking.

The best inspiration to excellence and artistry is excellence and artistry. The
best inspiration to excellent hacking is likewise excellent hacking. In the
"good old days" when professions were, for lack of understanding, more art than
technology, the next generation of professionals were trained by a process
known as apprenticeship. Today we call the same process mentoring, but it is
somewhat less formal. Successful older hackers should seek out and encourage
promising young hackers to help them develop the attitudes and skills that will
carry the tradition of innovation and excellence forward. Young, ambitious
comers could do worse than to listen once in a while to the old fogies.
Obviously this works better in a stable craft like stained glass or music, but
there is room for mentoring in our profession also. Michael Hogan, my
supervisor and mentor at the first full-time job I held, instilled in me his
(and now my) philosophy of programming.

Another way for the excellence to pass on to the next generation is by means of
inspired and inspiring program code. I learned the Macintosh mostly by reading
other people's code through a MacsBug disassembly window. Bill Atkinson and
Andy Hertzfeld and the whole Mac team have been an inspiration to us by virtue
of the excellent software they released to the world.

As important as mentoring and inspiration-by-example may appear to us, however,
all of the people I asked indicated that the dominant enabling influence for
them was the availability of adequate tools. The two most important tools are,
of course, a computer and a programming language. It almost does not matter
which computer or what language, as long as they are available and offer
opportunity to do "wonderful" things. "Available" in this context means that
they are essentially free, that is, (at least initially) subsidized by other
purposes. The MIT hackers in Steven Levy's book used an academic computer that
was free after hours. Apple Computer got its start because MOS Technology came
out with a $20 microprocessor when the average CPU chip sold for $300. A year
earlier MITS made a similar splash by selling a whole computer (kit) for $400,
about the same price as the CPU in it sold from Intel.

Peter Nagel tells me, "I had never written a line of code of any sort (except
possibly an Excel macro and small Red Ryder scripts) before I got my hands on
HyperCard shortly after its original release. At that time, I was between jobs
and sat down to write a stack to handle timekeeping and billing chores that I
still use in my law practice. Within several months, I bought a copy of
Lightspeed Pascal, read its manual and a couple of other books on Pascal, and
began writing short little XCMDs to improve my billing stacks." Now he is
selling MasterScript, a serious commercial programming tool, to other
programmers. An earlier version of this same software substantially influenced
some of the features added to HyperCard 2.0. Peter concludes, "The result of
all this is that I am deeply indebted to HyperCard for having introduced me to
fundamental programming concepts and techniques."

HyperCard turns out to have a disproportionate influence in opening up computer
programming to people. Swami Gurupremananda, another nascent hacker not quite
so far along, even went so far as to buy his first Macintosh computer just so
he could run HyperCard. "With my increasing knowledge of HyperCard," he said,
"I, who have not taken even one course in programming, started making stacks to
run our correspondence course. Now all our work is done in HyperCard. The IBMs
were forced into early retirement!" Most people, however, are more like Susan
Leschr, who told me, "The fact that HyperCard came with my computer (1987) had
a lot to do with it." Brian Molyneaux, the CompileIt product manager at Heizer
Software, reports that a number of our customers use CompileIt as a stepping
stone along the path from HyperCard to conventional programming languages such
as Pascal and C.

Bill Atkinson reportedly offered HyperCard to Apple on the condition that it
ship free in every Macintosh. That is the availability that develops unforeseen
hackers and inspires new programmers. Present Apple/Claris policy appears to be
based on seeing the vast majority of Mac owners who have not yet discovered
HyperCard - or simply lack the hacker mentality - and a few enlightened
companies using it in-house for prototyping and other "throw-away" programs.
This ignores the fact that HyperCard performs a valuable social service if and
only if it is available (read "free" again) to a large number of people, most
of whom will never take advantage of it. Bill had the right vision.

Bill Atkinson and Andy Hertzfeld have gone on to other ventures, which I assume
and hope will benefit us as much as HyperCard and the Mac did. We must not,
however, abdicate our own responsibility to benefit society and leave it all to
such luminaries. Peter Nagel is producing software tools to help other
programmers along the path he has come. Swami Gurupremananda gets online on
CompuServe and elsewhere, answering questions from beginners not far behind
himself. Each of us can do our part to help new programmers with advice and
tools.

I conclude this essay with a challenge: How were you helped to become what you
are? What have you done to pass that help along and repay your debt to society?
Our only hope of rescue from exponential complexity is likely to come from some
new hack not yet thought of. Perhaps you will invent it; perhaps you can only
help some other hacker along the way, and he will invent it. Who can know?
Quoting the medieval monastic Bernard, Isaac Newton said, "If I have seen
farther than others, it is because I stood on the shoulders of giants." You
could be the giant under the feet of the next generation of Newtons.