playing the

past
h i s to r y a n d n o s ta lg i a i n v i d e o g a m e s

EDITED BY

Zach Whalen
Laurie N. Taylor
Playing the Past
History and Nostalgia
in Video Games

Edited by Zach Whalen

and Laurie N. Taylor

Vanderbilt University Press • Nashville

© 2008 by Vanderbilt University Press
Nashville, Tennessee 37235
All rights reserved

Library of Congress Cataloging-in-Publication Data

Playing the past : history and nostalgia in video games /

edited by Zach Whalen and Laurie N. Taylor.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-8265-1600-8 (cloth : alk. paper)
ISBN 978-0-8265-1601-5 (pbk. : alk. paper)
1. Video games. 2. Video games—Psychological aspects.
3. Video games—Study and teaching. I. Whalen, Zach,
1979– II. Taylor, Laurie N., 1978–
GV1469.3.P483 2008
794.8—dc22
2007051878
 Contents

Preface and Acknowledgments vii

1 Playing the Past: An Introduction 1

Laurie N. Taylor and Zach Whalen

Part I. Playing in the Past

Negotiating Nostalgia and Classic Gaming
2 Why Old School Is “Cool”:
A Brief Analysis of Classic Video Game Nostalgia 19
Sean Fenty
3 Homesick for Silent Hill:
Modalities of Nostalgia in Fan Responses
to Silent Hill 4: The Room 32
Natasha Whiteman
4 Playing the Déjà-New:
“Plug it in and Play TV Games” and the
Cultural Politics of Classic Gaming 51
Matthew Thomas Payne
5 Hacks, Mods, Easter Eggs, and Fossils:
Intentionality and Digitalism in the Video Game 69
Wm. Ruffin Bailey
6 Screw the Grue:
Mediality, Metalepsis, Recapture 91
Terry Harpold
vi Playing the Past

Part II. Playing and the Past

Understanding Media History and Video Games
7 Unlimited Minutes:
Playing Games in the Palm of Your Hand 111
Sheila C. Murphy
8 Visions and Revisions of the Hollywood Golden Age
and America in the Thirties and Forties:
Prince of Persia and Crimson Skies 126
Andrew E. Jankowich
9 Toward a New Sound for Games 145
Thomas E. Gersic
10 Remembrance of Things Fast:
Conceptualizing Nostalgic-Play
in the Battlestar Galactica Video Game 164
Anna Reading and Colin Harvey

Part III. Playing with the Past

Nostalgia and Real History in Video Games
11 Just Less Than Total War:
Simulating World War II as Ludic Nostalgia 183
James Campbell
12 Performing the (Virtual) Past:
Online Character Interpretation
as Living History at Old Sturbridge Village 201
Scott Magelssen
13 Documentary Games:
Putting the Player in the Path of History 215
Tracy Fullerton
14 Of Puppets, Automatons, and Avatars:
Automating the Reader-Player
in Electronic Literature and Computer Games 239
Robert P. Fletcher

Contributors 265

Index 271
 5
Hacks, Mods, Easter Eggs, and Fossils
Intentionality and Digitalism in the Video Game

Wm. Ruffin Bailey

The term “interactive digital media” contains an often-overlooked adjec-

tive, digital. Espen Aarseth has given us a detailed study of the aesthetics
of cybertext; Nick Montfort, the textuality of interactive fiction; and Mark
J. P. Wolf, a strict review of the hardware requirements for a work to be
labeled a video game. These authors provide useful, high-level work that
introduces a young field, but only begin to provide an in-depth look at
the digital underpinnings of gaming software and digitalism’s undeniable
influence on the creation of virtual realities.
The “digital” in digital media needs to be examined to gauge its char-
acteristics’ influence upon the creation of virtual spaces. Every web page
lives on a digital host, forcing preschoolers to grandmothers to become
familiar with the highly technical standards of Universal Resource Lo-
cators (URLs) and Hypertext Markup Language (HTML) that enable
digital interpretation. Engines that drive video games are mediated by
similar digital hosts. Games are limited in size by their ability to access
their host’s physical memory. They are limited in complexity by their au-
thors’ ability to construct decision-making algorithms that approximate
the authors’ visions of virtual realities. Theorists may be able to observe
the results of these virtual realities, but in the absence of an interrogation
of the fundamentals of computer science, they will not be able to effect a
complete study of digital media and its creation.
This chapter seeks to establish a rhetorical method—of terminology
and taxonomy—by which to explore what is unique to software-based
digital media, starting with a nostalgic application of the method to the
Atari 2600 (the first truly modular home gaming console) and watching
how its workings continue to reflect and inform studies of games writ-
ten nearly two decades later. Such a methodology currently requires a

69
70 Playing in the Past

multidisciplinary cultural approach and is instrumental in grasping how

interactive digital media, here video games in particular, operate.
This study begins by reviewing Montfort’s study of interactive fiction,
which provides an accessible introduction to video game studies by fo-
cusing on the games’ use of narrative and riddle—two conventional ap-
proaches to text—before launching into code. I then review two founda-
tions of a digitalistic approach that are currently housed in the field of
computer science, the concepts of Boolean logic and memory addressing.
These prove crucial for studying the operation of video game engines and
creating a technical taxonomy for video game content. I end with a con-
sideration of several video games to demonstrate how a nostalgically in-
formed study of code can be employed in practice to shape game studies,
using examples as varied as programmers innocuously hiding graffiti of
their names in Atari 2600 games to more modern artifacts like Rockstar
Games’ “accidental” inclusion of a hidden sex game in Grand Theft Auto:
San Andreas (2004). In sum, because digital media is heavily influenced
by technical characteristics, game studies will not achieve its full potential
until considerations of computer science are more fully integrated into its
methods.

The Bias of Accessibility or Limited Approaches

Montfort’s enlightening study of interactive fiction (IF), Twisty Little
Passages, recommends approaching IF, a subset of video games, in three
ways: as narrative, as riddles, and as computer programs (14–15). The
first two have received a great deal of discussion, largely through scholars
like Montfort constructively framing games as texts. Purely by virtue of
being accessible, however, these two approaches risk garnering inordi-
nate attention within cultural studies at the unfortunate detriment of the
third.
In the field of cybertext, six years prior to Montfort, Aarseth expressed
his “wish to challenge the recurrent practice of applying the theories of
literary criticism to a new empirical field” (14). This “recurrent practice”
is one possible symptom of over privileging customary approaches—
including narratives and riddles. At the same time, scholars’ familiarity
with these approaches is one reason that Montfort’s book on interactive
fiction (a style of game with a high affinity for the common codex) is a
proven and popular introduction to the field.
Aarseth admits that “[traditional literary] approaches are useful for
establishing the legitimacy of the field [of hypertext literary theory]” (76).
Twisty Little Passages does a fine job establishing such legitimacy by dis-
 Hacks, Mods, Easter Eggs, and Fossils 71

secting IF-as-narrative and riddle. Still, both authors ultimately privilege

the experience of IF’s and cybertext’s content without giving the same
level of sustained attention to how that content was shaped by its digital
hosts, a topic crucial for the study of both IF and cybertext.
Guarding against unfairly strong remediations of old media ap-
proaches into the study of the new proves a major difficulty in many
studies, exemplified in part by John Muckelbauer’s critique of a recent
collection of essays regarding the rhetoric of film. Muckelbauer warns
against exploiting that which is unique to a medium to support conclu-
sions that were reached exclusively via historically more familiar means of
compiling evidence: “This [narrative bias] is particularly striking insofar
as discussion of things like images and sound—things that would seem to
be important elements of film’s distinctiveness—are relatively invisible in
this collection” (905). In the case of video games, this warning translates
directly, and it is this tendency to ignore games’ “distinctiveness” that lu-
dology and other multidisciplinary approaches must avoid.
The importance of truly digital approaches is evident in a section of
Twisty Little Passages where Montfort heuristically recommends initially
treating an IF work’s engine, called a parser, “as a black box that accepts
input and generates output.” He goes as far as to proscribe “making ref-
erence to a program’s specific data structures, functions, objects, and
so forth” (23). However, to complete his analysis, Montfort (himself an
author/programmer of several impressive works of IF) ultimately does
move to his third approach: looking at IF compositions as computer pro-
grams. When he compares Adventure (1976) to a successor, Zork (1982),
he lists a number of ways that Zork’s parser is superior to Adventure’s.
Montfort states that Adventure “only accepts commands of one or two
words” and that Zork’s parser understands longer directives. Adventure
requires explicitly mentioning the object a player wants to use in com-
mands, but Zork’s parser can “disambiguate” the item a user implicitly
requested to employ with certain actions (e.g., a player’s avatar might dig
with its hands if another object was not mentioned). Zork’s parser can
also understand prepositions, unlike Adventure (108–9).
Montfort’s conclusions about IF engines—that one parser is limited
to one or two words, that it never understands prepositions, or that its
content is limited to a certain size —cannot be stated with full confidence
without an in-depth, open-box understanding of the works’ parsers.1 Per-
haps Adventure anachronistically understands the sentence, “Climb over
the northern mountains, read Sam Lantinga’s blog, and then summarize
any advantages of using OpenGL over DirectX.” If one were to take Twisty
Little Passages’ initial proposal to treat game engines as black boxes until
72 Playing in the Past

every possible sentence has been tried, a task of ridiculously irrational

scope, its conclusions could not be stated in absolute terms and likewise
could not be authoritatively woven into critical works. Determinations
about maximum work lengths or the potential to display graphics could
not be made. As Montfort’s and Ian Bogost’s forthcoming Platform Studies
series’ concentration on just these issues attests, it is crucial for scholars
of digital artifacts to now move past accessible remediations and forward
toward understanding the inner workings of these artifacts’ platforms.
To analyze these inner workings requires the explicit introduction of a
number of concepts from computer science.

Digitalism
For digitalism, the first steps involve further pursuing Montfort’s third
approach to IF works: to treat them as computer programs. Commercial
video games, thus far, are digital creations.2 I argue that scholars studying
works of digital media must, as an essential position for rhetorical and
cultural studies of games, approach the content on an equal footing with
those games’ creator(s) or author(s).
Digital means a system of representation based on discrete digits
rather than any sort of continuous spectrum. What this means in prac-
tice becomes especially evident in the way computer programs make de-
cisions. The simplest flow chart that includes at least one decision step
makes the point obvious.
In each decision step, there are one or more discrete choices for exit.
Even in the most complicated piece of branching logic, entry into each
potential branch is, at its core, mediated by a simple test of true/false
(that is, Boolean) logic. Either the condition or conditions for taking the
branch are all true statements and the branch is taken, or at least one con-
dition is false and the branch is avoided. By convention, conditions that
point towards taking a branch evaluate to a Boolean value of “true,” usu-
ally represented by the digit “1.” Conditions that would stop the branch’s
execution are given a value of “false,” and are usually represented by the
digit “0.”3
In technical terms, Boolean data types that hold these values are of-
ten called switches, and these evaluations act like switches on a railroad
track. If “on,” the train of logic will travel down one track; if “off,” the
logic will proceed down another. For trains, taking a third choice is disas-
trous. In digital hardware, a third choice, much less an infinite spectrum
of choices, is—by design—impossible. This is precisely why cybertext so
 Hacks, Mods, Easter Eggs, and Fossils 73

Avator enters door

activation radius

Is it between game hours No Start police

of 7 pm and 6:30 am? alert timer.

Yes

Is the blue key in the No Activate handle jiggle

avatar‘s inventory? but locked sound.

Yes

Open door. Activate

door creak sound.

Figure 5.1. A simple flowchart

naturally gave rise to a system that, as Aarseth remarks, is “forcing the

reader to pay attention to the strategic links” between portions of text
(78). The reader is being forced to act digitally—to select one of a finite
number of branches for their progression in a work. The operation of a
cybertext interface can be reduced to nothing more than the most trivial
of covers above the Boolean logic being evaluated beneath.
Exactly like they do in the thinly veiled interfaces of most hypertexts,
Boolean evaluations or switches drive decision-making even in the most
complicated representational systems created by video games. A player’s
avatar may approach a locked door that only operates during the night
and if the avatar is carrying the appropriate key. Here there are at least
two conditions to be evaluated. If the value of the night switch is “true”
and a key inventory switch is also “true,” the engine will allow the code
to open the door to be executed. To the player, the checks in this example
happen behind the scenes. Regardless, the digital switches are still there,
evidenced by the consistent appearance of if-statements in the represen-
tation of Figure 5.1’s logic shown in Figure 5.2’s pseudo-code.
74 Playing in the Past

if (measurer.pointDist(p1,p2) < 20) {

if ( Game.currentTime() < Game.MORNING_TIME
&&
Game.currentTime() >= Game.EVENING_TIME
) {
if (inventory.key(BLUE_KEY)) {
blueDoor.open();
} else {
Game.sounds.
play(handleJiggle());
}
} else {
police.soundAlarm();
}
}

Figure 5.2. Pseudo-code for opening a door

This Boolean logic also introduces the method for creating “hacks.”
Hacks occur when an unsanctioned third-party or independent program-
mer changes a program’s code to function in non-standard but not nec-
essarily unintended ways. If a hacker could change the above example’s
logic to read, “If the night switch is ‘true’ and the key inventory flag is also
‘true’,” or “1 = 1” (that is, simply adding a trivial expression that makes the
evaluation of the whole always true), the hacker could easily short-circuit
the game’s door to open regardless of the time of day within the game or
whether the avatar is carrying an appropriate key.4 This is precisely what
happens in certain “god mode” cheats in games that, for instance, allow
players’ avatars to walk through walls. By locating the Boolean evalua-
tion where the collision detection occurs, a hacker can essentially insert
a switch that makes the game’s engine believe that collisions never occur.
The new switch permanently sets the collision state to “false” for every
check. The avatar can then move through any wall in the game. The game
is now full of “not-collisions.”
Certain games establish a documented standard where modifications
can be added by placing the homespun content at certain memory ad-
dresses on the computer. Memory addressing is the second important
digital concept computer science lends to ludology.5 Every instruction
in a computer has a memory address of one sort or another. Much as
 Hacks, Mods, Easter Eggs, and Fossils 75

street addresses, hotel room numbers, or the line numbers in a BASIC

program label potential locations for their users, digital computers use
discrete numbers to reference specific hardware memory addresses that
contain the code or data necessary to perform their tasks. Most, if not all,
contemporary games have abstracted this process so that the machine’s
literal memory addresses do not have to be known. These games can situ-
ationally reference the correct address for, as an example, a folder as rep-
resented by a user’s operating system rather than a specific spot on a stick
of RAM or a hard disk.
Still, anyone who has experienced a “Blue Screen of Death” on a Win-
dows computer has had every layer of abstraction ripped from her or
his interface when the memory address describing the location of a deal-
breaking error was displayed, usually as a particularly befuddling hexa-
decimal number similar to “71d633e5.” Memory addressing is also why
game console developers remain so intent on letting gamers know what
was an 8-bit, 16-bit, 32-bit, or 64-bit system. Each increase indicates an
exponential growth in the maximum number of addresses, meaning (at
least indirectly) more resources with which the consoles could construct
games.

Hacking Nostalgic
At this point, it is useful to revisit the term game engine to help establish
how games are “hacked.” Since the earliest days of home consoles, a game
engine has been the part of a video game that interfaced with a game’s
content and evaluated Boolean logic. An engine, if run alone, is not able
to provide a traditional gaming experience. In some ways, engines en-
force the rules of their virtual worlds, but they are not worlds themselves.
Engines are often created to be reusable, supporting more than one pack-
age of content, as with those displayed in Table 5.1.
It is sometimes difficult, especially in classic games, to know where
an engine ends and content begins, as they may contain engines that
were not designed for reuse and seem inextricably wedded to their con-
tent, like Space Invaders (1979) for the Atari 2600. Some contemporary
“games,” however, are hardly games at all; instead, they are little more
than specialized engines with a minimum of content added before their
release, begging third-party modification, like Quake 3 (1999).
Atariage.com remarks that Space Invaders for the Atari 2600 home
gaming console was the first arcade game licensed for home use (“Atari
2600—Space Invaders (Atari)”), and the ability to create the arcade expe-
rience in the home gave the 2600 its iconic stature. Yet Space Invaders, like
76 Playing in the Past

Table 5.1: Selected popular first-person shooter engines

with implementations

Engine Release Implementations

Wolfenstein 3D 1992 Wolfenstein 3D, Rise of the Triad, Spear of Destiny
Doom 1993 Doom, Doom 2, HeXen, Strife, HacX
Quake 1 1996 Quake, Half-Life,* HeXen 2, X-Men: Ravages of the
Apocalypse
Quake 2 1997 Quake 2, Soldier of Fortune, Heretic 2, SiN, Kingpin,
Daikatana
Quake 3 1999 Quake 3, Star Wars: Jedi Academy, American McGee’s
Alice
Unreal 1 1998 Unreal (original), Unreal Tournament, Deus Ex, Harry
Potter and the Sorcerer’s Stone, Star Trek: The Next
Generation: Klingon Honor Guard
Unreal 2 2002 Unreal 2, America’s Army, Thief: Deadly Shadows,
UT2003, UT2004, Unreal Championship 2, Tom
Clancy’s Splinter Cell, Star Wars: Republic Commando,
XIII
Lithtech 1.0 1998 Shogo, Blood II: The Chosen
Lithtech 1.5 2000 No One Lives Forever, Alien vs. Predator 2
Lithtech 2.0 2002 No One Lives Forever 2, Tron 2.0, The Matrix Online
Halo 2001 Halo, Stubbs the Zombie, Red vs. Blue (machinema)

*Half-Life contains a heavily modified Quake 1 engine with some code

from Quake 2 as well.

a number of 2600 arcade ports,6 was not nearly as true to the original as
even the limited 2600 hardware would allow. In 1999, using new tools like
modern emulators, Rob Kudla hacked Space Invaders’ code in a doubly
nostalgic attempt to re-create more faithfully the coin-operated arcade
version of Space Invaders on the 2600 (“Atari 2600 Hacks”). Kudla made
the invaders, the player’s tank, sounds, and colors more closely reflect the
original’s. His improvements succeeded impressively.
The first step in hacking Space Invaders’ graphics was to discover the
memory address where graphic content (as opposed to the location of the
game’s engine) was kept in the machine language code. This is not nearly
as hard a task as it may sound. Each invader is made up of ten “scanlines”
of graphics, where each scanline matches one pass of a television’s elec-
 Hacks, Mods, Easter Eggs, and Fossils 77

Figure 5.3. Original and hacked graphics, code, and memory addresses for Space
Invaders and Space Invaders Arcade. Images © Taito and Atari, Inc.

tron gun across its screen. Each scanline of the 2600 invaders’ graphics
is eight bits, or switches, wide. It is possible to decompile the machine
language code into a graphic representation of the program’s switches’
values, and the invaders’ graphics’ location in the code become obvious,
in spite of appearing upside-down in the decompiled representation.
In Figure 5.3, the graphics (rotated for easy comparison with the code)
for the original top-level, 2600 invader is displayed on the left next to
the hacked, more arcade-faithful version by Kudla on the right. There are
three columns in each section of disassembled code. The first column for
each invader’s disassembly shows the byte value of the eight-switch line of
graphics in hexadecimal notation.7 The second column shows the value
of each individual switch in every eight-switch line, displaying an “X” for
each switch that is “on” or set to “true.” The last column shows the address
in memory where the line is held, again in hexadecimal format.
Note that the addresses in the third column are the same for both
invaders’ listings ($FC96-$FC9F). The Atari 2600 does not rely on an op-
erating system to abstract memory addresses like most present-day plat-
forms, differing what happens here from the engines of “modern games”
that Bogost notes are modularly “split up into software objects and frame-
works” (55). These, then, are exact, static, unabstracted memory addresses
in fairly monolithic code. That the addresses for the graphics are the same
in the hack as the original suggests that when Kudla made his updates
in Space Invaders Arcade, the logic of the game’s original engine was not
changed, and only a few switches solely related to graphics were hacked
in place.
Kudla’s reliance on a hack in place is also an indication of the degree
to which the engine and content of this classic game are intertwined.
In the byte—a collection of eight switches—located at memory address
$FC96, two switches that were off were switched on and two that were on
were switched off to change the appearance of the bottom of the invader.
Kudla’s creation of two blank lines of graphics at the top of the invader
78 Playing in the Past

ensured the memory addresses remained synchronized in the modified

content. In other words, Space Invaders Arcade is still using the origi-
nal 2600 Space Invaders engine to provide the enforcement of its world’s
rules.
A crucial characteristic of the definition of a hack is that the original
author of the game did not intend for a third party to change the game’s
content. Space Invaders Arcade can be safely considered a hack through a
number of pieces of evidence. First and foremost, Space Invaders was ini-
tially released on a cartridge. Its code was permanently burned into Read
Only Memory (ROM) and this hardwiring could not be changed, similar
to the music on a commercial compact disc. Second, the game contin-
ues to be under copyright, legally forbidding just this sort of modifica-
tion. Third, the author of Space Invaders did not provide documentation
identifying the memory address of the content that Kudla changed, and
the tools that now enable a hobbyist to disassemble and reassemble Atari
2600 games easily did not exist at the time of Space Invaders’ release in
1980. Bogost described a similar situation with Tank and PONG, as well
as Combat for the 2600, a home version of the former arcade game. The
shared codebase in Tank and PONG, like that in Space Invaders and Spaces
Invaders Arcade, served as a proto-engine, tying the games together in a
manner that could be discovered only by a digitalistic critical approach.
“[T]heir common gameplay properties relied entirely on the same code-
base . . . Tank, PONG, and Combat’s relation to one another is far stronger
than interpretative notions like intertextuality or new media concepts like
remediation allow” (58). Kudla’s modification is different from the one
Bogost describes in that Kudla’s code-sharing was unauthorized. From
the medium of the game’s release to a complete lack of documentation
on providing new or altering old content, it is safe to say Space Invaders
Arcade is an unsanctioned hack.
There is another species of game alteration that works in a similar
fashion, but where the hackers’ addition of new content is anticipated
and encouraged by the games’ designers. Rather than a hacker finding
a memory address through subversive code disassembly, in these cases
designers expose a standardized memory address (usually abstracted by
a folder location) where new content can be placed and read by a game’s
engine. This content can be as simple as a new skin for an avatar’s frame
that gives the player a different color or set of clothes or as complicated
as a full modification, including new maps, textures, player models, and
even rules for in-game scoring or physics.
Taken together, hacks and the addition of new code show that games
are never done. From the very first games released on a home console to
 Hacks, Mods, Easter Eggs, and Fossils 79

Figure 5.4. From Quake 1: (top left) default skin; (top right) model wire frame
outline; (bottom left) Spider-Man skin; (bottom right) female skins, marine model.
Images © id Software.

the latest games to hit market shelves, every one is open to alteration and
additions in the digital age, with the Internet enabling mass consumption
even of hobbyist releases.

Engines and Intentionality: Easter or Fossilized Egg?

Hacks, skins, and modifications are means for technically savvy gamers
to coauthor the games they play. Another important concept in the video
game that deserves close attention is the Easter egg, a portion of a game
that is not added by savvy players but hidden by author(s) in the original.
If found, an Easter egg provides a metalepsis similar to what Montfort
describes in Infocom’s Planetfall (30), but the effects are usually much
more innocuous. Like literal Easter eggs, Easter eggs in video games are
meant to be found. Determining whether a hidden portion of a game is a
fossil (an unintended leftover from earlier development) or a true Easter
egg waiting for discovery can be difficult. Distinguishing between the two
and understanding how they affect the interpretation of digital artifacts is
the goal of this portion of this chapter.
The most famous Easter egg in a video game can be found in War-
ren Robinett’s Atari 2600 game, Adventure (1980). Game authors were
80 Playing in the Past

Figure 5.5. Adventure’s Easter egg with code disassembly. © Atari, Inc.

not credited in games released for the 2600 by Atari, and Robinett de-
cided to circumvent this rule by burying his name in a secret room of
the game. That Robinett intended players to find the secret room can be
deduced from three clues. The first and most obvious is that the mes-
sage is visible when playing the game. As will be shown, the game’s code
clearly holds his name, but Robinett provided a means for gamers to read
the message on their television screens without sifting through its digital
code. The second is that the graffiti reads, “Created by Warren Robinett.”
This message expects interpretation: it does not simply hold the author’s
name but unambiguously tells the gamer that the game was indeed made
by the named individual. The last comes from the method by which the
room is discovered. The room is accessed when the gamer’s avatar places
 Hacks, Mods, Easter Eggs, and Fossils 81

a “dot” Robinett hid in the game near the wall of a specific, easily accessed
room. Robinett had programmed the game’s engine to flash objects when
more were on the screen than the 2600’s hardware could easily support.
The 2600 will only easily display a maximum of two complex objects
called “sprites” (called “player graphics” on the 2600) on the screen with
each frame. When more needed to be present, Robinett circumvented
the 2600’s limitation by making Adventure’s frames flash quickly enough
(with two different objects displayed per frame) to let the gamer under-
stand that more objects were there, though with the side effect that the
items would seem to strobe constantly. Robinett put an extra object into
the room that contained the secret dot to ensure that the room would
flash when the player entered.8 A perceptive player would notice the
flashing in spite of there apparently being fewer items than what caused
flashing elsewhere and be curious enough to eventually discover the exis-
tence of the dot. Not only does Adventure allow a gamer to read the mes-
sage on his or her screen, it also provides hints for the player to discover
those means.
That the secret message was not discovered by Atari during the game’s
development testifies to the way games could be, in the classic era, created
by a single programmer without much oversight. A simple disassembly
today, as shown in Figure 5.5, displays Robinett’s message as easily as the
same method finds the invaders’ location in Space Invaders. Yet Atariage.
com lists five more games released by Atari that included some sort of
Easter egg containing “Programmer Credit” (“Tips, Cheats, and Easter
Eggs”).

San Francisco Rush: Time-Delay Easter Eggs

and the Sophisticated Gamer

Not all digital Easter eggs are created to resist a parent corporation’s at-
tempts to dehumanize programmers.9 San Francisco Rush: Extreme Rac-
ing, created for one of Nintendo’s home consoles, the N64, found itself
in a particularly precarious position. The N64 had enough power to ap-
proach very closely the game play and experience of the coin-operated
original,10 released the previous year. Gamers were also expecting the
home version to contain content from the more recent arcade update,
San Francisco Rush: The Rock. The arcade sequel’s title headlined a new
track for the racing game, which went to great pains to intricately recreate
the streets of San Francisco; the extra track was one created on the land-
scape of Alcatraz, “The Rock.” When San Francisco Rush: Extreme Racing
82 Playing in the Past

was released, however, not only was there no Alcatraz, Extreme Racing’s
project lead, Ed Logg, quashed any reports of plans to bring The Rock to
the N64 (IGN Staff “Rush”).
There was a rumor, propagated online, that claimed that Extreme Rac-
ing had the Alcatraz level hidden on the cartridge. A number of myths
regarding how to access the track sprouted, including winning a race on
each track in record time, using a manual transmission in those races,
and even using a specific car while racing. None of these in-game meth-
ods unlocked the track (Stevefel et al.)
The Alcatraz track was unlocked no later than January of 1998, when
a player posted his or her successful discovery of a code to hack the game
to Usenet (bunivfan). The code worked with a device called a GameShark,
which allows gamers to change values in their console’s memory, much
as Kudla did with the Space Invaders. Instead of changing bit values, the
hack pointed the game’s engine to the secret memory address that held
Alcatraz, tricking the engine into loading and playing the hidden track.
On March 19, IGN.com published a code from Atari Games that allowed
gamers to access the track without a GameShark hack (IGN Staff “You’re
Going”).
IGN.com interviewed Logg on April 1, 1998, regarding how the track
came to be hidden on the cart. In brief, Atari Games’ sales department
did not want Alcatraz in the N64 version so that it would not, as they
saw it, compete with the arcade game, and when the level was added it
was hidden so well that, short of the GameShark, no one was able to ac-
cess the track. The sales department was never contacted. The motivation
for hiding the level was not strictly commercial, however; Logg admit-
ted that the level was not likely to be as well tested as the other six in the
game, which is apparent when the level is played (IGN Staff “Exclusive
Interview”).
San Francisco Rush: Extreme Racing challenges the archetypal con-
cept of the Easter egg, like that found in Adventure, in two main ways.
The first is that Extreme Racing’s track was embedded to create a sort of
time-release content rather than exist as something gamers could hunt
and locate within the game. Logg and his team did not provide a hint
like Robinett’s flashing objects. At the same time, it was content that the
game’s authors intended for the gaming public to experience if and only if
they were in some way in contact with the circles that provided codes, be
it Usenet, IGN.com and other Internet sites; gaming magazines; or friends
who could access one of those outlets. In effect, they hid part of the game
outside of the physical media of, in this case, the cartridge. Alcatraz was
to be a multimedia Easter egg.
 Hacks, Mods, Easter Eggs, and Fossils 83

Multimedia Easter eggs have become the rule in video games, and the
means of their distribution mark a sharp change from Robinett’s in-joke
for perceptive players. Lists of newly released codes that almost certainly
could not have been guessed are standard in the back of magazines deal-
ing with video games in what has become a somewhat codependent state
of affairs. The publishing of video games’ codes amounts to free game ad-
vertisements at the same time that the anticipation of new codes sells each
new magazine issue, turning Montfort’s metalepsis into an industry.
The second way that Extreme Racing helps to redefine the Easter egg
is the way that it deprivileged the work’s authors. Alcatraz was discov-
ered without its authors’ consent and contrary to the authors’ design by
GameShark hackers. Gamers are now savvy enough to search for hidden
content that is not accessible solely using the game’s traditional interface.
Alcatraz was meant to be found. With its early discovery, however, gam-
ers proved they had the means to find unintended “eggs” as well. Game-
Sharks, disassemblers like that used by Kudla, model viewers like Pak-
rat (Naughton), and utility software that allows file inspection like hex
editors allow gamers to break from the intended interfaces and creatively
search for hidden content. These tools provide new directions for a player
or “reader” to access digital works.

Fossils: Unintended Easter Eggs

If a programmer creates a subroutine that may not be used in a later ver-
sion of their program, there is no technical requirement to remove it. If
Warren Robinett had decided to hide his secret message from gamers, he
could have deleted the extra room where the message was displayed and
left the data with his message in Adventure’s code, only to be found with a
thorough disassembly. Unlike text that has been struck through, unused
code does not create any noticeable effect for its end users. It simply re-
mains, fossilized.
In fact, there is often good reason to leave unused code in a program.
If a later version needs a similar function, the code is already integrated
with the codebase and is ready to be called. In some respects it can be
an extra tool in the toolbox for an anticipated need, but one that is cur-
rently not required. There is also a potential advantage for fossilized code
when testing an application. If the fossilized code has been cut out but
not removed, perhaps for expediency’s sake, and the program makes its
way through user testing, completely removing the code later would re-
quire more testing to ensure a sloppy removal did not accidentally intro-
duce new errors. Either reason can produce fossilized code in any digital
84 Playing in the Past

work. Fossilizing code is a uniquely digital process. One switch might be

changed (a Boolean expression that, if “true,” would have caused a sub-
routine to be called is set to always evaluate as “false,” fossilizing the sub-
routine) and the end user’s experience is, in theory, exactly the same as it
would have been had the code never existed.
Tomb Raider: The Angel of Darkness (2003) is an example of a video
game with fossilized content. Whereas the first five entries in the Tomb
Raider series shipped like clockwork, one coming out before each year’s
holiday buying season, Angel of Darkness was released over three-and-
a-half years after Tomb Raider: Chronicles (2000). One Tomb Raider fan,
upset by Angel of Darkness’ errors, created a list called the “AOD Bugs de-
finitive list” in which sixty-two bugs were listed (Dragoncarer). Bug 1.1,
“The Secret Garden,” describes fossilized content: Angel of Darkness opens
with a training level that allows users a chance to become familiar with
the complicated controls of the game. Much of the level was scrapped,
presumably to allow the game to be released more quickly. The fossilized
content is still within the game, and parts of it can be seen from conven-
tionally accessible portions of an extant level. Entering the fossilized con-
tent requires either hacking the game or loading a saved game where Lara
Croft, the game’s protagonist, begins at a location in the fossilized area.
As Figure 5.6 shows, in the fossilized area, large parts of the surround-
ing cityscape are missing and training instructions are displayed for the
player. One safe assumption from the archeological evidence is that the
training section of Angel of Darkness was originally intended to be much
more elaborate.

Hot Coffee and Possible Piltdowns

The skull planted in Piltdown, England, may be the greatest scientific
hoax, and is certainly the greatest in physical anthropology.11 No more
than an orangutan jaw placed next to a human skull, the false fossil
tricked anthropologists for more than four decades. The Piltdown con-
cept is a useful one for video games, especially when trying to evaluate
rhetorical intent with respect to apparently fossilized content. Are pro-
grammers cunning enough to disguise Easter eggs in their games as fos-
sils to trick the gaming public into believing they were accidents?
Rockstar’s Grand Theft Auto: San Andreas raises just this question. San
Andreas contains what appeared to be a fossil: a portion of the game that
is inaccessible without some sort of hack or download. Here, the osten-
sibly fossilized content allows a gamer to have limited control over rela-
tively graphic sexual scenes between the player’s avatar, named “CJ,” and
 Hacks, Mods, Easter Eggs, and Fossils 85

Figure 5.6. Fossilized training area in Tomb Raider: Angel of Darkness.

Image © Eidos Interactive.

his girlfriend(s). After dating, CJ’s current girlfriend may invite him into
her house for coffee. For a player without the “Hot Coffee” modification,
the camera shakes suggestively (like CJ’s car when he picks up a prosti-
tute) after he is invited into his girlfriend’s home, but for a player with the
modification, the camera follows CJ inside.
Patrick Wildenborg, who discovered the “fossil” and created a modifi-
cation that allowed others to access it easily, says the following about the
modification on his website:

After reading various discussion [sic] about this mod around the
internet, I would like to make the following statement:
All the contents of this mod was already available on the original disks.
Therefor [sic] the scriptcode, the models, the animations and the dialogs
by the original voice-actors were all created by RockStar. The only
thing I had to do to enable the mini-games was toggling a single bit in
the main.scm file. (Of course it was not easy to find the correct bit).
(Wildenborg, emphasis in the original)12
86 Playing in the Past

After Hot Coffee’s discovery, Rockstar quickly became trapped in a con-

troversy, and the Federal Trade Commission investigated how the game
received a “Mature” rating, which recommended its sale to people seven-
teen years old and older, and not an “Adults Only” rating, which would
have restricted its sale to those able to prove they were eighteen years old
and up. An “Adults Only” rating was then given to San Andreas due to the
apparent fossil discovery, and the game was pulled from mass retailers’
shelves. A revised version without the Hot Coffee content was written
and re-released, and an update was provided online to erase the content
from personal computers with the game already installed.
Rodney Walker, a Rockstar spokesman, quickly attempted to fashion
a metaphor for Hot Coffee’s inclusion favorable for his company:

An artist makes a painting, then doesn’t like the first version and
paints over the canvas with a new painting, right? . . . That’s what
happened here. Hackers on the Internet made a program that
scratches the canvas to reveal an earlier draft of the game. (Schiesel)

While it is true that fossilized content can “reveal an earlier draft” of a

game, the metaphor is flawed due to Walker’s inattention to digitalism’s
influence on the rhetoric of the San Andreas minigame. A copy of a paint-
ing does not include an embedded history. Owners will not discover a
sketch of Mary with her left hand to her breast in their copies of mass-
produced prints of Leonardo da Vinci’s Virgin on the Rocks. Every copy of
Tomb Raider: Angel of Darkness has the unfinished portions of the train-
ing level, every copy of San Andreas has the hidden sex game, and infra-
red reflectography is not required to discover the content (“The Hidden
Leonardo”). It is also possible to erase unused code from a digital prod-
uct’s final version, whereas it is nearly impossible to remove an unfinished
draft from below a masterpiece’s last layers of paint. Walker’s metaphor
may have been one of the best of the quick attempts to understand Hot
Coffee’s inclusion, yet it is one that unjustly favors the publisher in its
depiction that ignores digitalism.

The Lessons of Digitalism

Studying virtual realities dependent on digital hosts requires an empha-
sis on digitalism, which in turn enables readings that would otherwise
be missed. Conjoining what are now multidisciplinary concepts is re-
quired to question complex works, like Grand Theft Auto: San Andreas,
on a footing equal with authors, whether those authors are sanctioned
 Hacks, Mods, Easter Eggs, and Fossils 87

or unauthorized. These approaches help remove the limitations of legacy

methods, and instead of losing many of those tools, the integration rein-
vigorates them.
This chapter’s quick survey, moving from nostalgic to contemporary
digital artifacts, does little more than scratch the surface of a digital ap-
proach, as it concentrates on the contested fringe of the representation
of virtual worlds. Articles like Carolyn Miller’s “Writing in a Culture of
Simulation: Ethos Online” have opened the door for analyzing the next
wave of virtual reality, where great characters are created not simply by
great writing, but by skillful writing combined with convincing artificial
intelligence. The time and tools for integrating digitalism with cultural
studies are here, and the field awaits.

Notes
1. Even more stark a break from the “black box” approach is Montfort’s character-
ization of BASIC as a particularly difficult language with which to write IF and
one he has tried using himself. Here, Twisty Little Passages has clearly moved
from the accessible approaches of IF-as-narrative and riddle and on to the
third—complex computer programs.
2. An interesting study would be the history of video games based on analog com-
puters. I am only familiar with William Higinbotham’s Tennis for Two, created in
1958 (US Dept. of Energy). Regardless, this claimant for the title of the first video
game offers an intriguing alternative to the metaphors of digitalism.
3. There is always an exception that proves the rule. In Visual Basic 6.0, once argu-
ably the most popular programming language for the Microsoft Windows oper-
ating system, “true” is represented by “-1.” To make things even more convoluted,
once the Visual Basic code is translated into machine language, the “-1” becomes
a “1” again (Bailey et al.).
4. The is reminiscent of Captain Kirk’s answer to the Kobayashi Maru simulation in
cadet training, recounted in Star Trek: The Wrath of Khan. When presented with
a “no-win”/always “false” trial in training, Kirk, after failing the simulation twice,
reprograms/hacks the computer to create a winning solution.
5. Though the explanations given for these terms herein are arguably accurate,
a much better—and extremely accessible—primer text is Richard Mansfield’s
Machine Language for Beginners, which explains assembly language program-
ming and the operation of a series of chips found in the Atari 2600, Commodore
64, Nintendo Entertainment System, and the Apple II. These chips are simple
enough to allow a beginner to achieve a reasonable understanding of their opera-
tion and complete enough in their design that the conceptual lessons learned
easily extend to contemporary processors.
6. “Port” comes from “portable” and is something of a misnomer, as most early
“ports,” like Space Invaders for the 2600, were more precisely rewrites on alter-
nate hardware. Other arcade ports that did not fully exploit the 2600’s hardware
88 Playing in the Past

include the infamous Pac-Man (see Ms. Pac-Man on the 2600 as an example of
what the game could have been), Zaxxon, and Popeye.
7. Hexadecimal is a base-16 numbering system that allows eight binary switches
to be displayed in two digits. In the “ones” place, A, B, C, D, E, and F represent
10, 11, 12, 13, 14, and 15, respectively. An A in the “tens” place represents ten
sixteens, or 160. For this study’s purposes, all that is required is an understanding
that these are numbers and that the “$” preceding each hexadecimal number is a
signpost that we are using base-16, not base-10.
8. In this room, the “darkness” was technically an object. The darkness, the dot, and
one extra item were enough to create the flash.
9. An engaging introduction to the topic of the dehumanization of the programmer
is Edward G. Nilges’s “PRACTICAL DECONSTRUCTIVE CODING.”
10. Extreme Racing’s project lead mentions that tracks were shared between the ar-
cade team and the N64 team, meaning that they were both using approximately
the same format for their content, truly porting levels from platform to platform
rather than rewriting the game. The levels from the original arcade game were
going directly into Extreme Racing and three levels from Extreme Racing were
borrowed and placed into a later arcade release (IGN Staff “Exclusive Interview”).
11. This section can be found in an expanded form in “Inviting Subversion: Meta-
lepses and Tmesis in Rockstar Games’ Grand Theft Auto Series” in The Meaning
and Culture of Grand Theft Auto Critical Essays, ed. Nate Garrelts (Jefferson, NC:
McFarland Press, 2006), 210–25.
12. After a controversy appeared regarding the modification, Wildenborg appro-
priately hid this text and the locations of pictures of the modification in his
webpages through the use of “html comments,” a method usually used by html
coders to leave messages for other coders viewing their pages’ code. To read the
quoted material, web users had to hack Wildenborg’s page to access the trivially
hidden content from within the web page’s html code, a method not coinciden-
tally similar to what Wildenborg did to hack San Andreas. He has, as of 13 Janu-
ary 2008, restored the text to his site.

Works Cited
Aarseth, Espen. Cybertext. Baltimore: Johns Hopkins University Press, 1997.
“Atari 2600 - Space Invaders (Atari).” www.atariage.com (6 November 2004). Accessed
7 August 2005.
“Atari 2600 Hacks - Space Invaders Arcade.” www.atariage.com (19 November 2004).
Accessed 7 August 2005. .
Bailey, Wm. Ruffin, Cor Ligthert, Herfried K. Wagner, et al. “Why is minus one (-1)
equal to true in VB again?” microsoft.public.dotnet.languages.vb (Usenet, 21–23
June 2004). Accessed 13 August 2005
Barr, Roger, ed. “Hacked Rom Reviews!” www.i-mockery.com. Accessed 12 August
2005.
Blue Zircon. “Thompson 1928.” www.gtagaming.com (26 July 2005). Accessed 12
August 2005.
 Hacks, Mods, Easter Eggs, and Fossils 89

Bogost, Ian. Unit Operations. Cambridge: MIT Press, 2006.

bunivfan. “SF Rush Hidden Track Code.” rec.games.video.nintendo (Usenet, 13 January
1998). Accessed 10 August 2005.
DieselGT. “‘Tumbler’ Batmobile.” www.gtagaming.com (17 July 2005). Accessed 12
August 2005.
Dragoncarer. “AOD Bugs definitive list.” db.gamefaqs.com (9 December 2003).
Accessed 7 August 2005.
IGN Staff. “Exclusive Interview: Ed Logg.” ign64.ign.com (1 April 1998). Accessed 7
August 2005.
“The Hidden Leonardo.” www.nationalgallery.org.uk (2001). Accessed 12 August
2005.
“History of the 11th Annual Interactive Fiction Competition.” ifcomp.org/comp05/
history.html (2005). Accessed 8 August 2005.
IGN Staff. “Exclusive Interview: Ed Logg.” ign64.ign.com (1 April 1998). Accessed 7
August 2005.
___. “Rush: The Rock Not Coming to N64.” ign64.ign.com (17 November 1997).
Accessed 7 August 2005.
___. “You’re Going to Alcatraz.” ign64.ign.com (19 March 1998). Accessed 7 August
2005.
Mansfield, Richard. Machine Language for Beginners. Greensboro, NC: Compute!
Publications, 1983.
Montfort, Nick. Twisty Little Passages. Cambridge: MIT Press, 2003.
Miller, Carolyn. “Writing in a Culture of Simulation: Ethos Online.” In The Semiotics
of Writing: Transdisciplinary Perspectives on the Technology of Writing. Ed. Patrick
Coppock. Turnhout, Belgium: Brepols, 2001. 253–79.
Muckelbauer, John. Review of The Terministic Screen: Rhetorical Perspectives on Film
by David Blakesley. JAC 23.4 (2003).
Naughton, Tom. Pakrat “About” page. pakrat.fragland.net (10 October 2001). Accessed
14 August 2005
Nilges, Edward G. “PRACTICAL DECONSTRUCTIVE CODING, an essay in the
critical theory of computer science.” comp.lang.basic.visual.misc (Usenet, 16
November 2001). Accessed 10 August 2005.
Pomaville, Leann, ed. Quake Woman’s Forum. 1999. planetquake.com. Accessed 11
August 2005.
“San Francisco Rush–Cheat Codes & Secrets–GameFAQs.” gamefaqs.com. Accessed 7
August 2005.
Schiesel, Seth. “Video Game Known for Violence Lands in Rating Trouble Over Sex.”
New York Times (21 July 2005).
Stevefel; Tal A. Funke-Bilu; bliss; et al. “SF Rush Alcatraz track: Relase [sic] the code
Atari-Games !! (please).” rec.games.video.nintendo (Usenet, 18 January–18
February 1998). Accessed 10 August 2005.
“Tips, Cheats, and Easter Eggs.” www.atariage.com/hint_list.html (10 October 2004).
Accessed 7 August 2005.
US Dept. of Energy. “Brookhaven 1958 Video Game - DOE Research and
Development (R&D) Accomplishments.” Office of Scientific and Technical
90 Playing in the Past

Information (www.osti.gov/accomplishments/videogame.html). Accessed 7 May

2005.
Wildenborg, Patrick. “PatrickW GTA Modding | Home.” patrickw.gtagames.nl/index.
html. Accessed 10 March 2006.
Wolf, Mark J. P., ed. The Medium of the Video Game. Austin: University of Texas Press,
2001.

Games Cited
Adventure (Atari VCS/2600). Sunnydale, CA: Atari, 1978.
Descent (Windows). Champaign, IL: Parallax Software, 1995.
Grand Theft Auto: San Andreas (PlayStation 2). New York: Rockstar Games, 2004.
Grand Theft Auto III (PlayStation 2). New York: Rockstar Games,
2001.
Quake 1 (Macintosh). Santa Monica: id Software/MacPlay, 1996.
Quake 2 (Windows). Santa Monica: id Software/Activision, 1997.
Quake 3 (Macintosh). Santa Monica: id Software/Activision, 1999.
Space Invaders (Atari VCS/2600). Sunnyvale, CA: Atari, 1980.
Space Invaders Arcade (Atari VCS/2600). Atariage.com, 1999.
San Francisco Rush: Extreme Racing (N64). Milpitas, CA: Midway Games West, 1997.
Tank (Coin-Op). Kee Games, 1974.
Tomb Raider: Angel of Darkness (Macintosh). Austin: Aspyr, 2004.
Tomb Raider: Chronicles (Macintosh). Austin: Aspyr, 2000.
Zork (various platforms). Cambridge, MA: Infocom, 1979.