The Hidden Language of
Computer Hardware and Software
Charles Petzold
PUBLISHED BY
Microsoft Press
A Division of Microsoft Corporation
One Microsoft Way
Redmond, Washington 98052-6399
Copyright © 2000 by Charles Petzold
All rights reserved. No part of the contents of this book may be reproduced or
transmitted in any form or by any means without the written permission of the
publisher.
Library of Congress Cataloging-in-Publication Data
Petzold, Charles, 1953–
Code / Charles Petzold.
p. cm.
978-0-735-69260-2
ISBN
1. Computer programming. 2. Coding theory. I. Title.
-- ISBN 0-7356-1131-9 (paperback)
QA76.6 .P495 1999
005.7'2 21--dc21
99-040198
Printed and bound in the United States of America.
1 2 3 4 5 6 7 8 9 QWM 5 4 3 2 1 0
Distributed in Canada by Penguin Books Canada Limited.
A CIP catalogue record for this book is available from the British Library.
Microsoft Press books are available through booksellers and distributors worldwide.
For further information about international editions, contact your local Microsoft
Corporation office or contact Microsoft Press International directly at fax
(425) 936-7329. Visit our Web site at mspress.microsoft.com. Send comments
to mspinput@microsoft.com.
Macintosh is a registered trademark of Apple Computer, Inc. Microsoft, MS-DOS, and
Windows are either registered trademarks or trademarks of Microsoft Corpo-
ration in the United States and/or other countries. Other product and company
names mentioned herein may be the trademarks of their respective owners.
Images of Charles Babbage, George Boole, Louis Braille, Herman Hollerith,
Samuel Morse, and John von Neumann appear courtesy of Corbis Images and were
modified for this book by Joel Panchot. The January 1975 cover of Popular
Electronics is reprinted by permission of Ziff-Davis and the Ziff family. All other
illustrations in the book were produced by Joel Panchot.
Unless otherwise noted, the example companies, organizations, products, people,
and events depicted herein are fictitious. No association with any real company,
organization, product, person, or event is intended or should be inferred.
Acquisitions Editor: Ben Ryan
Project Editor: Kathleen Atkins
Technical Editor: Jim Fuchs
Contents
Preface
Best Friends
Codes and Combinations
Braille and Binary Codes
Anatomy of a Flashlight
Seeing Around Corners
Telegraphs and Relays
Our Ten Digits
Alternatives to Ten
Bit by Bit by Bit
Logic and Switches
Gates (Not Bill)
A Binary Adding Machine
Chapter One
Chapter Two
Chapter Three
Chapter Four
Chapter Five
Chapter Six
Chapter Seven
Chapter Eight
Chapter Nine
Chapter Ten
Chapter Eleven
Chapter Twelve
Chapter Thirteen
But What About Subtraction?
Chapter Fourteen
Feedback and Flip-Flops
Chapter Fifteen
Chapter Sixteen
Bytes and Hex
An Assemblage of Memory
Chapter Seventeen
Automation
Chapter Eighteen
From Abaci to Chips
Chapter Nineteen
Two Classic Microprocessors
Chapter Twenty
ASCII and a Cast of Characters
Chapter Twenty-One
Get on the Bus
Chapter Twenty-Two
The Operating System
Chapter Twenty-Three
Fixed Point, Floating Point
Chapter Twenty-Four
Languages High and Low
Chapter Twenty-Five
The Graphical Revolution
Acknowledgments
Index
iv
3
9
15
22
32
40
47
54
69
86
102
131
143
155
180
190
206
238
260
286
301
320
335
349
364
383
385
iii
Preface to the
Paperback Edition
ode rattled around in my head for about a decade before I started
writing it. As I was contemplating Code and then writing it, and even
after the book was published, people would ask me, “What’s the
C
book about?”
I was always reluctant to answer this question. I’d mumble something
about “a unique journey through the evolution of the digital technologies
that define the modern age” and hope that would be sufficient.
But finally I had to admit it: “Code is a book about how computers work.”
As I feared, the reactions weren’t favorable. “Oh, I have a book like that,”
some people would say, to which my immediate response was, “No, no, no,
you don’t have a book like this one.” I still think that’s true. Code is not like
other how-computers-work books. It doesn’t have big color illustrations of
disk drives with arrows showing how the data sweeps into the computer.
Code has no drawings of trains carrying a cargo of zeros and ones.
Metaphors and similes are wonderful literary devices but they do nothing
but obscure the beauty of technology.
The other comment I heard was, “People don’t want to know how
computers work.” And this I’m sure is true. I personally happen to enjoy
learning how things work. But I also like to choose which things I learn about
and which I do not. I’d be hard pressed to explain how my refrigerator works,
for example.
Yet I often hear people ask questions that reveal a need to know something
about the inner workings of personal computers. One such common question
is, “What’s the difference between storage and memory?”
That’s certainly a critical question. The marketing of personal computers
is based on such concepts. Even novice users are expected to know how many
megas of the one thing and gigas of the other thing will be necessary for their
particular applications. Novice users are also expected to master the concept
of the computer “file” and to visualize how files are loaded from storage into
memory and saved from memory back to storage.
The storage-and-memory question is usually answered with an analogy:
“Memory is like the surface of your desk and storage is like the filing
cabinet.” That’s not a bad answer as far as it goes. But I find it quite
unsatisfactory. It makes it sound as if computer architecture were patterned
after an office. The truth is that the distinction between memory and storage
is an artificial one and exists solely because we don’t have a single storage
medium that is both fast and vast as well as nonvolatile. What we know
today as “von Neumann architecture”—the dominant computer architecture
for over 50 years—is a direct result of this technical deficiency.
Here’s another question that someone once asked me: “Why can’t you run
Macintosh programs under Windows?” My mouth opened to begin an
answer when I realized that it involved many more technical issues than I’m
sure my questioner was prepared to deal with in one sitting.
I want Code to be a book that makes you understand these things, not
in some abstract way, but with a depth that just might even rival that of
electrical engineers and programmers. I also hope that you might recognize
the computer to be one of the crowning achievements of twentieth century
technology and appreciate it as a beautiful thing in itself without metaphors
and similes getting in the way.
Computers are constructed in a hierarchy, from transistors down at the
bottom to the information displayed on our computer screens at the top.
Moving up each level in the hierarchy—which is how Code is structured—
is probably not as hard as most people might think. There is certainly a lot
going on inside the modern computer, but it is a lot of very common and
simple operations.
Although computers today are more complex than the computers of 25
years or 50 years ago, they are still fundamentally the same. That’s what’s
so great about studying the history of technology: The further back in time
you go, the simpler the technologies become. Thus it’s possible to reach a
point where it all makes relatively easy sense.
In Code, I went as far back as I could go. Astonishingly, I found that I
could go back into the nineteenth century and use early telegraph equipment
to show how computers are built. In theory at least, everything in the first
17 chapters of Code can be built entirely using simple electrical devices that
have been around for over a century.
This use of antique technology gives Code a fairly nostalgic feel, I think.
Code is a book that could never be titled The Faster New Faster Thing or
Business @ the Speed of a Digital Nervous System. The “bit” isn’t defined
until page 68; “byte” isn’t defined until page 180. I don’t mention transistors
until page 142, and that’s only in passing.
So, while Code goes fairly deep into the workings of the computer (few
other books show how computer processors actually work, for example),
the pace is fairly relaxed. Despite the depth, I tried to make the trip as
comfortable as possible.
But without little drawings of trains carrying a cargo of zeros and ones.
Charles Petzold
August 16, 2000
code (kod) ...
3.a. A system of signals used to represent letters or numbers in
transmitting messages.
b. A system of symbols, letters, or words given certain
arbitrary meanings, used for transmitting messages
requiring secrecy or brevity.
4. A system of symbols and rules used to represent instructions
to a computer…
— The American Heritage Dictionary of the English Language
Chapter One
Best Friends
Y
ou’re 10 years old. Your best friend lives across the street. In fact,
the windows of your bedrooms face each other. Every night, after
your parents have declared bedtime at the usual indecently early
hour, you still need to exchange thoughts, observations, secrets, gossip, jokes,
and dreams. No one can blame you. After all, the impulse to communicate
is one of the most human of traits.
While the lights are still on in your bedrooms, you and your best friend
can wave to each other from the windows and, using broad gestures and
rudimentary body language, convey a thought or two. But sophisticated
transactions seem difficult. And once the parents have decreed “Lights out!”
the situation seems hopeless.
How to communicate? The telephone perhaps? Do you have a telephone
in your room at the age of 10? Even so, wherever the phone is you’ll be
overheard. If your family personal computer is hooked into a phone line, it
might offer soundless help, but again, it’s not in your room.
What you and your best friend do own, however, are flashlights. Everyone
knows that flashlights were invented to let kids read books under the bed
covers; flashlights also seem perfect for the job of communicating after dark.
They’re certainly quiet enough, and the light is highly directional and prob-
ably won’t seep out under the bedroom door to alert your suspicious folks.
Can flashlights be made to speak? It’s certainly worth a try. You learned
how to write letters and words on paper in first grade, so transferring that
knowledge to the flashlight seems reasonable. All you have to do is stand
at your window and draw the letters with light. For an O, you turn on the
flashlight, sweep a circle in the air, and turn off the switch. For an I, you make
a vertical stroke. But, as you discover quickly, this method simply doesn’t
work. As you watch your friend’s flashlight making swoops and lines in the
3
4
Chapter One
air, you find that it’s too hard to assemble the multiple strokes together in
your head. These swirls and slashes of light are not precise enough.
You once saw a movie in which a couple of sailors signaled to each other
across the sea with blinking lights. In another movie, a spy wiggled a mir-
ror to reflect the sunlight into a room where another spy lay captive. Maybe
that’s the solution. So you first devise a simple technique. Each letter of the
alphabet corresponds to a series of flashlight blinks. An A is 1 blink, a B is
2 blinks, a C is 3 blinks, and so on to 26 blinks for Z. The word BAD is 2
blinks, 1 blink, and 4 blinks with little pauses between the letters so you
won’t mistake the 7 blinks for a G. You’ll pause a bit longer between words.
This seems promising. The good news is that you no longer have to wave
the flashlight in the air; all you have to do is point and click. The bad news
is that one of the first messages you try to send (“How are you?”) turns
out to require a grand total of 131 blinks of light! Moreover, you forgot
about punctuation, so you don’t know how many blinks correspond to a
question mark.
But you’re close. Surely, you think, somebody must have faced this prob-
lem before, and you’re absolutely right. With daylight and a trip to the li-
brary for research, you discover a marvelous invention known as Morse code.
It’s exactly what you’ve been looking for, even though you must now relearn
how to “write” all the letters of the alphabet.
Here’s the difference: In the system you invented, every letter of the al-
phabet is a certain number of blinks, from 1 blink for A to 26 blinks for Z.
In Morse code, you have two kinds of blinks—short blinks and long blinks.
This makes Morse code more complicated, of course, but in actual use it
turns out to be much more efficient. The sentence “How are you?” now
requires only 32 blinks (some short, some long) rather than 131, and that’s
including a code for the question mark.
When discussing how Morse code works, people don’t talk about “short
blinks” and “long blinks.” Instead, they refer to “dots” and “dashes” be-
cause that’s a convenient way of showing the codes on the printed page. In
Morse code, every letter of the alphabet corresponds to a short series of
dots and dashes, as you can see in the following table.
A
A
J J
B b
K K
S
T
S
T
C C
L L
U U
D D
M M
V V
E
E
N
N
W W
F F
O O
X X
G G
P P
Y Y
H
I
H
I
Q Q
Z Z
R R