Transcript Digital Dilemmas Introduction

Digital Dilemmas Introduction
Stanford University FSP
CS99R, Fall 2001
Armando Fox, Emre Kiciman,
and a cast of many
© 2000-2001 Armando Fox
Enrollment and Logistics

Waitlist posted tonight: www.stanford.edu/class/cs99r

If you don’t intend to enroll please let us know so we can remove
you from the enrollment/wait list!

If you do intend to enroll and your name is below, you must
email Emre ([email protected]) before Friday AM 9/28
Chris Anderson
Melissa Burns
Julianne Cuellar
Peter Deng
Jacqueline Dozier
Matthew Henick
Wendra Liang
Nicholas Lukens
Karan Mahajan
Chelsea Nicholls
Marisa Nicopoulos
James Rodgers
JP Schnapper-Casteras
Yuka Teraguchi
Weisong Toh
Anthea Tjuanakis
Noah Veltman
Melissa Wong
Organizational stuff



What to expect today

Logistics: readings, email, Web, etc.

No-count, group-interactive intro quiz and Fun Facts™

Five millennia of computing history highlights, in a nutshell
What to expect rest of quarter

Technology and society discussions

Great guest speakers

Interdisciplinary crowd
What not to expect

A lot of technical details…I will pass over these in favor of the big
picture

Discussion in which the focus is the technology itself
Logistics



Communication between you and us:

Subscribe to the mailing list! (instructions on course web site)

Course web site: http://www.stanford.edu/class/cs99r

Slides will be online
Readings

Books: Cyganski (at Bookstore); Lessig (we have some copies)

Course reader: available at Bookstore

Everything else on Web
Grading

Participation and discussion, 50%

Projects, 50%

No exams
Intro quiz and Fun Facts
“Every city will want to have one”

What am I?
More fun facts
Original purpose: recording/capturing
important live events and performances

What am I?

Of note… what was not its original purpose?
More fun facts: No-Count Intro Quiz

Fill in the blanks:

The Web as we know it has existed since ____

The Internet has existed since about ____

The basic ideas of reliable, point-to-point communication
over a large distance have been well known since _____

Most technological advances in communication and
computing have come from/been funded by ______
Don’t believe me?

Charles Babbage and Ada Lovelace, late 1700’s, the
Analytical Engine

Claude Chappe, 1799-1850, optical telegraph system

J. Presper Eckert and Charles Mauchly, ENIAC (one of the
earliest digital electronic computers), mid-1940’s

COLOSSUS, British-built early supercomputer, early
1940’s

Paul Baran and Donald Davies, early 1960’s, packet
switching

Seymour Cray & others, mid-1970’s, early
supercomputers
Some “Everyday” Technologies
What was the originally envisioned purpose? How are they
used now?

Telephone

Copier

Internet/Web

“Phonograph” [sic]
What have we learned, if anything?

A milestone in the lifetime of an invention: people start
using it in ways the creators didn’t (couldn’t?) envision.

The network externality effect: [Metcalfe’s Law]
If nobody has one, it’s useless.
If everyone has one, it’s indispensable.
More fun facts: Legal/Social Trivia

What section(s) of what legal document(s)
guarantee Americans the right to privacy?
Yet more fun facts…

What is the primary purpose of U.S. patent and
copyright law?
Now what have we learned?

(Your turn to answer.)
Preview of What’s Coming Up



Rest of today, part of next time

Tour of computing history and the forces that have driven it

Whirlwind technical introduction to how the Internet works
Next 2-3 weeks

Technical and Legal vocabulary for intellectual property

Three case studies in progress: DeCSS, Napster, Sklyarov
An assignment

Start reading (inhaling?) Cyganski

Email me with 1 thing you’d like to see covered in the “technical
intro” , or generally in the course

Optional: email me a cool travel photo
Condensed History of Computing
“Five Millennia In Fifty Minutes Or Less”

Prehistory

Electro-mechanical computers (1800’s-1930’s)

Electronic vacuum-tube computers (1940’s-1960’s)

Integrated circuit (silicon chip) computers (1970’spresent)

Communications history

If all goes as planned, you’ll see some of these machines
during our Field Trip near end of quarter
Prehistory: Numbers


The Need to Count

The Abacus

Roman Numerals

An Innovation: Numeration Systems
Turning hard tasks into easy ones: Napier’s Bones
Babbage and Lovelace

Babbage’s Analytical Engine

First Government-sponsored computer research project

British military wanted it for ballistics calculations

Genesis of the stored program concept

Never finished!

Ada Lovelace, the Mother of Programming

Saw the Analytical Engine’s potential for processing
arbitrary symbols encoded as numbers
Electro-Mechanical Computers


Hollerith and the CTR Company

Early American entrepreneur & Govt.
contractor

Punch card tabulating machines for US
Census

First high-tech startup to make its founder a
millionaire
CTR evolved into IBM under Thomas J.
Watson, Sr.
ENIAC

John Mauchly and J. Presper Eckert, Moore School of
Engineering, Univ. of Pennsylvania

“Electronic Numerical Integrator And Calculator”

WW II Contract for US Air Force

Direct ancestor to modern computers

Weighs in at 30 tons
Commercialization:

First UNIVAC delivered to Census
Bureau in 1951; in service till
1963

Instant celebrity predicting outcome
of Presidential election in 1952

But Eckert & Mauchly weren’t
businessmen…

“I have had very bad experiences
with [patent attorneys] over the
years.”
Transistors and Integrated Circuits


Bardeen, Brattain & Shockley at Bell Labs

Shockley Semiconductor: the first in Silicon Valley

Replaced vacuum tube technology

Cheaper, cooler, more reliable, faster
Integrated circuit chips

Transistors and other elements in a “silicon
sandwich” on a wafer

Bob Noyce & Gordon Moore found Intel in
1969
Microprocessors & Moore’s Law
Moore’s Law: Transistors = K* 2(N-1968)


Ted Hoff: the first microprocessor
(Intel 4004, 1971)

A “computer on a chip” -- just add
memory!

Volume production

Intel 4004, 8008, 8080, 8086, 80286,
80386, i486, Pentium, Pentium Pro,
Pentium II…
Innovation: high integration
For comparison...

CPU actual sizes
(Diagram © 2000 Intel
Corp., used without
permission)

Disk sizes:

Check out the display
case in Gates lobby

Hint: state of the art
microdisk is about
this size
“Order of Magnitude effect”


Microprocessors were a catalyst!

1968: 30,000 computers worldwide

Today: > 40 million sold per year
Allowed discontinuous leap in what
could be created affordably

1979: Steve Jobs sees demo of Xerox Alto
prototype at Xerox PARC

Q: What else was invented at PARC or by
PARC alumni?

1984: Macintosh

1987-present: MS Windows

1995-present: portables
History of Distnace Communications

Optical Communication

Electrical Telegraphy and Telephony

Wireless Telegraphy and Radio

The Internet

The order-of-magnitude effect, again
Fire Beacons and Alphabetic Codes
150 BC: Polybius, a Greek historian, documents first known
system for transmitting arbitrary messages
1
2
3
4
5
1 2 3
A B C
F G H
L M N
Q R S
V W X
4
D
IJ
O
T
Y
5
E
K
P
U
Z
• Major innovations: “instantaneous” transmission and
fully alphabetic codes
Napoleon’s Secret Weapon

Claude Chappe, 1763-1805: The Optical
Telegraph

Emergence of a Network



1799: Napoleon seizes power: “Paris is
quiet, and the good citizens are content.”
1814: Extends from Paris to Belgium, Italy
1853: 3000 miles, 556 stations
Early Defense
Contractor
Scientific Advances, 18th-19th c.


Relationship between electricity and magnetism

Oersted (Copenhagen): demonstrated electricity’s ability to deflect a
needle

1831, Faraday (Royal Institution, London): demonstrated
electromagnetic induction, the basis of electric motors

1880’s: James Clerk Maxwell develops electromagnetic “field
equations”, Heinrich Hertz demonstrates EM wave propagation
experimentally
Innovations enabled…

Telegraph (Wheatstone & Cook, 1830’s; Morse Code, 1837)

Transatlantic telegraph cable (1858)

Telephone (Alexander Graham Bell, 1876)

Radio (Guglielmo Marconi, 1895)
Packet Switching
Paul Baran & Donald Davies

Early 1960s: New approaches for
survivable comms systems

packet switching, decentralized
architecture

1967: ARPAnet Interface Message Processors (IMP’s)
connect computers at UCLA, SRI, UCB, UofU via
leased telephone lines

1973-75: Internetworking via common protocols

1980: Ethernet invented by Metcalfe at Xerox PARC

1981: Berkeley UNIX, free TCP/IP

1990s: NSF privatizes NSFnet

1993-4: World Wide Web developed (by whom? why?)
What are the important innovations?

Alphabetic code vs. word-based code


Any message can be represented
Digital vs. analog

The ability to make perfect copies

Non-degradation of messages via error correction codes: the
ability to repair defective copies

A way of thinking about point-to-point communication

Packet switching is cheaper and more robust than circuit
switching

Any message can be encoded
“Order-of-magnitude” effect again

An order of magnitude (10x or more) change in a
technology can cause a qualitative change in the impact
of that technology.

Example: computer speed


1971: 108,000 cycles per second
2,300 transistors

2000: 1,000,000,000 cycles per second
28,000,000 transistors (a factor of 10K in both)
Why is this relevant?

Hint: CD-quality audio is 88,000 audio samples/sec

Hint: decompressing and converting audio is hard work
Info source: CPU InfoCenter, Chipgeek.com, Intel Museum
Another example: Storage


Hard disk evolution (largely by IBM)

Hard disk invented in 1956; modern “Winchester” design ~1965

1956: 50 platters, 24” each, 5 MB total, $10K/MB

2001: 2-5 platters?, 2.5”, 30,000-80,000 MB total, < 1 cent/MB

A factor of 100,000 in cost per storage, 5,000 in size, 10,000 in
gross capacity
Why is this relevant?

Hint: audio takes up a lot of space

“Raw” audio: about 88,000 bytes per second of music; about 15
million bytes (Megabytes) for a 3-minute song

Compressed audio using MP3: about 1.5 to 2 millilon bytes for a
3-minute song (a factor of 10 smaller)
Last example: Network

Speed of transmitting data from machine to machine

1960’s-70’s: Leased lines, 56,000 bits per second (today’s
dialups!)

Today: 1,000,000,000 bits per second locally
up to 10,000,000 bits per second metropolitan-area

Why is this relevant? …I think you get the idea

The moral: Encoding audio, compressing it, transmitting
it over the network, storing it, and playing it back were
barely conceivable just a couple of decades ago.

Networks were too slow to make transmission practical

Computers were too slow to encode/decode and playback

Disks were too small to make storage realistic

Everything was way too expensive
Today’s Travel Photo: Where Am I?