The World in the Network:The Interop Trade Show, Carl Malamud's Internet 1996 Exposition,

and the Politics of Internet Commercialization MICHVEMASSACHUSETTS INSTITUTE

by OF TECHNOLOLGY

Colleen E. Kaman JUN 2 3 2015

B.A. AnthropologyBates College, 1995 LIBRARIES

SUBMITTED TO THE PROGRAM IN COMPARATIVE MEDIA STUDIESIN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE IN COMPARATIVE MEDIA STUDIESAT THE

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

June 2010

0 2010 Colleen Elizabeth Kaman. All rights reserved.

The author hereby grants to MIT permission to reproduce and to distributepublicly paper and electronic copies of this thesis document in whole or in part in any

medium now known or hereafter crerad.

nature of Author: Signature redactecProgr in Comppative Media Studies

17 May 2010

Sia nature redactedy:Certified b

Accepted b

I

William Charles UricchioProfessor of Comparative Media Studies

Director, Comparative Media Studies

,-7 . Thesis Sufervisor

y: Signature redactedH'ny'Jenkins III-

Provost's Professor of Communication, Journalism, a Vd Cinematic Arts

Department of Communication, University of Southern CaliforniaThesis Committee Member

Accepted by: Signature redactedNick Montfort

Associate Professor of Digital MediaProgram in Writing and Humanistic Studies

Thesis Committee Member

Sig

77 Massachusetts AvenueCambridge, MA 02139hftp://Iibraries.mit.edu/askMITLibraries

DISCLAIMER NOTICE

Due to the condition of the original material, there are unavoidableflaws in this reproduction. We have made every effort possible toprovide you with the best copy available.

Thank you.

Slight cropping of page numbers at the bottompage margin.

Prologue

One starting point of this study was a curiosity about the meteoric transformation of theInternet from an experimental research network into a global communications medium.

Figure 1: "International Connectivity" in 1991. This map shows whatto electronic networks, including the Internet. However, this mapquality of that connectivity.

countries had permanent linksdoes not indicate the level or

Figure 2: "International Connectivity" in 1997. This map shows how dramatically permanentinternational links to the Internet had expanded in just six years.

Copyright 1991 and 1997 Lawrence H. Landweber and the Internet Society.Unlimited permission to copy or use is hereby granted subject to inclusion of this copyright notice.

2

INTERNATIONAL CNNECTIVITY

SIntsa.t

EMU.1 Only(UUCP,Fid.N.t)N0Conntity

INTERNATIONAL CO NECTIVITY

Blinet but not InternetEU.Niny (UUiCP, Fkd.N.* : 1.

1No Connwctivfty rn

The World in the Network: The Interop Trade Show, Carl Malamud's Internet1996 Exposition, and the Politics of Internet Commercialization

Abstract

In the early 1990s, the Internet emerged as a commercially viable global communicationsmedium. This study considers the role that representatives of the military-industrialresearch world played in the physical expansion of the Internet. It does so by examining thesocial practices and processes of the semi-annual "Interop" computer-networking tradeshow, and one affiliated "exposition." Beginning in 1987, and for nearly a decade, Interopoperated as a forum that brought representatives from industry and the research and usercommunities into strategic alliance to tackle the practicalities of expanding the Internet'score networking protocols and assembling diverse networks into a global Internet. Theperiod examined culminates with the Internet 1996 World Exposition. Through that event,technologist Carl Malamud drew on the rhetoric of turn-of-the-century world's fairs todemonstrate the value of faster networks but also argued for a conception of "the commons"that could ideally be served by the rapidly privatizing Internet. In the absence of acomprehensive history of the commercial expansion of the Internet, analysis of thesepractices provides a pioneering analytic narrative of a crucial strand of this development.This thesis moves between levels of analysis, specifically between the Interop network, theInternet 1996 Exposition event, and the perspective of Malamud himself. By highlightingthese hitherto neglected practices, this examination deepens our understanding of theforces that proved critical to the Internet's commercial success.

Thesis Supervisor: William Charles UricchioTitle: Professor of Comparative Media Studies

2

Acknowledgments

I'd like to extend my deepest thanks to the many individuals who helped me along the way.CMS mentors William Uricchio, Henry Jenkins, and Nick Montfort provided intellectualguidance and encouragement that greatly influenced this project as well as many otherendeavors.

I am grateful to Glorianna Davenport, Lucy Suchman, Michael Fischer, Fred Turner, andStefan Helmreich, who helped along the way, and to Lisa Williams, whose sketches helpedme understand protocol layers and whose stories kept my spirits high.

I would like to extend my thanks to numerous interviewees who generously gave of theirtime to speak to me about their experiences as well as the technical aspects of their work inperson, by phone, and over email. These include Karl Auerbach, David Brandin, DavidClark, Dave Crocker, Tom Keating, Ole Jacobsen, Dan Lynch, Tom Keating, Carl Malamud,Howard Rheingold, Andy Lippman, Marty Lucas, and Marshall Rose. Without theirpatience and assistance, this work would never have been possible.

A special thanks goes to my entire family, who have always supported my various interestsand never failed to offer words of encouragement. I am particularly grateful to Bridget andAnthony Barron who so generously offered their home for my numerous trips to the SanFrancisco Bay area. Finally, thanks to Abdulrazzaq al-Saiedi, who kept me company andlistened to me ramble on about my thesis at all hours of the day and night.

4

List of Figures

Prologue

Figure 1: "International Connectivity" in 1991

Figure 2: "International Connectivity" in 1997

Chapter One

Figure 3: Advertisement for the October 1, 1982 Launch of EPCOT Theme Park

Figure 4: The AT&T Network Operations Center scene, Spaceship Earth, 1984

Figure 5: AT&T's International Fiber Optic Cables, circa 1998

Chapter Three

Figure 6: Screenshot, Construction of Interop Show Network, date unknown

Figure 7: Diagram of the INTEROP90 Show Network Configuration

Chapter Four

Figure 8: Screenshot, Internet 1996 Expo website

Is

Contents

Prologue

Abstract

Acknowledgments3

List of Figures4

Introduction: The Commercial Sphere as a Site of Social Change7

Chapter One: As our Thirst for Knowledge Grew, the World Began toShrink: Spaceship Earth as a Networked Utopia20

Chapter Two: Internet Explorers and Digital Worlds36

Chapter Three: I Know it Works, I Saw it at Interop49

Chapter Four: In Truth, All the World Was There: The Internet 1996 Expo63

Chapter Five: Conclusion79

Appendix A: List of Interviewees89

References90

Introduction: The Commercial Sphere as a Site of Social Change

In 1994, Kevin Kelly -- information technology pundit and founding executive editor of

Wired, and co-founder of the online community the WELL' -- argued in "Out of Control"

that the marketplace in the emerging networked society was the site of social change. The

text, which was organized in a format similar to the Whole Earth Catalog, outlined deep

interconnections between the biological, the technological, and the social (Turner 2006,

200). Describing living systems in computer science terms, Kelly suggested that organisms

advanced by "hacking," or working-around, challenges that, over time, naturally led to

ubiquity and complexity. Likewise, Kelly asserted that technology itself had evolved such

that computer networks had transformed the corporation into a living organism,

"distributed, decentralized, collaborative, and adaptive." Such a process, Kelly believed,

signaled the emergence of a global information system that naturally guided an economy

within which men and machines would be effortlessly integrated. In other words, Kelly

downplayed the physical aspects of the global economy, including the computer-networking

hardware and production lines as well as the physical labor and relationships embedded in

these objects.

As Fred Turner has demonstrated, Kelly's argument synthesized influences that had

first formed around the Whole Earth network. The emerging society he depicted integrated

1960s-era countercultural ideals with corporate interests and the collaborative practices

and rhetoric of interconnectedness associated with the military-industrial research world

(Turner 2006, 199-206). According to Kelly, the emerging post-industrial economy was a

powerful demonstration of the deep integration of computers and computer networks in

society, revealing "a common soul between the organic communities ... and their

manufactured counterparts of robots, corporations, economies, and computer circuits"

(Kelly 1994, 3). The world itself had become an information system, and with it, new forms,

such as the bee swarm (and with it, the "hive mind") and complex adaptive systems,

emerged to replace the hierarchical logic of the previous era. For corporate executives

trying to understand the technological and economic changes they faced, Kelly encouraged

them to "obey the logic of the net" if they hoped to succeed in the emerging economy, a

1 The WELL, or Whole Earth 'Lectronic Link, was founded in 1985 by Stewart Brand and LarryBrilliant. Many of the WELL's core members were previously associated with Brand's Whole EarthCatalog, and like the catalog, quickly became a highly influential computer conferencing system andvirtual community.

7

system in which the intangibles of the network would supersede the world of physical

objects" (1998, 160).

This countercultural worldview depended heavily on the cybernetic theories of

information management that drew connections between system social theories and objects

and systems; yet in the process of translation, the counterculture downplayed and even

obscured the physical aspects of the technologies built in the Cold War-era research labs.

Still, the physicality of computer networks represents a critical aspect of the Internet and

continues to be a site of conflict. Those conflicts range from "Denial of Service" attacks, to

edicts of national and international courts limiting the reach of information online2 and the

control mechanisms of corporate providers and national governments, to lagging broadband

infrastructures that cause "information traffic jams" and fragment network connectivity.

The scope and increasing severity of these conflicts surrounding the physical network have

led Harvard Law School professor Jonathan Zittrain (2008) to predict that the Internet is

increasingly likely to become a "closed" technology as aspects of the technological system

that encourage experimentation and exchange are replaced by consumer "appliances" that

offer little in the way of participation.

What is it about the physical aspects of computer networks that have bedeviled

idealistic visions of the networked society? External forces, such as commercial influences

or national interests, are not simply corrupting an exceptional technology and the ideal

society it promised, as many countercultural figures supposed. Part of the answer lies with

the nature of the technology itself. When the Internet and then the World Wide Web3 first

2 LICRA v. Yahoo (2000) was the first successful international challenge to the Internet community'sargument that the Internet represents an exceptional technology that should be governed bydifferent means than by national laws, as are traditional communications technologies. The caseexamined whether it was illegal for a Yahoo! online auction site to sell Nazi artifacts in France.3 The World Wide Web, sometimes confused with the Internet by people who first encountered themboth at the same time (in the mid-1990s or later), was a system for making information widelyavailable that was conceived and pioneered by Tim Berners-Lee, a British citizen working at theCERN research institute in Switzerland. It consisted of 1) "web sites" (electronically accessible"places") for storing text and images with a protocol for assigning each one a name (formed ofstandard alphabetic and typewriter keyboard characters)-termed a URL (for Universal ResourceLocator); 2) "hypertext," text with certain words appearing on-screen as underlined or differentlycolored and serving as "links" that when "clicked on" with a computer mouse, bring to the screen anassociated web site; and 3) a programming language, originally HTML ("hypertext mark-uplanguage"), for giving each web site a standard, widely interpretable format for its information. Byproviding a network of physically connected computers on which web sites can reside, to be accessedat any time, the Internet served as the communication infrastructure for the World Wide Web.Conversely, the World Wide Web, by offering ever richer information content, undergirded and

A

emerged into public view in the mid-1990s, enthusiasm for networked exchange and

distributed communities all but obscured the tangle of cables and "cyberspace-warping

wires" (Stephenson 1996) as well as the significance of networked computing's history. Yet,

the Internet had a history. It is a distributed computer network created by linking together

previously existing smaller computer networks, of which the best known was the ARPAnet

(the Advanced Research Projects Agency (ARPA) network for rapid communication among

Department of Defense-linked researchers). In other words, it has its roots in the military-

research culture that emerged in the wake of World War II and the Cold War. The network

was developed to be independent of centralized control, flexible, and readily adaptable, such

that the technology could withstand nuclear attack. At its core, the Internet operates

according to a suite of protocols known as TCP/IP (Transmission Control Protocol/Internet

Protocol) that specifies how to structure, transmit, and receive information between

dissimilar networks. 4 These protocols allowed for the ubiquitous connectivity upon which

the modern Internet is based.

Another physical aspect of distributed network technologies is their tangible

infrastructure. Since this technology often bootstraps onto existing telecommunications

wires and cables, the computer network becomes a point of conflict within existing

infrastructures, laws, and norms. In the early 1990s, for example, large-scale commercial

providers (like America Online, CompuServe, and Prodigy)5 fought the organizational logic

of the Internet that allowed for peer-to-peer transmission of data packets regardless of

source or terminus. In contrast, they envisioned closed communities that offered easy-to-

motivated the improvement of the capabilities of the Internet far beyond its original function ofrelaying messages. Each one, an enthusiast might say, sustained and nourished the other, in asymbiotic co-evolution powered by human sociability and curiosity.4 TCP/IP had been developed as an experimental, U.S. military-funded solution to the technicalproblem of connecting dissimilar "packet-switched" networks and earlier radio relay technologies. Bystrict definition, TCP/IP is only two protocols - TCP (Transmission Control Protocol) and IP(Internet Protocol) - each performing a distinct function. However, the term "TCP/IP" is commonlyused to describe an entire family of protocols known as the TCP/IP protocol suite. For example, itspecifies protocols for performing tasks such as file transfer (FTP or File Transfer Protocol),electronic mail (SMTP or Simple Mail Transport Protocol), and remote access to a computer (telnet).The TCP/IP protocols are standards for formatting, addressing, fragmenting, delivering,reassembling and checking transmitted information. Any computer network, even a physicallyisolated one having no connection to the Internet can use TCP/IP protocols. However, many considerthe public Internet synonymous with these protocols because it is a global TCP/IP network. TheInternet is, among other things, an enormous TCP/IP network.5For a period account of Prodigy, see Howard Rheingold's chapter, "Disinformocracy" in The VirtualCommunity: Homesteading on the Electronic Frontier (2000), available online athttp://www.rheingold.com/vc/book/.

9

use services for their customers that included managing online access, exchanges on public

forums and even e-mail. By 1996, explicit regulations tempered the utopian assertions that

networked computing would (or could) challenge the legitimacy of institutions and

traditional governance structures. A law passed by Congress in 1996 marked the first

legislative attempt to regulate speech on the Internet. That same year, the World

Intellectual Property Organization (WIPO) drafted the so-called "Internet Treaties" that

would go on to play a major role in copyright disputes.6 In fact, some of the most powerful

structuring agencies on the Internet today - the protocols and standards as well as

legislation that govern the Net (what Internet legal scholar Lawrence Lessig (1999) calls

"West Coast Code" and "East Coast Code," respectively) - largely function as invisible

infrastructures that appear as "natural" characteristics of the system and thus don't reveal

the profound relationship between discourses around a technology and its physical

attributes.

Continuing debates over the shape and limits of the Internet reveal deeper truths

about modern communications infrastructures and their relationship to previous

communications systems. These debates also point to larger shifts between the relative

power of the State and private enterprise. They reveal that these technologies did not

replace Industrial-era infrastructures so much as facilitate their reorganization, and then

build upon them a new distributed management system that carried with it its own set of

operational logics. These struggles suggest questions about the role that engineers and

organizations affiliated with the military-industrial research world might have had in the

physical expansion and commercialization of the Internet: How did they understand their

roles as architects of this emerging global infrastructure? How were they able to leverage

the cybernetic discourses and interdisciplinary, collaborative practices into strategic

alliances and practical strategies for computer network expansion that worked to ensure

the global success of the Internet? Given what we already know about the military-

industrial research world's contributions to the commercialization of the Internet, what do

their efforts to construct the physical networks reveal about the organizational strategies

that ensured the Internet's successful commercial transition?

6 These copyright laws include the WIPO Copyright Treaty (WCT) and the WIPO Performances andPhonograms Treaty (WPPT). In the U.S., these treaties were implemented with the passage of theDigital Millennium Copyright Act (DMCA) in 1998. The DMCA outlaws technologies intended tocircumvent efforts to control access to copyrighted works.

10(

One network of individuals who focused on the practicalities of Internet expansion,

this research suggests, were affiliated with the largely overlooked "Interop" computer-

networking trade show and conference. 7 These semi-annual events, as well as the trade

shows company's associated publications and gatherings, were important for the physical

implementation of the Internet's core networking protocols that made interoperability

between distinct networks possible. Interop founder Dan Lynch assembled a core group of

Silicon Valley network engineers, vendors, and entrepreneurs associated with the military-

industrial research world. Beginning in 1987, and for nearly a decade, these engineers

engaged with a network of people and interests from the commercial and user communities,

addressing the considerable technical and organizational challenges of creating

interoperable hardware. These network developers included engineers and entrepreneurs

such as Vint Cerf, David Clark, Karl Auerbach, Paul Mockapetris, Dave Crocker, and Carl

Malamud, as well as representatives from Sun Microsystems, Cisco Systems, Apple, and

Digital Equipment Corporation (hereafter, DEC). Out of these encounters emerged shared

understandings of the viability of the Internet community's TCP/IP core networking

protocol, as well as how the interconnection of distinct networks might be accomplished.

The Interop trade show became a sensation, becoming one of the few places that actually

demonstrated functioning inter-networks: distinct networks that connected to one another

but also linked outward to the Internet, as well as products that functioned across the

networks themselves. Interop became one of the most respected and popular trade events in

the industry; by the early 1990s, the gathering had expanded from the U.S. (largely

California) to international locations such as Sydney, Paris, and Tokyo.

Lynch brought these different communities together in a series that since have been

described by scholars as (Turner 2006) "network forums." Comprising a series of

conferences, events, affiliated publications, and an informal membership of scientists and

engineers, these network forums functioned as critical sites for the "translation" of

computer internetworking technologies that allowed the Internet to expand across physical

boundaries into new realms. Successful exchanges between industry, academe, and

government extended the legitimacy of the Internet community's practices and processes

7 There are numerous explanations for the Interop trade show's relative obscurity today, chief amongthem the choices of the network developers themselves. They have deeply influenced the popularhistory of the Internet, yet their accounts largely downplay the role of the Interop trade show and itsnetwork, perhaps because the commercial orientation and focus on the practicalities ofimplementation didn't easily map to more strictly defined technical standards-setting efforts.

11

more deeply into the realm of the massive economic and technological forces reorganizing

the global economy. These actors shared an understanding of themselves as architects of

the emerging networked society, freely integrating economic, technical, and social frames as

they envisioned a global system of interconnected computer networks crisscrossing the

globe, and what the society that supported it might be like. With each "translation" across

another domain, the vision of the Internet attracted more allies. The emerging project grew

to include previously established overseas university research relationships with

international representatives like Joichi Ito (Japan) and Jun Murai (Japan). Together,

they would not only create the first prototypes of the global Internet but also establish the

collaborative processes that proved critical for the mutual accommodation and adaptation

required for the Internet's commercial success.

The narrative reach of this study starts in the early 1990s, as the Internet's place as

the global standard seemed increasingly fixed and the Interop's show network was in high

production. It focuses on the Interop network's role in the standardization of the Internet,

and more specifically two projects affiliated with Interop, Carl Malamud's 1993 survey of

the emerging global Internet and his Internet 1996 World Exposition. The second project,

ambitious in scale and concept, constituted an "exposition" that drew on the rhetoric of

turn-of-the-century world's fairs - first, to demonstrate the feasibility of global inter-

networking, but also to argue for a conception of "the commons" that could ideally be

served by the Internet, which was rapidly becoming privatized. The 1996 exposition

launched just as the most influential engineers and entrepreneurs in the Interop network

began to drift away. Although computer networks were still an "unfinished" technology -

they "broke down" with some frequency, were as yet unable to accommodate real-time audio

and video streams, and had yet to extend much beyond industrialized nations - the

affiliates of the Interop network had helped to create the social and technical conditions

necessary to fulfill a vision of the Internet as a global, commercially viable communications

medium.

By recounting the history of the Interop network, 8 this study considers how the

trade show network functioned alongside more explicit (and more researched) technical

8 Undoubtedly, Interop warrants a standalone analysis that might explore the trade show's role intechnical advances as well as its role in the eventual success of Internet standards in the TCP/IPversus OSI standards war. Here, I focused on the network engineers and have not been able togather material on corporate projects from company archives.

12

standardization efforts, and underscores the instrumental role that the military-industrial

research world's culture had in the commercial expansion of the Internet. Alongside the

imperatives of developing and implementing computing technologies, this research culture

facilitated the development of deeply entrepreneurial and collaborative practices. These

practices coalesced in the 1980s during the computer industry's debates over "open

systems" and the creation of particular information infrastructures. At the core of these

debates were battles over different versions of standardization, which were largely fought

between the Internet protocols and those stipulated by a traditional governmental

standards process. For network engineers, as the catchphrase "rough consensus and

running code" (coined by David Clark in 1992) implies, these struggles became framed in

terms of the "social and moral order of society" (Kelty 2008, 8).

Interop founder Dan Lynch was a former ARPAnet researcher and a member and

industry representative at the Internet Architecture Board (or IAB - it was originally called

the Internet Activities Board), the core architectural leadership organization that guided

the development of the Internet. As these primarily research-oriented practices became

increasingly difficult to implement in the complex commercial and highly litigious

standards environment, Lynch and the other engineers affiliated with Interop reoriented

Internet standards-setting by applying these practices to the practical imperative of

assembling functional links between networks. By doing so, they fashioned a hybrid model

of network standardization that exposed the broader commercial community to the Internet

engineers' manner of condensing the "process of standardization and validation into

implementation" (Kelty 2008, 173) and offered useful knowledge related to the practicalities

of linking networks. Such instruction also "routinized" Internet practices: that is, Internet

leadership imposed a kind of "system" for linking computer networks and developing

products that would run on such networks that allowed them to achieve better control of

implementation and expansion processes (Yates 1993, xvii). In these ways, Interop

functioned as a critical intervention for an information technology industry in flux. The

networking industry, as well as many companies, wanted to use the standards they

themselves had chosen, which were often proprietary, rather than accept the interoperable

standards that made interconnected networks and even open markets possible (DeNardis

2009, 38; Kelty 2008, 144). Convincing them to set aside their commercial rivalries and

build functioning, testable products that were also compatible with one another (as opposed

13

to creating competing, proprietary systems to "lock" customers into specific products and

associated support resources) was both a political and a technical feat. Yet Interop's

approach proved persuasive because, in order to participate in the trade show, Interop

required vendors otherwise uninterested in the success of Internet per se to connect their

products to the show network. Lynch and the other researchers leveraged their

considerable influence to encourage commercial networking companies to work together to

address substantial inter-networking challenges in an experimental research setting. For

vendors the hybrid setting afforded them the privacy to take risks and make mistakes away

from the competitive pressures of the marketplace.

A Note on Methodologies

This thesis builds on analytical frameworks that examine how people and things can be

translated into forces that shape society and technologies (Pinch and Bijker 1987; Turner

2006; Abbate 1999; Callon 1987), and focuses in particular on the social processes through

which a diverse set of interests can be recruited and brought into alignment. By doing so,

this analysis shifts away from an emphasis on protocols and standards as purely technical

and instead considers the expansion of technologies across domains as a complex process of

"translation" that is as much social and organizational as technical. Drawing on Janet

Abbate's definitive history of the Internet, this study demonstrates how the "kinds of social

dynamics that we associate with the use of networks also came into play during their

creation" (1999, 4). In particular, this study traces the practices and processes, which

include demonstrations and trade show exhibits, that reveal the visions that bound various

actors working to scale technologies (Nye 1994; Flichy 2007), and also the organizational

achievements that helped coordinate new methods of management that established

processes of coordination between different actors (Callon 1986; Thrift 2005; Yates 1993).

Most significantly, this examination builds on Turner's concept, mentioned earlier,

of "network forums": texts and experiences where a varied set of players meet to

collaborate, exchange ideas and legitimacy, integrate new networks, and envision

themselves as a part of a single (albeit distributed) community assembling a global,

seamless, and fundamentally liberalizing information economy and accompanying

information society. Turner's work traces what he terms the Whole Earth network, an

intertwining of the military-industrial research world's culture and the American

14

counterculture that helped shape the public understanding of computers and computer

networks as tools for personal expression and the creation of new social frontiers. To do so,

Turner links two theoretical perspectives from science and technology studies - in

particular Star and Griesemer's "boundary-object" concept, referring to objects that

circulate between several different social worlds but are independently meaningful for each

world - as well as Peter Galison's "trading zone," sites where representatives from various

disciplines come together to exchange ideas and collaborate, establishing "contact

languages" that facilitate shared understandings and collaboration. For example, Turner

argued that core members of the Whole Earth network came together to help create Wired

magazine, a prototype of the utopian society that networked computing would make

possible. MIT's Nicholas Negroponte used Wired as a site to claim that the Internet was

about to "flatten organizations, globalize society, decentralize control, and help harmonize

people" (1995). Turner has also argued that, by the late 1980s, the Whole Earth network

functioned as a vehicle that reinvigorated the influence of the cooperative practices and

systems rhetoric of the military-industrial research world's culture in the corporate sphere.

In turn, this worked to more deeply integrate countercultural utopian visions with the

massive economic and technological forces already reorganizing the industrial world.

Expanding on Turner's framework, this study attends to the guiding visions that

mobilized multiple communities, persuading them to undertake the work of assembling the

physical networks necessary to transform the Internet into a global commercial

infrastructure. As Wiebe Bijker has noted, a technology's successful expansion is as much

dependent on these shared visions as on any qualities or affordances that technologies

might themselves possess (1997, 15). Leo Marx (1964) has termed this a "technological

sublime," referring to the notion that from new technologies would flow social and moral

progress that would liberate the human spirit and improve society. Others have written

about this imaginary; David Nye (1996) on the first transcontinental railroad, Carolyn

Marvin (1990) on electricity, Susan Douglas (1986) on wireless and the invention of

American broadcasting, and more recently Patrice Flichy (2007) on the early Internet and

Chris Kelty (2008) on the practices of the distributed collaborative creation and distribution

of software source code. 9 Kelty has suggested that proponents of these practices "mix up

operating systems and social systems" and are driven by "imaginations of order that are

9 These practices are generally referred to as Free Software, or the Free Software Movement.

15

simultaneously moral and technical" (2008, 43, 9).10 Here, Charles Taylor's work on social

imaginaries becomes useful as it recalls "the ways in which people imagine their social

existence, how they fit together with others.... [It] draws on our whole world, that is, our

sense of our whole predicament in time and space, among others, and in history" (2004, 23,

28).

This research also examines the mobilization of network engineers as "system-

builders" (Hughes 1983),11 that is, they thought about their work constructing physical

networks not only in technical but also in social and economic terms. They focused in

particular on "project management" styles that emerged from the highly collaborative and

interdisciplinary work style and entrepreneurial sensibility of the military-industrial

research world. Through a variety of efforts, engineers enacted these visions by imposing

protocols, the internal logic of networks, and the expansion of those protocols through

flexible partnerships and a system of coordination. Understanding this "routinization of

innovation" (Thrift 2005, 7) has been greatly helped by JoAnne Yates' (1989) work on the

ways in which the first data processing machines led to the development of communication

systems. She has suggested that normalization occurred as management conveyed

procedures and rules to coordinate processes at lower levels and as communication flowed

upward in the form of data and analyses. As Alexander Galloway (2004) has shown in his

research on protocols, the Internet's community's codification of these technical standards

(which comprise the core functionality of the Internet) through the Request for Comments

(RFC) process suggests the importance of also examining the operational logics at the core

of complex technological systems like networks. In essence, the complex interactions

required to build such systems reveal the ways in which standards fully realized operate as

socially constituted values at every level.

Roadmap

The Internet is a complicated tangle of technologies and practices that are under constant

construction and defy easy analysis. Its history is no less complex. This study focuses on

what might be learned about the Internet's commercial transition by considering how the

10 Kelty has described this "social imaginary" as one that is shared between the individuals thatwork to create and build Free Software and "defines a particular relationship between technology,organs of governance (whether state, corporate, or nongovernmental) and the Internet" (2008, 12).11 Similarly, these engineers have been termed heterogeneous engineers (Law, 1987)

16

network engineers and entrepreneurs, members of the Interop network, and many affiliates

of the military-industrial research world, focused on the implementation and expansion of

the TCP/IP core networking protocols. To do so, they forged strategic alliances with

commercial interests. This study extracts one analytic narrative of the Internet's

emergence as a global and commercially viable communications medium. Since

infrastructural network development operates across multiple registers (Law and Callon

1992; Jackson et al. 2007; Bijker, Hughes, and Pinch 1987), this examination links the

"micro stories" of individual actors to the teams of Interop network developers as well as to

larger social processes around the emergence of the Internet. Carl Malamud provides a

through-line. He was deeply involved in the construction of computer networks in the 1980s

and 1990s and was an articulate promoter of the visions that helped drive network

construction, and also of a vision of the emerging networked society. Even so, this analysis

is not intended to be a biographical account of Malamud, or to recapitulate the entirety of

Malamud's projects in the first half of the 1990s. Many studies of the networked computing

infrastructures, and of the Internet, emphasize the innovations of Internet practices and

processes. Since, in most cases, the individuals I interviewed are still actively working in

the information technology industry (see Appendix A), and belong to groups that actively

maintain their own versions of events, some will doubtless disagree with each other, and

with the history that I have constructed.

Chapter One explores mobilizing visions as a critical element in the standardization

of the Internet. Standardization is often primarily thought of as a technical, and therefore

socially neutral, process of change. This chapter examines the more purely social and even

"commercial" aspects of achieving wider agreement on standards, focusing in particular on

idealized visions around emerging technologies and on the challenges of enacting those

visions in the midst of larger technological and economic reorganization in the global

economy. To do so, this chapter explores the Epcot theme park's "Spaceship Earth," an

exhibit that presents a corporate futurism inspired by cybernetic visions of

interconnectedness. It traces one aspect of the Internet's transition from a research

network into a commercially viable global infrastructure, driven by frames of connectivity

and modifiability.

Chapter Two turns to the practices by which network engineers affiliated with the

Interop trade show assumed the role of "system builders" of the physical networks, and

17

thus architects of the emerging networked society and economy. Mobilized by visions of

global connectivity and their imagined intellectual connection to the makers of earlier

modern technological systems, they helped drive the consensus and collaboration required

for the construction and assembly of a global Internet.

Chapter Three focuses on the Interop trade show itself, focusing in particular on the

semi-annual event's network, one of the most complex in the world, that functioned as a

demonstration of the emerging global Internet. This construction not only helped mobilize

engineers and vendors around Internet standards and practices but also functioned as a

hybrid research and development site that coordinated collaboration and partnerships

between representatives of a range of interests, many of whom were also fierce competitors.

Assembled by a core group of researchers with strong ties to the military-industrial

research world, Interop attended to the practicalities of implementing the Internet's core

technical standards while also negotiating powerful commercial needs as well as the larger

economic and technological forces sweeping the industrialized world.

In Chapter Four, the analysis shifts to an affiliate of Interop, Carl Malamud, and

the yearlong Internet 1996 Exposition that he conceived and produced with ample support

from the Interop Company itself. This analysis opens with Malamud's growing interest in

the ways in which better connectivity and faster networks might lead to new services and

uses, and ultimately new communities of users and consumers. A showman-intellectual in

the spirit of Marshall McLuhan, Malamud developed his exhibition in the spirit of a

"world's fair," a metaphor that reflected his preoccupation with the development of earlier

technological systems, especially railroad transportation, that promoted a particular vision

regarding the latent tension between privately managed communications systems, public

access, and the "politics of the commons." This project was realized through a series of

offline and online events, a website (http://park.org) aggregating numerous pieces of online

material, and a coffee-table book chronicling the exposition from inception through the

launch and conclusion of the event. Paradoxically, although many people do not consider

the exposition to have been a success, commercially or otherwise, it can still be looked to as

an alternative vision of how the networks that comprised the Internet might have

continued to develop and as a critical record of the models and discourses that existed

around Internet infrastructures.

1s

Together, these chapters attend to an aspect of Internet expansion and

commercialization that has been largely overlooked in historical accounts to date. They take

seriously the challenges of translating utopian visions into commercially viable technologies

and infrastructures, and in the process, interrogate a widespread assertion that the

Internet was largely developed in the academic world that existed apart from larger

economic forces. The Internet represents significant technical achievements. This study

focuses on the degree to which technological systems must be consciously created in order

to be successful at scale. At the heart of this research, then, lay questions about the

influence of the military-industrial research world and how particular technical visions and

practicalities shaped the Internet as it transformed into a commercially viable global

infrastructure. How did the computer engineers and entrepreneurs building computer

networks employ organizational strategies and alliances that helped ensure the Internet's

place in the global landscape? How did discourses around testability and connectivity

reflect their efforts to shape the emerging information landscape? How might the Interop

trade show have functioned as an important site of negotiation for developers who worked

to shape these critical discourses, and, in the process, ensure the commercial success of the

Internet? This research suggests that the global success of the Internet should be attributed

to the reemergence of the collaborative work styles and systems rhetoric of the military-

industrial research culture into the commercial sphere.

19

Chapter One: As Our Thirst for Knowledge Grew, the World Began to Shrink:Epcot's Spaceship Earth as a Networked Utopia

Numerous theories of technological change have portrayed the form and function of

technologies as determined by the cultural values, interests, and interpretations of social

groups (Bijker, Hughes, and Pinch 1987; Bijker 1995). Among the concepts introduced is

that of "interpretive flexibility," a process suggesting there is no one, or best, way to

construct a technology. Rather, a technology's design and use is flexible. This view

emphasizes how different social groups examining the same technology will not only

identify distinct technological problems, but also present distinct solutions to these

perceived challenges. In "Inventing the Internet," Janet Abbate suggests that the "TCP/IP

protocols, gateways, and uniform address scheme were designed to create a coherent

system while making minimal demands on the participating networks" (1999, 219). These

"minimal demands" gave the Internet, as a locally successful technological system, the

flexibility to survive commercial and political pressures as the system expanded among new

users and into new geographic areas. Yet Abbate also suggests that the very success of the

TCP/IP protocols refutes the general assumption that technical standards are socially

neutral, establishing that "standards can be politics by other means" (1999, 179).12 In

particular, computer networks and inter-networks were designed according to various

technical specifications that revealed distinct operational logics. The Internet's core

networking protocols reflected the values of the social groups that emerged from Cold War

military research culture, an environment that fostered practices that were not only highly

collaborative but also interdisciplinary and entrepreneurial in spirit.

These values continue to infuse the Internet today. In fact, that networked

computing in the early 1990s often did not operate "as advertised" is an irony that reveals

the deeply social nature of protocols. 13 Only as networked computing became tied up with

utopian visions of empowered individualism and a meritocratic marketplace, did it become

technically possible to redeem its promises. Abbate's analysis of this process largely focuses

on technical objects, yet she might also have usefully examined the more purely social and

12 Abbate's work is a definitive account of the history of network protocols through individualdevelopers, U.S. Department of Defense mandates, and international standards conflicts.13 For a sense of the "state" of the Internet in the early 1990s, see the Computer Chronicles episodeon "The Internet." Computer Chronicles was hosted by Steve Cheifet and produced in San Mateo,California by KCSM-TV. http://www.archive.org/details/episode 1134.

20

even "commercial" aspects of achieving wider agreement on standards. These aspects

include, as we shall see, encounters at trade shows and exhibitions.

Paul Edwards contends that constructing and maintaining standards is a complex

process interwoven with social practices:

Ideally, standardized processes and devices always work in the same way, nomatter where, what, or who applies them... Most standards also involvediscipline on the part of human participants, who are notoriously apt tomisunderstand and resist. As a result, maintaining adherence to a standardinvolves ongoing adjustments to people, practices, and machines. (2004, 827-828)

Thus, even the process of getting the core structures of the Internet to "work" elicited the

"ongoing adjustments" needed to create a coherent, effective research network. Trevor

Pinch and Wiebe Bijker point out that the social environment shapes the technical

characteristics of technologies, and emphasize the critical role that social groups play in

defining and addressing problems during a technology's development. A technology can be

considered stabilized once consensus emerges and "the social groups involved in designing

and using technology decide that a problem is solved" (Pinch and Bijker 1987, 12). Since a

technology is not a fixed object per se but rather emerges amidst interactions with

numerous social groups, this process of "closure and stabilization" occurs numerous times

(and even continuously) as a technology is developed, expanded, and improved (Pinch and

Bijker 1987, 17-50). This characteristic suggests that although technological change may

appear to follow a linear path (even if appearing as a disruptive force), the process is in fact

more nuanced. The tools of standardization, namely the technologies, organizational

solutions, and/or inter-connection protocols, also function as "gateways" that make it

possible to transfer technical as well as social, and cultural practices across otherwise

incompatible domains:

Standardization in its various guises (formal and informal, top-down andbottom-up) is perhaps the leading example of a gateway technology on thesocial/organizational side ... It is at this point of heterogeneous connectionamong systems that the eventual power, scope, and world-building quality ofinfrastructure begins to take shape. (Jackson et al. 2007)

This quality recalls another aspect of standardization: the degree to which it favors the

politics and practices of a specific group of actors to the exclusion of another. In other

words, a technology has certain attributes because inventors design a technology to express

their personal visions and desires. Understanding what is required to standardize a

21

technology becomes a critical part of tracing the technical, organizational, and political

negotiations and adaptations that were necessary for the Internet to become more widely

successful. Many of the same qualities likely helped the Internet as it scaled beyond a

locally constructed system and expanded into other domains, linking with other networks to

emerge as a commercially managed global information infrastructure. Deploying

technologies required the mobilization of network engineers and technologists, such as

those affiliated with Interop, who shared a vision and collaborative methods of making

meaning.

In the standardization and expansion of communications networks, technologies

have physical qualities that are central to how they operate locally or as part of larger

infrastructures. Modern infrastructures are technical systems-say, transportation,

telecommunications, or energy-that rationally engineer the world and order it in a way

that facilitates the circulation of goods and ideas. They are also conceptual, cultural devices

that are powerful as a mode of regulating societies by "publicly performing the relations

between the individual and the state" (Larkin 2009, 245) while at the same forging

architectures of the sublime that join the technological with our imaginations and notions

of progress. The infrastructure of computer networks appears to function in another

manner altogether, in a kind of chaos, an unpredictable structure without a center.

Although the distributed and flattened organizational structures of computer networks

appear to resist control, they are in fact governed by a particular logic that functions as a

form of management. These mechanisms, which are deeply imbedded in the free market,

deregulation, and enterprise, drive partnerships with the promise of "openness" and

"connectivity" that occurs through the global integration of the networked information

technologies.

To begin to explore the complex questions around technological change and the

particular visions that drove the physical construction of the Internet, it is worthwhile to

set the stage by visiting one part of a vast realm that is "commercially viable" while being

wholly a product of an alternative utopia "embodied" in the animated image of a talking

mouse. Under his patronage, we are offered a view of a high-tech utopia dominated by

benign corporate sponsorship and the guarantee of technological progress.

22

Spaceship Earth

Walt Disney's Spaceship Earth exhibit, located in Orlando, Florida's Epcot theme park,

presents an "Animatronic" tour of the history and future of communications. The Disney

exhibit exists as a series of dioramas - a cinematic recycling of the past cast as iconic

moments of technological achievements - strung together as a narrative of progress that

draws visitors into a future that is already upon them. Spaceship Earth and enterprise

computer-networking trade shows share little with one another in terms of operational

logics and visions, yet by way of this ambivalently defined relationship between

computational technologies, corporate interests, and individual agency, the two operate in

critical tension with one another. Each powerfully evokes the idealism and attention toward

social and moral progress that has infused technological innovations in the U.S. since at

least the 1 9 th century (Marx 1964).

The term "mobilizing utopias" will be used here to refer to the implementation of

idealistic models into experimental projects or prototypes, and even the practicalities of

bringing a technology to scale. Epcot realized Walt Disney's vision of an "Experimental

Prototype Community of Tomorrow" (EPCOT), a near-future world inspired by a faith in

the ability of cybernetic information systems and corporations to solve social ills and

advance society more effectively than individuals and democracies. The iconic Spaceship

Earth sphere figuratively anchors the park. It was also Epcot's guiding metaphor, 14 a vision

equally inspired by popular science fiction and the cybernetic information systems of the

military-industrial research world. Promotional material created for the park's launch in

1982 consisted of an illustration of a half-shrouded geodesic "planet" encircled by what

appeared to be a monorail track or the contrails of a rocket. The image's intention is clear:

it presents a "usable future" that has moved beyond the polarizing Cold War and

traditional economics of scarcity to reveal the planet as a globally integrated system

connecting all living things to a future of ever more efficient technologies (Deese 2009, 1-2;

Turner 2006, 56-58) (see Figure 1). The layout of the park itself is divided between "Future

World" and "World Showcase" pavilions, presenting spectacular displays of technological

innovations and cultural identities. In a style that first became popular with the 1939 New

14 "Spaceship Earth" is most often associated with inventor R. Buckminster Fuller, who publishedOperating Manual for Spaceship Earth in 1969; see also 1996 works by Barbara Ward and KennethBoulding. Science fiction writer Ray Bradbury wrote the original narrative for the Spaceship Earthexhibit.

York World's Fair's "World of Tomorrow," Disney's simulated landscapes market the idea of

progress itself, brought by corporations whose "expertise would create a harmonious world"

(Nye 1994, 213). Corporate sponsors support each exhibit, entertaining visitors with

glimpses of technology-infused prototypes of future worlds packaged for middlebrow tastes

(Bukatman 1991, 56; Nye 1994, 199-224).

Figure 1: Advertisement for the October 1, 1982 launch of EPCOT theme park. A BuckminsterFuller-inspired Spaceship Earth dominates this illustration, underscoring the cybernetic influenceon Walt Disney's futuristic visions of the ideal society.

"Mobilizing utopias" as a concept also implies the complex and often contradictory

processes of transforming emergent technologies into everyday tools that support and even

shape modern lives. The Spaceship Earth exhibit, like the Interop network, (re)negotiates

the transition between the first stage of exploration, a period of innovation and glory, and

the next, when emerging technologies become familiar, practical, and even invisible. The

24

term also suggests a future that is technologically intense, and inevitable - even already

upon us (Kelty 2008; Bukatman 1991, 59). In New Rules for the New Economy, one of the

most widely read business manuals of the 1990s, the executive editor of Wired magazine,

Kevin Kelly, celebrated a new order in which "the world of the soft-the world of

intangibles, of media, of software, and of services-will soon command the world of the

hard-the world of reality, of atoms, of objects, of steel and oil." While Kelly expressed a

faith in technological progress, he also took as a given that there would be losers in the

passage to an inevitable future in which "those who play by the new rules will prosper,

while those who ignore them will not. We have seen only the beginnings of the anxiety, loss,

excitement, and gains that many people will experience as our world shifts to a new highly

technical planetary economy" (Kelly 1998, 2).

In late 2007, Spaceship Earth underwent the first substantial renovations in more

than a decade.1 5 Since Siemens AG had assumed sponsorship of Spaceship Earth in 2005,

the exhibit received new signage, an updated narrative including interactive video screens

installed in the exhibit trams, and new scenes depicting computers and computer networks,

as well as a redesigned post-show exhibit space. As it always has, the exhibit consciously

draws on metaphors and iconic moments propagated in popular society and repackaged by

a corporate entity. The first half of the exhibit has generally remained the same since

Spaceship Earth first launched in 1982. Visitors board motorized trams fashioned as "time-

machine spaceships" in a fog- and lightning-filled "dawn of recorded time" before ascending

on a spiral track past dioramas depicting historical technological milestones as well as

idealized near-future scenarios. The first dioramas depict early man scrawling mammoth

figures on cave walls; later ones, Egyptians creating papyrus scrolls and Greeks

establishing schools and the study of mathematics. The ride skips forward to the invention

of the printing press and the subsequent flourishing of culture during the Renaissance,

then to the Age of Invention and with it the telegraph, telephone, radio, motion pictures,

and television. Finally, it moves on to the era of space travel, satellites, and computing

technologies.

The Spaceship Earth ride takes visitors past a family sitting around a television set

that is showing Neil Armstrong's first footsteps on the moon. The enduring collective

' Martin's Videos blog. Spaceship Earth 2007 - Ultimate Tribute. http://www.martinsvids.net/?tag=spaceship-earth.

25

memory of this technological achievement dwells on American's global supremacy, and

glosses over the Cold War politics and pervasive threat that infused this period. To explore

the unknown geographies of space, the narrator explains, "Society had to invent a new

language" of computation, represented in the next tableau as the banks of blinking lights

and speeding magnetic tapes of a room-sized computer. There is little sign of the

government-sponsored research that drove the invention and development of computing

technologies; instead, the Spaceship Earth narrative places these achievements within

market logics. "In 1977, young people with the passion for putting computers in everyone's

hands,"16 the narrator suggests, helped to miniaturize and personalize once-massive

machines. As computer networks have grown increasingly ubiquitous, the 2007 version of

Spaceship Earth suggests, humankind has become seamlessly integrated into the network,

part of a truly global community. Riders are driven onward into the future, through a

green-hued data stream tunnel, before arriving in outer space, suggesting the frontier of

the future. Framed by an image of planet Earth on the horizon, visitors design their own

futures by answering a series of multiple-choice questions on interactive LCD touch screens

mounted on the trams. When the ride ends, visitors can enter "Project Tomorrow: Inventing

the Wonders of the Future," the post-show exhibit. Here, interactive games showcase

Siemens technologies, including medical devices, transportation, and energy management

systems. A "Spaceship Earth Online" website has also been added.

These latest renovations to Spaceship Earth might have gone largely unnoticed -

both Epcot and the exhibit itself are nearly three decades old and both offer visions of

technological innovation that have always tilted toward the mundane - except that in the

days before the refurbished exhibit opened, one of the new scenes in Spaceship Earth

sparked interest. On December 2, 2007, the technology blog Boing Boing reposted a rumor

that the renovations included a diorama of the California garage where marketing whiz

Steve Jobs and computer programming genius Steve Wozniak co-founded Apple Inc. in the

late 1970s, though showing only Jobs. 17

As the rumor spread through the blogosphere, what seemed to arouse the curiosity

of a number of readers was not why this scene had been chosen but rather which history it

16Apple was incorporated in 1977.17 Doctorow, Cory. 2007. Steve Jobs (and not Woz) to come to Epcot's Spaceship Earth?? Boing Boing.December 2. http://www.boingboing.net/2007/12/02/steve-jobs-and-not-w.html.

26

emphasized. 18 Many disapproved of Disney's apparent decision to include Jobs, the current

CEO of Apple and Disney's largest shareholder, and exclude Wozniak. Some speculated

that Jobs had used his considerable influence to garner a top spot in Disney's

communications exhibit. One Boing Boing reader observed, "Jobs? Better then [sic]

[Microsoft's Bill] Gates, I suppose." Another commented, "Jobs never influenced anything

until later on when the first Mac was being made. That's when the first of his visions

started to be seen (closed system, no expansion, etc.). ... Jobs influences products today and

does so with a near 100% record of success, but to suggest that he was the primary brain

behind the personal computer revolution (i.e., the garage intro of the Apple computer) is a

huge untruth and deceptive." 19 When Disney reopened the exhibit a few days later, it had

indeed recreated many aspects of an early press photograph of the Apple co-founders,

except it was Jobs, not Wozniak, who was excluded from the scene. A mechanical likeness of

Wozniak sat in a garage-turned-office in front of what resembled a prototype of the Apple II

computer, a machine that was evidently meant to stand in for a number of Apple's early

advancements that helped turn the start-up into a successful business. Boing Boing posted

an update, yet few readers commented on Jobs' absence from the Disney exhibit.2 0

Jobs and his Apple engineers translated utopian ideals into computer design by

replacing highly technical keyboard commands with a radically new graphical interface

that included easy-to-use point-and-click systems. 21 Jobs' marketing genius helped produce

the iconic "1984" advertisement that introduced the Macintosh personal computer,

18 The significance of the Wozniak and Jobs' different perspectives has been the subject of ongoingdiscussions between technologists about the nature (and future) of the Internet, and generallyframed as a tension between technologies that encourage experimentation and exchange, and onesthat offer little in the way of participation but whose closed functionality makes them moreaccessible - and marketable - to a wider public. For example, in The Future of the Internet-AndHow to Stop It (2008), Harvard Law professor Jonathan Zittrain predicted that the Internet is .increasingly likely to become a "closed" technology, and used the "iPhone" as an example of what hedescribed as Jobs' determined effort to replace the personal computer with consumer "appliancestethered to a network of control." (3) The release of the "iPad" in 2010 further inflamed tensionsbetween Jobs and the "Internet community."19 Doctorow, Cory. 2007. Steve Jobs (and not Woz) to come to Epcot's Spaceship Earth?? Boing Boing.December 2. http://www.boingboing.net/2007/12/02/steve-iobs-and-not-w.html.The technical history of personal computing is also obscured, both in the exhibit and in the readercomments on Boing Boing. For example, Apple's first personal computer with a graphical userinterface (GUI) was the "Lisa," not the "Mac" as one reader suggested.20 Doctorow, Cory. 2007. Animatronic Steve Wozniak comes to Epcot Center ride, animatronic SteveJobs nowhere in evidence. Boing Boing, December 9. http://boingboing.net/2007/12/09/animatronic-steve-wo.html.21 This built on the work of Xerox's Palo Alto Research Center.

27

portraying the device's arrival as an unnamed heroine defeating the dehumanizing

bureaucracy of the corporation. Author Steven Levy later described the release of the

Macintosh computer as one that moved digital worlds out of "the arcane realm of data

processing and science fiction. After Macintosh, it began to weave itself into the fabric of

everyday life. Macintosh provided us with our first glimpse of where we fit into the future

... [It] brought just plain people, uninterested in the particulars of technology, into the

trenches of the information age" (1994). Levy focused on the social and technical visions

inscribed into the computer itself, although this vision of an empowering and intensely

personal technology was soon extended to include computer networks.

Yet Jobs' rumored inclusion in the Spaceship Earth exhibit stirred controversy, in

all likelihood, because technophiles actively and consciously maintain a utopian vision

about how the Internet came to have its present order and how it should be ordered in the

future (Kelty 2008; Streeter 1993, Abbate 1999, Turner 2006). Chris Kelty suggests that

these protective behaviors relate to ideas around openness and collaboration on the

Internet,2 2 and that individuals work together to defend the network's "legitimacy and

independence ... not only from state-based forms of power and control, but from corporate,

commercial, and non-governmental power as well" (2008, 9). These social practices around

openness have also flourished outside of technical communities, informing shifts in

intellectual property, music, films, databases, and education. One of the most powerful

demonstrations of this ethos is "Creative Commons," an alternative method of issuing

copyright licenses that allows for sharing information. Yet the assumptions implicit in

these practices - namely that the Internet functions as a tool that citizens use to

collaborate, share, create, and distribute knowledge-in a way that reorients power and

knowledge-do not represent the only, or earlier, articulations of "openness."

As a number of scholars have noted, when the Internet first emerged in the 1980s

and early 1990s, the most prominent mechanisms and logics emerging around networked

computing were intensely focused on openness as achieved through free market promotion,

deregulation, and privatization (Kelty 2008; Streeter 2003; Turner 2006). The rapid

integration of computing and telecommunication technologies into the international

22 Kelty defines Free Software as a "set of practices for the distributed collaborative creation ofsoftware source code that is then made openly and freely available through a clever, unconventionaluse of copyright law" that also "exemplifies a considerable reorientation of knowledge and power ...with respect to the creation, dissemination, and authorization of knowledge" (2008, 2-3). FreeSoftware is also known as Open Source Software as well as FOSS or FLOSS.

28

economy had created a "new economy" that brought with it more flexible corporate

organizations and a greater emphasis on entrepreneurs and "knowledge workers" (Turner

2006; Thrift 2005). As we shall see in later chapters, these changes promised to transform

America - and indeed the world.

Through the likeness of Steve Wozniak, the 2007 version of Spaceship Earth

reframes these shifting technical visions to suit its more corporate one. The exhibit

highlights Wozniak in 1977, the year that he co-founded Apple. Wozniak is romanticized

among technophiles for hacking massive Cold War-era research computers and integrating

miniaturized versions of them into everyday life. The Animatronic version of the scruffy

programmer reflects the hippie/hacker ethos of the 1970s. Yet other aspects of the scene

suggest that it actually shares more in common with the techno-utopian politics and

market populism that emerged alongside the Internet and then the World Wide Web in the

1990s. The exhibit designers accomplish this through an assemblage of iconic, and

historically incongruous cultural markers. The garage has become a well-known trope for

Silicon Valley-based entrepreneurialism, championing countless individuals and corporate

enterprises that helped to integrate computing and telecommunications technologies into

national and international economic life. Likewise, an issue of Wired magazine - which

began publication just as networked computing reached public consciousness - was

prominently displayed among the cans of paint, greasy pizza boxes, and hardware

components in the garage. The magazine, Fred Turner suggests, portrayed the Internet as

"a prototype of a newly decentralized, nonhierarchical society" and depicted computing

industry and telecommunications executives as the engineers constructing the social

infrastructures of this new world (Turner 2006, 208). In the end, Wired and the entire

garage scene become a chapter in the larger narrative of Spaceship Earth, which celebrates

individualized access to global networks.

Conveniently (for the present study), the last substantial renovation of Spaceship

Earth was completed in 1994, in the midst of the same massive economic and technological

restructuring that the current version of the exhibit now references. 23 Perhaps the two

metaphors that best connote the massive shifts that occurred in this era are the terms

23 "EuroTraveler" blog, Remembering Walter Cronkite at Spaceship Earth at Walt Disney World.http://www.zimbio.com/Epcot+Center/articles/cJhlIF75ziA/Remembering+Walter+Cronkite+Spaceship+Earth.

29

"information and control system" and "information superhighway." From the launch of the

Spaceship Earth exhibit in 1982 until 1994, the final scene of the exhibit depicted a control

room with a global map monitoring networks worldwide (see Figure 2). For much of this

time, signage within the Spaceship Earth scene identified this control room as the "AT&T

Network Operations Center." AT&T, the world's largest telecommunications provider,

sponsored Spaceship Earth for twenty years (1984-2004). The exhibit was designed to

reflect Cold War visions of the planet as a "closed world" (Edwards 1996), an information

system bounded by the militaristic metaphor of global information and control. This scene

deployed a version of this metaphor, tweaked to represent AT&T's vision of a single

corporate communications and computing empire (Warf 1998, 257).

Figure 2: The AT&T Network Operations Center scene from Spaceship Earth, circa 1984. This scene

illustrates the corporate giant's explicit reliance on Cold War-era framing that computing

technologies promised global technological oversight.

I0

Likewise, the metaphor of the "information highway"24 represented the tensions

between mobilizing utopias, corporate interests, and government control. First, there was

envisioned a high-speed, high-capacity fiber-optic network provided by established

telecommunications institutions to offer interactive television, movies-on-demand, and

telechat (Flichy 2007, 18-20). By 1991 the strategy to improve the country's

communications infrastructure was envisioned as a high-capacity, fiber-optic network that

would drive future economic competitiveness as well as provide information and services to

citizens. During Bill Clinton and Al Gore's 1992 election campaign and victory, the

"information highway" became a concrete program: the state would finance and build a

national fiber-optic network while the private sector (under public sector supervision)

would provide the services (Markoff 1993). As it had in the construction of the interstate

highway system in the 1950s, the government would be a key player in the emerging

knowledge economy, providing traditional investment in public infrastructures, as well as

additionally providing high-tech research programs that previously had been funded by

military funding (Flichy 2007, 21). Within a year, however, the Clinton administration had

abandoned its grand technological vision; and the "information highway" became

synonymous with the telecommunications liberalization of the 1980s. According to Flichy,

the vision had become reductive: "democracy = information highways = deregulation. In

this sequence of translations, the first relates to an idea of technical determinism (a new

technique promotes democracy), and the second to a political choice (deregulation promotes

the construction of that technique)" (2007, 31). John Malone, chairman of one of the first

cable operators, suggested in a 1994 Business Week interview that the government should

be "mainly a cheerleader," that is, relegated to the sidelines. 25

At the International Telecommunications Union Conference (ITU) in Buenos Aires

in 1994, Vice President Al Gore touted the Global Infrastructure Initiative (GII), a private

and international network that promised to bring all the communities in the world

together. "We now can at last create a planetary information network that transmits

messages and images with the speed of light from the largest city to the smallest village on

every continent."26 In his presentation, Gore invoked Nathaniel Hawthorne's vision of

24 This metaphor, incidentally, has been around since the 1970s. For one example, see Ralph LeeSmith's article, "The Wired Nation", published in The Nation, May 18,1970.25 Malone, John. 1994. Business Week, January 24: 89, quoted in Flichy, 2007, 30-31.26 Nash, Nathaniel C. 1994. Gore Sees World Data Privatizing. New York Times, March 22.

21

nearly one-hundred-fifty years earlier, that a global telegraph system would transform the

world into a vast "brain" whose "nerves" would link all human knowledge. Gore continued,

"to accomplish this purpose, legislators, regulators, and businesspeople must do this: build

and operate ... information superhighways on which all people can travel."

In the wake of telecommunications deregulation in the 1980s, AT&T followed a

similar plan, expanding aggressively overseas, ventured outside of the traditional

telephony market, and focused increasingly on global computing hardware and

telecommunications equipment like fiber-optic cable, switching and routing systems, and

computer chips (Warf 1998, 258) (see Figure 3).

In response to substantial social and economic shifts, as well as AT&T's focus on

other markets, the simulated landscapes of Spaceship Earth now offered a boundaryless

vision of the world. Spaceship Earth had adapted to shifting popular visions of networked

computing. AT&T no longer dominated the exhibit, although their corporate interests

remained central to the depiction of ubiquitous networked computing. In the 1994

renovation of the exhibit, the "Networked Operations Command" was replaced with several

new scenes. One featured a woman sitting in front of a monitor in a darkened office,

revealing computers and computer networks as technologies that "integrate the individual

ever more closely into the corporation" (Shoshanna Zuboff 1988, quoted in Turner 2006, 2).

Another depicted an American boy communicating via video screens with a Japanese girl,

their exchange linked via fiber-optic "highways" of light that leapt across cities and oceans.

'2

Gore, Al. 1994. Global Information Infrastructure Speech. Presentation at the InternationalTelecommunications Union, March 21 in Buenos Aires, Argentina. http://www.interesting-people.org/archives/interesting-peole/99403/msgOO1 12.html.

Figure 3: AT&T's International Fiber-Optic Cables, circa 1998 (existing or in progress). Compiled byBarney Warf of Florida State University from data on AT&T website, http://www.att.com.

Cultural theorist Scott Bukatman has suggested that Walt Disney World functions

as a kind of virtual reality that, behind user-friendly interfaces, conceals technologies of a

"fundamentally conservative and historically bound vision of 'the future"' (1991, 73). These

interfaces, Bukatman continues, are analogous to the structures of a computer system, from

the rides and attractions, or files, to the pervasive transportation systems, or operating

systems. Bukatman's emphasis on the computational qualities of Epcot suggests that

Spaceship Earth can be further thought of as a content management system capable of, as

the exhibit website currently suggests, storing and organizing the history of "human

connection and collaboration over 40,000 years."27 As such, the exhibit recasts disparate

historical achievements within the modern technological era until finally humankind's

various historical narratives function as nodes in a massive computer network. Time,

culture, and space collapse into a universal utopian present so that ancient Greeks, Islamic

scholars, Western monks, and American scientists simultaneously work to advance global

communication systems, and, ultimately, networked computing. The ride narration in the

2007 renovation suggests that Romans constructed a system of roads to move their armies

around, thus "creating the world's first World Wide Web"; ten years earlier the Romans'

system of highways had been a metaphor for the "information superhighway."

27 Walt Disney World in Florida. Spaceship Earth attract at Epcot. http://disneyworld.disney.go.com/parks/epcot/attractions/spaceship-earth/.

22

When the renovated Spaceship Earth exhibit opened in 2007 - "relaunched," in

Disney parlance - longtime Walt Disney "Imagineer" Bob Zalk (2008) suggested the

changes to the exhibit constituted a substantial shift in Spaceship Earth's representation of

computers and computer networks. "The old story of Spaceship Earth was the history of

communications. The new story is each generation invents the future for the next

generation." 28 In typical Disney fashion, Spaceship Earth presented this "new story" of

networked individualism through spectacular (and historically incomplete) moments - such

as humankind's forays into space and the public emergence of the Internet and the World

Wide Web - in a manner reminiscent of twentieth century world's fairs. Epcot's simulated

landscapes and future worlds were first built to replicate many aspects of the 1939 New

York World's Fair (Bukatman 1991). Epcot's exhibits position technologies within a larger

narrative of historical progress such that the technology itself, as a consumer good and

artifact of an ideal future, becomes "the last act in a scientific drama" (Nye 1994, 220) that

is an inevitable - and distinctively American 29 - achievement. As in previous versions of

the exhibit (and indeed at world's fairs beginning in the 1930s), this narrative is a sanitized

one that approximates the anticipatory excitement of frontier exploration even as it lacks

the uncertainty and risk of invention. Although Spaceship Earth's interactive screens invite

viewers to "invent the future" (as Zalk suggests), this effort actually casts visitors as

consumers that must be "cajoled into modernization" (Nye 1993, 221-222). This recalls

Nye's description of corporations at the 1939 New York World's Fair that constructed a

technological sublime that "sought not to enlighten but to impress and pacify. ... The

spontaneous crowd, which had been one important element of a sublime event, had been

turned into paying spectators, who were told in detail how to interpret the wonders

presented to them" (Nye 1994, 222). In other words, the users who have always actively

shaped the practices and processes of the Internet are nowhere to be found.

For this study, Spaceship Earth's machine-aided futures are most interesting for

how they frame technological systems - the actors and the physical technologies themselves

- in complex modern societies. The exhibit's reliance on popular 1990s-era utopian

28 Zalk, Bob. 2007. Spaceship Earth Re Launches into the Future at Epcot. Siemens AG promotionalmaterial. Orlando, Florida. http://www.metacafe.com/watch/1144320/spaceship earth re launches into the future at epcot/.29 Based on my reading of David Nye, I suggest that the exhibit's reliance on sensory discontinuitiesof the sublime as well as its reliance on technological achievements as "measures of cultural value"are distinctively American.

.4

constructions that networked computing would signal the arrival of an ideal society

disguises the exhibit's related assumptions about the profound relationship between the

American military-industrial research world, market economies (including deregulation,

privatization, and open economies), and multinational corporate interests. These

statements situate the rapid emergence of the Internet as a commercially viable global

system within histories of modern technological infrastructures that invariably lead to

dramatic social restructuring. By drawing heavily on the vernacular elements of world's

fairs, Spaceship Earth underscores the importance of the symbolic dimensions and

discourses (the metaphors, frames, narratives, and enactments) of emerging technological

systems.

It is tempting to argue that the developers who helped transform the Internet into a

commercially viable communications medium had little in common with Epcot. From the

vantage point of the early 21;t century, numerous groups have described the Internet as

reorganizing knowledge and power (Kelty 2008). Yet, as we shall see, the network

engineers assembling the global Internet were often mobilized by what Chris Kelty has

termed "openness through privatization." They understood themselves as architects of the

emerging society and "new economy." This concept refers to the tension that existed in the

1980s and 1990s between idealistic visions of a democratic free market and a concurrent

push toward "openness" and interoperability that was framed as "the freedom to buy access

to any aspect of a system without signing a contract, a nondisclosure agreement, or any

other legal document besides a check" (2008, 150-151). In other words, the Internet

developers in this era tended to have less interest in the freedom to copy and modify but

instead were driven by the practicalities of ensuring the commercial success of the Internet.

Chapter 2: Internet Explorers and Digital Worlds

Silicon Valley comprises sprawling suburbs dominated by corporate landscapes that

seamlessly fade into one another. In the past four decades, this region has been best known

as a locus of innovation, entrepreneurship, and extraordinary economic growth. Its success

can be attributed to the numerous forums that brought together individuals from different

companies and organizations, from the public and private sectors, and from academic and

educational institutions. These encounters encouraged allies and competitors alike to

discuss common problems and consider solutions that often helped the interests of

numerous independent firms. These forums also encouraged individuals to form flexible,

innovative partnerships serving a shared recognition of the need to assure the Internet's

global success. This study explores a series of Silicon Valley-based forums affiliated with

the Interop trade show network, an enterprise with direct ties to the highly collaborative

and entrepreneurial Cold War-era military research world, and that network's role in the

commercialization of the Internet.

Over the course of several months, beginning in April 2009, I visited the Silicon

Valley and San Francisco Bay area to conduct a series of interviews with individuals,

almost all engineers, affiliated with the Interop trade show at the height of its influence. In

our conversations, I focused in particular on the artifacts - trade show publications and

research collaborations - that typified the Interop network at the height of its influence. I

also focused on the particular visions that have mobilized programmers and engineers.

The system builders involved in the conceptual and physical construction of the

Internet devoted a lot of time to telling stories and writing about the impact that new

technologies might have on society. As with Disney's Spaceship Earth, these narratives

were often a combination of fact and fiction that helped make sense of the present and

order the future in which the relationship between time, space, and progress would change.

These stories also allowed individuals to legitimize their visions for the emerging utopian

society by making themselves into credible representatives of the communities that they

were helping to build. Turner suggests that members of the Whole Earth network,

including Kevin Kelly and Stewart Brand, did this by using their conversations to turn

"digital media into emblems of network members' own, shared ways of living, and evidence

of their individual credibility" (2006, 7). In her research on computer engineers, Janet

Abbate argues that engineers working to expand and popularize the Internet employed

technical standards and documentation practices that had support within a large segment

of the computer science community (Abbate 1999, 178). Chris Kelty has described the kinds

of stories that computer programmers and engineers tell as "usable pasts" (Kelty 2008, 64-

94) that reflect their ideas about the relationships between "operating systems and social

systems" (2008, 43). Kelty has argued that, for technical actors, these stories are an

important process of "meaning-making" because they occupy a world "finely controlled by

corporate organizations, mass media, marketing departments, and lobbyists" even as they

"share a profound distrust of government regulation" (2008, 72). He writes about the

technical actors affiliated with Free Software, and the particular ways they have

maintained a space for the "critique and moral evaluation of contemporary capitalism and

competition" (2008, 76).

In contrast, network developers in the early 1990s possessed a "double aspect." Like

Kelty's "geeks," the network engineers affiliated with Interop often employed "usable pasts"

that helped them understand their practices in relation to the technical and political

economy of the early 1990s. Yet these visions also focused primarily on the practicalities of

expanding the Internet through the privatization of the physical networks (and later the

establishment of private services), integrating the Internet into the emerging networked

economy. Like Disney's depiction of Wozniak alone in his office, shaping the future,

network engineers held romantic notions of themselves as explorers crafting the prototype

of a future ideal society. At the same time they worked as system-builders (Hughes 1983),

adopting a "managerial ideology" (Flichy 2007, 6) as they operated across multiple

(technical, economic, political, and social) registers to assemble a global information

infrastructure. These resources included not only massive investments in labor and capital

but also a diverse range of interests. The assembly of diverse networks into a singular

infrastructure was a social and organizational feat as much as a technical one. In their

struggle to "work out" their relationship to governance, the global capitalist economy, and

personal liberties, Interop's network engineers actively sought to place themselves in

intellectual connection with the actors of previous technological systems. They often focused

on stories about infrastructures and standardization, integrating their visions into larger

questions about governance and larger global economic flows.

27

Exploring Global Connectivity

I met Carl Malamud in 2009.30 He was at the Tech Policy Summit, an event focused on

issues around regulation, spectrum policy, and America's lagging broadband

infrastructures. Malamud was in attendance to speak about the need for the greater

accessibility of information such as government data and public archives. 31

Malamud has. been an open access advocate for more than twenty years. In that

time, he has taken on not only the Securities and Exchange Commission (SEC) 32 but the

Smithsonian Institution,33 the Government Printing Office (GPO),3 4 and, most recently, the

U.S. federal judiciary. In early 2002, Malamud made unsuccessful bids to run the Internet

Corporation for Assigned Names and Numbers, or ICANN, which handles the most crucial

functions of the Internet, pushing to run it as a public trust.35 For many years, Malamud

was also an author of technical resource manuals who also wrote for industry journals and

Interop trade show publications, explaining complicated networking technologies to a

technical audience. Malamud's projects have almost always been provocations - equal parts

public spectacle and demonstration - that highlight larger technical or social issues and

then offer "work-arounds" to address them. These projects are prototypes that mobilize

actors to imagine themselves at the forefront of an emerging ideal society and offer tools to

manage that change (interview 2009; a list of all interviewees appears in Appendix A).

30 Malamud (born, 1959) had nearly completed a PhD in economics at Indiana University-where,incidentally, he focused on the deregulation of AT&T-when he left to build computer networks inthe late-1980s. The son of a Fermi National Accelerator Laboratory physicist, Malamud becameacquainted with the world of high-energy physics, mainframe computers and other computingtechnologies, and international scientific research [Lausanne/CERNI at an early age.31 Malamud, Carl. 2009. Tech's role in promoting greater government transparency andaccountability. Appearance on panel at the Tech Policy Summit, May 11-13, in San Mateo, CA.http://www.techpolicvcentral.com/media-vault/2009/06/2009-tech-policv-summit-podcas.php.32 In the early 1990s, Malamud took the SEC's corporate filings, which were public documents butdifficult to find, and made them freely accessible and searchable on the Internet. When Malamudlater threatened to close the site, public demand forced the SEC to set up its own site.33 In 2006, Smithsonian Business Ventures, which is affiliated with the Smithsonian Institution,sought to partner with Showtime to create "Smithsonian on Demand." Malamud protested and latertestified before the Senate on the matter. The testimony is available here:http://public.resource.org/smithsonian.html.34 In 2009, Malamud began an online campaign (YesWeScan.org) to oversee the office that publishesdocuments and other publications generated by the three branches of government, in part to drawattention to the need for the government to make public domain information more broadly accessibleonline.35 ICANN has overseen this function since the late 1990s. For news about the contract, seehttp://www.zdnet.co.uk/news/it-strategy/2003/02/10/flak-flies-over-icann-contract-renewal-2130201/.For a copy of Malamud's bid, see http://trusted.resource.org/org-proposal.htm.

In the early 1990s, Malamud published a global survey of the emerging Internet,

entitled Exploring the Internet: A Technical Travelogue.3 6 Dan Lynch and the Interop

Company had funded Malamud's travels, and his published account was distributed to

attendees at the Interop93 conferences. In his travelogue, Malamud provides an account of

the various sites around the world that were gradually linking themselves to the global

Internet. Casting himself as "one of the free-spirited aboriginal technologists on the new

frontier" (Fischer 1995, 271), Malamud recounted his travels around the world,

crisscrossing the United States from Silicon Valley to Washington, D.C. to Chicago, Europe

from Prague to Geneva to Amsterdam, the Pacific from Honolulu to Tokyo to Hong Kong to

Kuala Lumpur, Singapore, Canberra, Seoul and various other cities. In each of these places,

Malamud discovered the heterogeneity of the actual hardware, wiring, design, and

organizations of various components of a global computer network infrastructure: from

CERN's global Internet hub; a Czech university's reverse-engineered network, made from

old IBM mainframes; and Torben Nielsen's local area network (LAN) made from salvaged

military aircraft material in Hawaii.

Malamud's Technical Travelogue was emblematic, making manifest the global

connectivity that numerous network developers envisioned through the construction of a

fully-operational show network at the Interop trade show - and, likewise, the "correctness"

of their project, and their own role in the physical assembly of these far-reaching

architectures. Malamud celebrates their technical skills and showmanship. He writes that

in the hours before the Interop9l event, a team of network engineers installed more than

thirty-five miles of cable-enough "to wire a 20-story high-tech skyscraper"-as well as fifty

subnets, a microwave link, two different backbones, and a connection to the NSFNET3 7 so

that 300 vendors could demonstrate the interoperability of their products 38 in Interop-

themed groups known as "Solutions Showcases" (Malamud 1993, 29-33). In the book's

foreword, Lynch celebrates these achievements, proclaiming that "this book demonstrates

36 Malamud's Technical Travelogue was also edited by Lynch, Ole Jacobsen, and Dave Brandin (vicepresident of Programs at Interop). Malamud appeared as a speaker at Interop93 to discuss his book.37 In 1990, the U.S. Department of Defense's Advanced Research Projects Agency (ARPA) transferredto the National Science Foundation (NSF) control of the Internet backbone, which was subsequentlyknown as the NSFNET. The NSF was actively involved in the expansion and privatization of thenetwork in this period. In April 1995, NSF gave up control of the Internet. Fred Turner hasreferenced this transfer as a moment that facilitated "the interlinking of commercial, alternative,and government-sponsored networks and the mixing of for-profit and not-for-profit uses across thesystem" (2008, 213).38 These products included Frame Relay, SMDS, X.400, and SNMP.

2 9

what many of us have long felt: the worldwide network is here. Interoperability is not some

imaginary goal at vendor briefings but a concrete part of networks all over the world" (in

Malamud 1993a, vii). Like the Interop show network, which was known as the

INTEROPnet or the "ShowNet," the Technical Travelogue mobilized developers (both the

network engineers and the users) themselves. In this regard, Malamud's Technical

Travelogue functioned as a spectacle, less an account of the various states of the distinct

networks that would comprise a global Internet than a celebration of technological forces

that network developers had unleashed.

Innovation depends on actors who invent technologies as well as construct the

problems that these technologies address (Carlson 1992). In the early 1990s, network

developers worked to solve the complex routing problems of linking networks while they

simultaneously interwove technical protocols and strategies into the massive social and

economic restructuring already underway. Such pronounced shifts required individuals and

organizations to conform to their new protocols and strategies, regardless of their proximity

to these changes. In this way, Malamud's multiple trips around the globe not only captured

a snapshot of the Internet under construction (as it existed in 1991) but also revealed the

resistances to, and efforts to limit or control, connectivity. By doing so, he also created the

rhetorical space for network engineers to act as global problem-solvers.

In Taipei, for example, Malamud writes that he found Taiwan's networks, such as

SEEDNET, unable to adequately connect to other regional networks or to U.S.-based

NSFNET without dropping packets, or information, or cutting off communication

altogether.

The SEEDNET problem was certainly just a temporary one, but it showedthe strains that were beginning to appear on the routing infrastructure of theInternet. ... Cutting off people who probably wouldn't talk to you is certainlya rational response to the problem of saturating the Internet. The problem,however, was that this didn't solve the long-term problem of scaling theInternet. The Internet was doubling every 7 to 10 months and there ... wasobviously a need for many types of networks: the day of "the" network hadlong passed. Yet, this diversity meant that the network was starting tofragment and splinter into subsets of connectivity. (Malamud 1993a, 335-336)

As Malamud suggests, these experiences led him to conclude that an "integrated global

Internet" would be difficult without greater attention to interoperability. One solution he

offered was the engineers themselves, even suggesting that they functioned alongside

protocols and hardware as a critical layer in the technical infrastructure: "Technology alone

40

doesn't make a network, though. The next layer is the people layer where technology is

applied, deployed, and networks start being used" (Malamud 2006, 364). In this way,

Malamud explicitly intertwined technical and social solutions, infusing the physical

construction of computer networks with a moral-technical framing (Kelty 2008) that

equated openness with liberal democratic ideals. Put another way, for Malamud, there was

a "correct" way to build networks.

Is your routing protocol complex? You've raised the cost of entry. Do you havean acceptable use policy? You've limited your population. Have you inventedan anonymous FTP mechanism and an RFC series? You've encouraged thespread of the network. ... Infrastructure ... reflects how we apply ...fundamental human values. Privacy, for example, can be protected ordestroyed by a network. (1993a, 364-365)

Malamud conceptualized the creation of an "ideal market infrastructure that would allow

open systems to flourish" (Kelty 2008, 14) and support "fundamental human values." For

Malamud, the articulation of these values would include practices such as organizing

people and machines across locales and time zones. It would also include sharing

documentation of core operating standards, which in this case was a global communications

infrastructure, the technical standards of the International Telecommunications Union

(ITU).

One underlying narrative of Malamud's account was the ongoing conflict between

the OSI and the TCP/IP standards, a struggle that has become so heated over the years

that it has come to be known as a religious battle, although the conflict could be better

described as the "struggle between the Corporation and the State" (Kelty 2008, 67). Kelty

has described this conflict as one that focuses on the relationship between information

technology (IT) as a reorientation around the ownership of ideas and IT as an economic

driver. The battle between Apple and Microsoft is the most famous, although the tensions

that arose around the decision to include Animatronic versions of Wozniak or Jobs in the

Spaceship Earth exhibit illustrate the degree to which this struggle, the deep ambivalence

it provokes, has become a central component of the network cultures that have around

arisen around the Internet. Malamud's articulation of this dispute was his 1991 effort to

"liberate" the technical standards from the ITU. The "Blue Book," 39 as it was known,

39 According to an Interop press release in advance of Interop 91 Fall, these internationalspecifications regulated "high speed modems (V series), X.25 packet-switched networks, ISDN and

41

comprised international specifications that were normally only available in paper form for

purchase. Malamud had come to believe that the inaccessibility of international standards

was endangering the future of the Internet by "hindering technical progress" (1993a, 3) and

the development of new products. With the help of key figures in the Interop network,

Malamud persuaded the ITU to publish their complete standards (totaling more than

19,000 pages) on the Internet at no charge. 40 In his account, Malamud suggested that "once

the data was digital, we could all start using advanced services, write better code and ...

enter a state of standards equilibrium, a nirvana of documentation" (1993a, 9). This

"experiment," as it was called in the trade show press release, was announced over "live

video link" at the Interop9l Fall trade show. Malamud would coordinate the conversion of

the standards into accessible data files and the publication of the data onto the Internet. 4 1

That was the plan.

Malamud's argument was that the rapid commercial growth of computer networks

necessitated a radical change in the ITU's policies to adapt to the competitive economic

pressures of open markets and open standards that had shaped the information technology

industries since the 1980s. In this case, the drive toward "openness" became an attempt to

make telecommunications standards more widely accessible by posting them online. In the

end, however, the "experiment" had mixed results: the ITU gave Malamud half of its

standards (the other half had been lost in the organization's outmoded filing system), which

he converted and posted on an FTP server, before the ITU abruptly canceled the project

months later.42

Malamud had initially "hacked" the ITU under the rubric of "The Documentation

Liberation Front" (Malamud, 1991). By the time Malamud published his "technical

Broadband ISDN, X.400 message handling systems, fax, telex, teletex, and the X.500 globaldirectory." http://www.scribd.com/doc/2571592/INTEROP-91-Fall-to-Feature-Maior-Announcement.40 He did do by enlisting the support of Tony Rutkowski, Counsellor to the Secretary-General at theITU, as well as Vint Cerf, Chairman of the IAB. Richard desJardins, one of the leading authoritieson the Government OSI Profile (GOSIP) and an architect of the Open Systems Interconnection (OSI)standards, was also involved.41 This was handled through an anonymous FTP (or File Transfer Protocol) file-sharing site.42 The letter canceling the project can be found here: http://www.scribd.com/doc/2571598/Dear-Mr-Malamud. As an aside, the ITU didn't revisit the question of posting their standards online and freeof charge until 2007. In the press release announcing the decision, Malcolm Johnson, Director ofITU's Telecommunication Standardization Bureau (TSB), suggested that posting the standardsonline would help "bridge the 'standardization gap' between countries with resources to pursuestandardization issues and those without." Retrieved from:http://www.itu.int/newsroom/Dress releases/2007/21.html.

42

travelogue" two years later, however, he had further developed the narrative around his

provocation with the moniker "Project Bruno," thus adopting a "usable past" involving

philosopher Giordano Bruno, who was killed for revealing secret knowledge to the rest of

the world. 4 3 In this sense, his project demonstrates the gulf that existed between the two

models of standardization-Open Systems Interconnection (or OSI) and the Internet

community's TCP/IP-and the degree to which these differences mobilized network

engineers like Malamud to work to build a commercially viable network infrastructure.

Each model represented different avenues of legitimacy. TCP/IP had been developed to

allow for the linking of diverse networks, an imperative that was reflected by its emphasis

on implementation, and its availability to anyone via the network. By contrast, OSI"4

seemed likely to define global network architecture (Abbate 1999, 172-179; Russell 2006,

48-49). Endorsed by governments around the world (as well as the U.S. Department of

Commerce), these standards were based on a model of comprehensiveness and consensus

that had grown out of more than a century of coordination and standardization of

international telecommunications. OSI allowed businesses to create proprietary standards

for products; the standards body would function as the validating body that would

determine that various standards could interoperate with one another (Kelty 2009, 167-

168). Developed by the same organizations that had coordinated and standardized

international telecommunications for more than a century, its proponents assumed that

once OSI standards were fully implemented, competing internetworking protocols,

including TCP/IP, would be phased out completely.

Malamud's project also reveals another tension around standardization: that it

implies consensus. Within the Interop network, the need for consensus likely related to the

practicalities of establishing partnerships to ensure the commercial success of the Internet.

Yet even within the confines of the Interop network, standardization was a complicated

experiment that tacked between cooperation and competition, with various social groups

jockeying for the ability to translate their practices across domains, often employing tactics

4 Giordano Bruno was a 16th century philosopher and mathematician burned at the stake for heresy.Although he is remembered as martyred for his beliefs, it is unclear why he was declared a heretic.4 The battle between TCP/IP and OSI has been analyzed in depth elsewhere. For an internalhistory, see Hafner and Lyon's Where Wizards Stay Up Late (1996). For a technical history, seeAbbate's Inventing the Internet (1999) as well as Kahin and Janet Abbate's (eds.). Standards Policyfor Information Infrastructure (1995). For a discussion of TCP/IP debates in the context of thehistory of open source, see Kelty, Two Bits (2008).

42

to limit the capacities of other groups. This occurred most dramatically between proponents

of competing models of standardization and, as we have seen, with the Internet leadership

as they sought to retain control over a commercializing network, but it also occurred within

factions.

In his 1993 recounting of the project, Malamud concluded that the strategies of the

"Bruno" project were not sustainable on a larger scale. "Bruno was a stopgap, and even if a

few people working on their own could come up with a new stopgap, what we need is a real

solution" (1993a, 366). Malamud considered getting countries, perhaps Korea, to set up

standards havens with the professed hope of forcing organizations such as the ITU to

widely distribute the material so that it would be available to citizens and developing

countries alike. Janet Abbate notes that "efforts to create formal standards bring system

builders' private technical decisions into the public realm; in this way, standards battles

can bring to light unspoken assumptions and conflicts of interests" (1999, 179). For

Malamud, by beginning to identify a particular course of action around the documentation

of technical standards, he expressed "openness" in a manner that might have appealed to

the sensibility of many Interop affiliates. However, given the pressing demands of

privatizing and commercializing the network, Malamud's projects were often seen, at least

by people like Dan Lynch, as troublesome provocations that drew attention away from the

most critical tasks at hand (interview 2009: see Appendix A). Malamud was equally

interested in commercializing the Internet, although, as we shall see in later chapters, his

impulse to promote a "people layer" in infrastructures would later lead him to very different

conclusions about what commerce might look like online. He would come to believe that a

"coherent business environment" required parks and schools and museums that would

attract "visitors" (1996).45 He would see himself as the right man for the job.

The Romance of Network Operators

On a sunny morning in June 2009, I meet Ole Jacobsen, a tall, affable man who is the

longstanding editor and publisher of the Internet Protocol Journal at Cisco Systems. 46

45 Byczkowski, John. 1996. World of Fun. Norwich Bulletin. January 28. http://www.scribd.com/doc/2576777/World-of-Fun.46 Today Jacobsen is the Editor and Publisher of The Internet Protocol Journal, a quarterlypublication at Cisco Systems, and continues to be active in Internet governance issues, primarilythrough the Internet Corporation for Assigned Named and Numbers (ICANN).

44

Moments after I met him, Jacobsen pulled a yellowed Radio Corporation of America (RCA)4 7

advertisement from his briefcase to show me. Jacobsen was the editor and publisher of the

Interop trade show's monthly newsletter, ConneXions - The Interoperability Report, from

1987 until 1998 (interview 2009). He thrust the advertisement into my hands, promotional

literature that appeared to be from the 1920s. On it is drawn a family seated around a large

radio. He pointed out the complicated wires coming out of the box, encouraging me to

consider how complicated it would be for the general public, and that somebody needed to

make radio simple enough to use the technology without thinking about how it works. The

radio became simpler, Jacobsen suggested to me, because of operators who experimented

and made changes.

Jacobsen's story evokes the nostalgia of the wireless signals and crystal receivers of

early radio. From its public unveiling at the start of the 20th century, wireless gripped the

American imagination as a technology that provided the means to achieve a happier, more

abundant society. In newspapers, vaudeville, and popular science magazines, the

invisibility of wireless networks fostered visions that the new technology would allow

individuals to communicate through "the air" with whomever they wanted, whenever they

wanted, and without the help (or obstruction) of governments or corporations. Despite this

idealism, corporations and the U.S. military tended to view radio as a long-distance, point-

to-point communications technology, not as an invention for the masses. Amateur

operators, an estimated several hundred thousand strong, "dominated the air"48 and

transformed wireless into an intensely participatory communications medium, first

articulating a public right to access the communications network.

As Susan Douglas demonstrates, these enterprising "radio boys"-primarily white

middle-class males-constantly tinkered with their radio sets. "Amateurs didn't just adopt

this new technology," she writes, "they built it, experimented with it, modified it, and

sought to extend its range and performance. They made radio their own medium of

expression" (1986, 44).49 Relishing the relative exclusivity afforded by the mastery of

wireless and the romance of marrying the science of wireless with the mysteries and

adventures of the new medium, amateur operators understood themselves as laying the

47 The Radio Corporation of America was formed in 1919 to establish an American-controlledinternational network.48 De Soto, Clinton B. 1993. Two Hundred Meters and Down: The Story of Amateur Radio. AmericanRadio Relay League, (West Hartford, Connecticut), 3, quoted in Douglas 1986, 44.49 For a history of radio, see Susan Douglas' Listening In: Radio and the American Imagination.

45

groundwork for the future ideal society. For example, one amateur operator turned book

author reminded his readers, "radio is still a young science, and some of the most

remarkable advances in it have been contributed by amateurs-that is, by boy

experimenters... Don't be discouraged because Edison came before you. There is still plenty

of opportunity for you to become a new Edison, and no science offers the possibilities in this

respect as does radio communication" (Binn 1924).50 Despite their visions for the

technology, amateur operators were soon relegated to the shortwave spectrum as the

majority of the airwaves were either militarized or sold off to commercial interests.

Although radio would increasingly accommodate corporate interests, amateur visions of

wireless continued to shape the technology. Their practices helped demonstrate the benefits

of a national communications network and the public as "rightful heirs to the spectrum" 5 1

that later proved critical to the legitimization of the public's claim to the air.

Seven decades later, the same aspects of a technological utopia were marshaled in

the construction of the Internet, a new democratizing communications medium that

promised to empower individuals and transform society. Network developers like Jacobsen

envisioned themselves as building this new future through computer technologies. Like the

amateur operators before them, the network developers were animated by their willingness

to scientifically test and implement new ideas, and by seeing themselves as members of an

exclusive technical community working to ensure the public right to the computer networks

increasingly crisscrossing the globe.

Jacobsen's decision to begin an interview about Interop with an early RCA

advertisement suggests that Jacobsen envisioned himself in relation to the amateur

operators who pioneered shortwave broadcasting, ensuring that even as early visions of

wireless merged with the reality of corporate-controlled airwaves, the technology retained

critical utopian elements. Despite the limits of comparing the emergence of radio and the

Internet (chiefly, that it tends to highlight some influences and obscure others), Jacobsen

was not alone in seeing the network development of the early 1990s in relation to the

50 Excerpt from Binn, Jack. 1924. Foreword to The Radio Boys with the Iceberg Patrol by AllenChapman. New York: Grosset and Dunlap, quoted in Douglas 1986.51 The Communications Act of 1934, which established the Federal Communications Commissionand required the licensing of all radio stations, mandated that these stations serve "the publicinterest, convenience, and necessity."

emergence of earlier technologies. For instance, network engineer and author Ed Krol 2

compares the development of the Internet to the early automobile in his 1992 best-selling

book, The Whole Internet Users Guide & Catalog.3 Krol implies that the global computer

network infrastructure was built by technically skilled developers who innovated through

experimentation.

In the early 1900s, if you wanted to tinker with horseless carriages, you fell

in with other tinkerers and learned by doing. There were no books aboutautomobiles, no schools for would-be mechanics ... [E]arly cars were sounreliable that they could hardly be called transportation. ... Eight years ago,the Internet was in much the same state: ... slow and unreliable. Its majorpurpose was not to do anything useful, but to help people learn how to buildand use networks. (1992, 5)

Like many commentators in the early 1990s, Krol suggests that the Internet became

"useful" once the technology became accessible for a non-technical audience. According to

Krol, these users "demand reliability, and don't want to be mechanics. ... They are

computer-literate, but not network-literate." In other words, Krol expects the general user

to be uninterested, and even put off by, the technical details of the Internet's operation. His

emphasis on network literacy, which he goes on to describe largely in terms of the "rules of

the road" on the Internet, offers the basic skills that newcomers will need to know to

navigate online and join the prophesied future that developers already inhabit. This

included governance of the Internet, which Krol described as organized by a "council of

elders," a 12-member organization 54 that included Interop co-founder Dan Lynch as well

52 Critically, Krol worked at the National Supercomputing Center at the University of Illinois, thelab where the first version of the graphical web browser Mosaic was released in 1993. It was not thefirst web browser, yet Mosaic was instrumental in making the Internet more publicly accessible.53 Krol's book was first published by O'Reilly Media, a company that has played a substantial role inthe popularization of networked computing for a larger general audience. This first printing is afascinating artifact because it pre-dates the public introduction of the World Wide Web, whichoccurred in 1993. Tim O"Reilly has suggested that he convinced Krol to include a short chapter onthe Web at the last minute. O'Reilly also posted a number of chapters from Krol's book on the GlobalNetwork Navigator, the first commercial website. Krol's book was also one of the first publicationsposted online. It can be found here: http://www.archive.org/stream/wholeinternetOOkrolmiss#page/nl/mode/2up. In both cases, Krol's book illustrates the distinctionbetween the Internet, a network of physically connected computers that serves as the communicationinfrastructure for the World Wide Web, and the sites and hyperlinks that comprise the Web itself.54 I refer to the Internet Architecture Board (IAB), which provided technical oversight and guided thearchitecture of the network protocols, overseeing the standards-setting arm, the InternetEngineering Task Force (IETF) and the Internet Engineering Steering Group (IESG) as well asmanaging the Request for Comment (RFC) document series. Working in conjunction with theInternet Corporation for Assigned Names and Numbers (ICANN), the IAB is responsible for theadministration of the Internet Assigned Numbers Authority (IANA). The IAB was originally an

47

Vint Cerf and David Clark. This group met "regularly to 'bless' standards ... decid(ing)

when a standard is necessary, and what that standard should be" (1992).55

Krol's characterization of the Internet reflects a romantic narrative around the

emergence of the Internet that has long since hardened into legend - namely, that the

Internet was crafted by entrepreneurial "boy wonders" whose innovative principles of

cooperation and open exchange encouraged its global spread. At the same time, these

utopian stories disseminated to the public downplay the degree to which network

developers actively engaged with existing technical, social, and political infrastructures.

These parallel stories exist for a reason.

For the network engineers affiliated with the Interop network, their focus on

standards and infrastructures are provocative because they shed light on the degree to

which technological systems are consciously constructed. It is a testament to the Internet's

success that its infrastructures - both the technologies of networked computing (the

physical hardware, and also practices and embedded knowledge) and the graphical

interfaces - have come to be understood as a singular phenomenon, naturalized into

everyday life, simultaneously (at least theoretically) global and local. Except when Internet

connectivity "fails" (or becomes outdated, as has occurred with the United States' lagging

broadband coverage), the Internet's fundamental infrastructures - the "bottom" layers of

core networking protocols and physical hardware like routers - are largely invisible. Yet in

the 1980s and early 1990s, long before networked computing as a communications medium

had begun to operate "as advertised," and, in fact, even before it was clear that the Internet

would become the dominant global information infrastructure, these bottom layers, the

Internet itself, were under construction. It is here that we turn to the Interop trade show

itself.

acronym for the Internet Advisory Board, although the committee was renamed the InternetActivities Board in 1986, and the Internet Architecture Board in 1992.55 Krol also released a modified version of the first chapter of his book, The Whole Internet UsersGuide & Catalog, as an RFC. The RFC can be accessed here: http://www.rfc-editor.org/rfc/rfc1462.txt.

48

Chapter 3: I Know It Works, I Saw It At Interop

We join the Interop5 6 in the early 1990s, at the height of the trade show's influence. The

semi-annual event had become one of the most respected and popular trade events in the

industry.5 7 The events were based in the U.S, usually in San Jose, San Francisco, and Las

Vegas, although by 1992, the trade show had expanded to international venues such as

Sydney, Paris, and Tokyo. "Interop was like a rocket ship," former Interop Vice President of

Programs David Brandin recalled in a recent interview. "For years it was the only place

where you could see the stuff work" (Appendix A). "Seeing the stuff work" at Interop

entailed the spectacle of a real-time "demonstration" of the emerging communications

infrastructure and a "process" for assembling those networks. Interop offered lectures and

in-depth tutorials by leading researchers 58 as well as "Bird-of-a-Feather" informal meetings.

The event showcased vendors' latest computer networking hardware, including routers,

access points, storage arrays, and security appliances through a functioning show network,

or "INTEROPnet" that demonstrated these technologies in practice. "Most trade shows are

satisfied to leave individual networks to vendor booths, or to put a simple Ethernet cable

into place on the show floor. Interop manages to put in one of the more complex networks in

the world in the space of just two days," Malamud wrote about the Interop9l San Jose

event in Communications Week (1991). "A real network means vendors can prove new

technologies work. There is nothing like an interoperability demonstration featuring dozens

of competing vendors to convince users that a new technology is real. The vendors get to

help build new markets. The engineers get to test their products in a real environment. And

dozens of talented computer engineers get to stay up all night and pull off a technical tour

de force."

56 The present-day gathering is one of the oldest and largest information technology trade shows,billed as "the event where the global IT community comes together to see all the latest technologiesin action" (2009), which includes the latest in security, networking, storage, and software products.Interop has replaced COMDEX, or Computer Dealer's Exhibition, which dominated the technologyindustry for decades with shows that offering computer hardware, software and associatedcomponents to all levels of manufacturers and developers. Despite its success, the present-day tradeshow shares little in common with Interop as it existed from its founding in the mid-1980s until1995. In this earlier iteration, Interop was a smaller, specialty conference and series of publications,more narrowly focused on enterprise computer networking and on integrating the efforts of theengineers and vendors working to connect various networks together.57 According to the ConneXions publications, attendance at the event averaged 30,000 attendees.58 MIT's Dave Clark and Purdue University's Doug Comer taught the most popular of these tutorials.

49

Pulling this off required Interop founder Dan Lynch to bring together individuals

from different firms and research groups in a series of encounters that occurred around the

Silicon Valley-based trade show that (re)-infused the hybrid production strategies of the

region with the military-industrial practices of collaboration and implementation from

which computer networking technologies had first emerged. He assembled the somewhat

overlapping worlds of military-industrial research, enterprise networking firms, and user

communities. Forming flexible partnerships, representatives of multiple groups came

together, driven by a shared vision of a grand scheme of inter-networks and the recognition

of the practical need to ensure the global success of the Internet.

These groups assembled as part of a broader struggle to standardize and expand the

Internet technology. In the 1980s and early 1990s, the network system's place in global or

national information infrastructures was uncertain. For the embattled Internet leadership,

Interop gave them a venue to retain their authority over a rapidly changing information

infrastructure. For the network engineers, Interop allowed them to build the prestige and

entrepreneurial networks vital in the emerging freelance patterns of employment. Through

Interop, these engineers and researchers also worked to ensure that the Internet did not

simply become a global infrastructure in name only but retained the collaborative practices

within which the technology developed. For networking firms, Interop presented an

opportunity to respond to the massive global reorganization of the information technology

industries that had ushered in "open systems," and with it the demand for open markets

and open-standard processes for high-tech networked hardware and software. Interop

afforded them a space to not only learn how to link networks but, more critically, to field-

test "interoperable" products before releasing them on the open market. By tacking between

the commercial demands of global markets and the moral-technical visions (Kelty 2008) of

engineers, Interop translated the cybernetic dream of boundless connectivity (or seamless

integration) through a global inter-linking of computer networks. However, the popularity

of the trade show only tells part of the story. In the 1980s networked computing was

expanding exponentially, although the Internet faced a raft of competitive corporate and

government forces that left its long-term success uncertain (Abbate 1999, 147-179). It was

in this moment of ambiguity that Interop emerged.

50

Struggles Over "Open Systems"

On a sunny June afternoon in 2009, I met Dan Lynch at his home in Napa Valley, Northern

California's wine country. I had come there to learn more about the Interop trade show that

the former ARPAnet researcher and computer manager at Stanford Research Institute

(SRI) had founded more than two decades earlier. In many ways, his story parallels the

critical role that the military-industrial research world had in the standardization and

commercial success of the Internet. In our conversation, Lynch recalled that contests

around standardization, and the campaigns toward commercialization that subsequently

became possible, were at the heart of the Interop trade show.

In 1986, Lynch received a request from the Defense Communication Agency (DCA),

which was charged with centralizing communications throughout the military as an

attempt to bring operations under central Department of Defense control (Abbate 1999,

20),59 to help further develop applications and products for the core networking TCP/IP

protocols. In response, Lynch organized an invitation-only meeting to brainstorm about the

future of the Internet protocols. A few hundred current and former ARPAnet researchers -

computer scientists and engineers from industry, government, and academia that had been

instrumental in the development of the Internet - were in attendance. 60 Those first

interoperability meetings functioned as collaborative workshops between vendors and

researchers that focused on existing and emerging issues with the Internet protocol, both

within the ARPA research community and among the vendors in the field. 61 They were

intended to promote solidarity between vendors and researchers at a critical stage in the

59 The DCA was founded in 1960 as the combat support agency of the U.S. Department of Defense(DoD) focused on providing real-time information technology. DCA is now known as the DefenseInformation Systems Agency (DISA), which is responsible for planning, developing, fielding,operating, and supporting command, control, communications, and information systems for the DoD.60 The attendee list at these first meetings read as a who's who of the military-industrial researchculture that had developed and implemented key elements of Internet functionality. It included MITprofessor David Clark, Purdue University professor Douglas Comer, Dave Crocker, Vint Cerf (co-inventor of TCP/IP protocols), and Jon Postel (editor of the RFC documents and administrator of theInternet's names and numbers process). [Interop Company. 2nd TCP/IP Interoperability Conference,December 1-4 1987 Attendee List, SRI Network Information Center Records, Lot X3578.2006,Interoperability Materials, Computer History Museum, Mountain View, California.]61 In 1991, the five key areas identified were: Routing and Addressing, Multi-Protocol Architecture,Security Architecture, Traffic Control and State - to accommodate real-time applications - andAdvanced Applications, among them the "increased need for innovation and standardization inbuilding new kinds of applications." Source: David Clark et al. 1991. "Towards the Future InternetArchitecture," RFC 1287: 3. Presumably these goals had remained the same, or nearly the same,since 1986.

51

Internet's development and, within a larger climate of anxiety around U.S. global

technological and economic leadership.6 2 Within a year, in 1987, Lynch restructured the

meetings into a commercial conference and trade show and shortened the name to

"Interop." The relationships that these early meetings fostered would last for more than a

decade, ensuring the construction of an increasing number of projects built according to

TCP/IP protocols and later the Internet's long-term success in the battles over a global

network standard. These relationships would also work to a more immediate effect: Lynch

wanted to build a permanent display of TCP/IP's capabilities, an exhibit he called the

"Connectivity Showcase." In an email thread posted August 28, 1987, Dan Lynch relayed a

plan that would involve dozens of vendors "demonstrating TCP/IP interoperability to the

public. It will be open daily ... and will be paid for by the vendors who want to clearly

demonstrate that their products run harmoniously together. Users will be able to come in

and run demos between any machines they wish to find out about."6 3 Lynch later altered

his plan to have a year-around exhibition. The following year, Interop launched the first

trade show network, a "fairly ambitious demonstration of TCP/IP interoperability" that

included an intranet (a private computer network within an organization) that linked 49

different vendors to one another and to as many different pieces of hardware as possible 64

in order to illustrate that although the Internet TCP/IP protocol suite was developed to link

distinct networks, it was "not tied to any particular physical medium" (Almquist 1989, 2)

62 Although it is outside the scope of this thesis, efforts to build global infrastructures amenable toU.S national and corporate interests should be read against the larger geo-political struggles overlong-term technological and economic competitiveness. President Ronald Reagan's mid-1980s foreignpolicy was defined by a return of Cold War era closed world politics. In his 1985 State of the Unionaddress, Reagan defined the U.S. mission one "to nourish and defend freedom and democracy, and tocommunicate these ideals everywhere we can." As former Interop international vice-president DavidBrandin suggested to me in a phone interview in August, 2009, the emergence of Interop wasdirectly related to the considerable anxiety over America's ability to retain global technological andeconomic control. Japan presented a particular threat because it had launched a joint government-industry-university research effort focused on artificial intelligence, parallel processing, andmicroprocessor technologies (Brandin 1987; Edwards 1996, 298-299). For a period account, seeBrandin, David R., and Michael A. Harrison. 1987. The Technology War: A Case of Competitiveness.New York: John Wiley & Sons.63 Lynch, Dan. 1987. Message to Re: Sun routers. http://www-mice.cs.ucl.ac.uk/multimedia/misc/tcp ip/8706.mm.www/0067.html.64 According to the February, 1989 ConneXions - The Interoperability Report, the media includedseveral versions of Ethernet, IBM/802.5 Token Ring, and amateur packet radio. The original articlecan be accessed here: http://www.cbi.umn.edu/hostedpublications/Connexions/ConneXions03 1989/ConneXions3-02 Feb1989.pdf.

52

and thus able to run on intranets as well. The show network also provided two connections

out to the Internet.

In many ways, Lynch was an ideal candidate to spearhead the Internet's transition

into a commercially viable communications medium. He already had extensive experience

with the assembly of the Internet itself. In the early 1980s, he65 had managed the painful

"cutover" that many consider the birth of the Internet. That massive, multi-year effort

transitioned two hundred or so U.S. government contractors and research teams from

Network Control Program (NCP) and their own proprietary networks to the more flexible and

powerful TCP/IP protocol suite that made interconnected networks possible.66 The

magnitude of this change was as much cultural as it was technical. From the 1960s until

1980, networks were relatively closed systems managed by a single entity - whether by a

government agency, company, or utility. Components were made by a handful of companies

who had built them to their specifications. Many had built their own networks with

proprietary hardware, software, and architectures (Abbate 1999, 148-151; Kelty 2008, 145;).

The transition had proven so unpopular with vendors that Vint Cerf, then DARPA research

manager, had twice shut down the ARPAnet in order to convince the companies that they

would be forced to comply with the changeover (interview, 2009).67 In fact, failure to adopt

these new protocols would have meant getting cut off from the network itself. In a 1991

interview Lynch recalled, "Dozens of us systems managers found ourselves on a New Year's

Eve trying to pull off this massive cutover. We had been working on it for over a year. There

were hundreds of programs at hundreds of sites that had to be developed and debugged."

Once the changeover was complete, Lynch commemorated the occasion by making buttons

that read, "I Survived the TCP Transition."68

65 As Director of the Information Processing Division at the Information Sciences Institute (ISI-USC)66 Cerf, Vint. 1987. Message to [ih] NCP to TCP/IP Transition. http://www.postel.org/pipermail/internet-historv/2009-April/000796.html.One of the best accounts of this time period is Abbate's Inventing the Internet (1999). An insideraccount of the time period can be found in Hafner and Lyon's Where Wizards Stay Up Late: TheOrigins of the Internet (2000).67 Kahn, R.E. 1994. The Role of the Government in the Evolution of the Internet. Communications ofthe ACM, 37(8): 16. Kahn, who co-invented TCP/IP with Cerf, was director of ARPA's InformationProcessing Techniques Office from August 1979 until September 1985. In addition, Cerf left ARPA inOctober 1982 and Barry Leiner did not replace him until August 1983, Kahn personally managed thetransition to TCP/IP. [R.E. Kahn, oral history interview, OH 192, CBI.]68 Lynch, Dan. 1991. Message to comp.protocols.tcp-ip, 23 June. McKenzie box 2, Bolt Beranek, andNewman library, quoted in Abbate 1999, 140-141.

5.2

Within three years, by the mid-1980s, Lynch and the other ARPAnet veterans once

again found themselves working to convince an unwilling user base to conform to the

TCP/IP Internet protocols. This time, they needed to convince a rapidly expanding vendor

and user community to adopt these protocols without the authority to compel them to

employ them, and in the midst of an increasingly acrimonious contest over "open systems."

At the time, there was a proliferation of proprietary hardware, software, protocols, and

systems (Kelty 2008, 147). Chris Kelty suggests that although the concept of "openness"

held many different agendas - variously articulated as open source code, self-publishing,

specifications available to certain third parties, or standards set by governments and

professional societies - all carried an antagonism toward "proprietary" systems (2008, 147).

For the developers and consumers (users) alike, struggles over open systems and

"interoperability" took on the guise of a "cultural imperative" that integrated the ideals of

the free market and of the free exchange of knowledge (Kelty 2008, 148). As Carl

Malamud's attempt to provoke the ITU into releasing their "Blue Book" of international

standards suggests, the struggle between the TCP/IP and the OSI core inter-networking

standards had become emblematic of the split between a bureaucratic, monopolistic

telecommunications industry and a flexible, networked computing industry.

The Internet TCP/IP protocol suite's eventual success over the international

standard have been commonly described as a triumph of the Internet model-that is, of a

culture of "openness"-over proprietary industries. This success has been framed primarily

as a technical achievement, and the groups tasked with overseeing this standard-setting

and architectural design processes - the Internet Activities Board (IAB) and the Internet

Engineering Task Force (IETF) - as exemplars of the collaborative and technical practices

that emerged as part of the Internet. The IETF and the IAB, were not political structures in

the traditional sense - they operated without legal mandate or any enforcement mechanism

to promote their standards - yet they functioned as the primary mechanism for governing

the Internet. It occurred primarily through a set of technical documents, known as

"Request(s) for Comments," or RFCs, that oversaw the development and implementation of

the specific technical protocol standards that comprise the Internet. RFCs had initially

emerged while the network was a military research project as a consensus-style process

through which a technical document became an Internet standard only after it had been

placed in the "standards track," significantly developed, and then reviewed by a number of

F54

parties. The IAB (as well as the Internet Engineering Steering Group, which oversaw the

IETF) had control over which documents entered the process (Galloway 2004, 134). The

RFC process had emerged to promote informal communication in the "absence of technical

certainty or recognized authority" (Abbate 1999, 74); the process also helped ensure that

the scattered collaborators who were involved in these conversations were able to easily

communicate with one another. On the 30th anniversary of the RFC, in 1999, Vint Cerf

described the RFC process as a conversation:

When RFCs were first produced, they had an almost 19th century characterto them - letters exchanged in public debating the merits of various designchoices for protocols in the ARPANET. As email and bulletin boards emergedfrom the fertile fabric of the network, the far-flung participants in thishistoric dialog began to make increasing use of the online medium to carryout the discussion. 69

According to conventional wisdom, these innovative technical practices gave the fledging

network a leg up over numerous competitors, most notably OSI standards.

Although the RFC process has become part of the mythology of the Internet, it does

not address how standards are able to move across other domains, or how a version of the

RFC process survived in the commercial environment. As the Internet's user base rapidly

expanded, the leadership of technical organizations like the IAB and the IETF had become

increasingly unable to manage the network architecture and standards (Abbate 1999, 207-

208).70 Perhaps more revealing, RFCs effectively documented the processes of

interoperability, yet, according to Lynch, they were too obscure to be implemented as

published. "If you tried to build a network just using RFCs, you'd run into a lot of problems

that had already been worked out in the field. This was a major issue with corporate

engineers who weren't part of the RFC process" (interview 2009). Put another way, the RFC

process would have been of limited use in the expansion of the Internet. Herein lies the

Interop network's great contribution to the standardization of the network: it provided a

mechanism for the practicalities of physically implementing RFC standards in a chaotic

69 Cerf, Vint. 1999. RFC 2555 - 30 Years of RFCs. http://www.faqs.org/rfcs/rfc2555.html.

70 Abbate argues that these organizations lacked accountability and international representation aswell as faced increasing legal challenges and a host of other challenges that compromised theirability to manage the rapidly growing network. This crisis was such that, in 1991, senior Internetleadership, specifically the Internet Activities Board (which included Dan Lynch, Vint Cerf, andDavid Clark) called a series of meetings about the future of the Internet. For a compelling account ofthis crisis in Internet governance, and the related conflicts around Internet addresses, see LauraDeNardis' Protocol Politics: The Globalization of Internet Governance MIT Press. 2009.

55

market climate that largely was not interested in the Internet per se, but rather only in

creating networks. It provided, as Janet Abbate suggests, a site that "internalized the

competitive forces of the market together" (1999, 145). Once there, out of a mixture of

practical need and technical desire, these parties would forge partnerships that would

rapidly address some of the thorniest issues around interoperating networks.

A close look at the corporate practices in Silicon Valley in the 1980s and 1990s

reveals that Interop and its role in the expansion and commercialization of the Internet

might best be understood within the industrial community that helped drive the regional

critical capacity in the global economy. In contrast to the Internet's technical organizations,

which attempted to define their authority over the Internet's protocols, architecture, and

practices, the Interop trade show network functioned as a hybrid implementation and

production environment. Individuals from different companies and organizations, often

fierce competitors, came together to discuss common problems and consider solutions, and

the process encouraged them to form flexible, innovative partnerships driven by a shared

recognition to keep the Internet advancing globally. At Interop, standards were negotiated

informally, and competitive standards like OSI were actively incorporated into the

conference talks as well as the INTEROPnet. Put another way, the trade show translated

the IETF's standardization formulation of "rough consensus and running code" into a

commercial environment, and did so at a time when the growing user base and complexity

of the network connections meant that testability had become increasingly difficult. The

trade show also promoted the TCP/IP standards by constantly demonstrating the protocol

suite's capacities, and by driving the implementation of products based on TCP/IP.

Although it is impossible to determine how much impact Interop ultimately had, there is

evidence to suggest that these tactics must have contributed to the Internet's later success.

By the late-1980s, as increasing numbers of products based on TCP/IP began to show up,

OSI standards no longer wore the guise of an apparent global standard. Computer scientist

Carl Sunshine (1989) put it this way, "It is ironic that while a consensus has developed that

OSI is indeed inevitable, the.TCP/IP protocol suite has achieved widespread deployment,

and now serves as a de facto interoperability standard" (Sunshine 1989, 5, quoted in Kelty

2008, 175).

After managing the transition to the Internet, Lynch had gone on to open several

businesses of his own. All failed, but the experiences had drawn him into the industrial

56

economy of Silicon Valley, a global competitive environment infused with the collaborative

and deeply entrepreneurial working style as well as the systems thinking of the military-

industrial world. In this way, Silicon Valley also functioned as a location where the military

industrial research world re-inserted itself into the commercial environment. One of the

most influential groups in the region was Douglas Engelbart's Augmentation Research

Center (ARC) at the Stanford Research Institute (SRI) (and later at Xerox's Palo Alto

Research Center). Himself a veteran of SRI and a friend of Engelbart's, Lynch was likely

deeply influenced by Engelbart's philosophy of "bootstrapping" which attempted to leverage

man's collective capacity to address the world's complex, urgent problems. Fred Turner has

suggested that Engelbart "worked to create an environment in which individual engineers

might see themselves both as elements and emblems of a collaborative system designed to

amplify their individual skills" (2006, 108). That idea can be understood as an

organizational strategy to retain the flexibility of small research groups as they grew in

size. These concepts helped shape Lynch's thinking as he considered how to implement

technical advances in an environment that demanded that the Internet leadership convey

procedures and coordinated processes to a user base that was only somewhat willing to go

along. That Lynch and the other network engineers accomplished this "routinization" of the

RFC processes helped determine the success of Interop, and also of the Internet more

generally.

Depicting Technological Change: The INTEROPnet as Prototype

Numerous theories of technological change have posited the form and function of

technologies as determined by the cultural values, interests, and interpretations of social

groups (Bijker, Hughes, and Pinch 1987; Bijker 1995). Among the concepts introduced are

closure and stabilization, processes where a social group involved in designing a technology

decides that a problem has been solved, which in turn defines (and limits) how the

technology is understood and used in society (Pinch and Bijker, 1989). The most visible

element of the trade show was the INTEROPnet. This demonstration simultaneously

illustrated "connectivity" and "openness" in practice and outlined the capacities of the

TCP/IP protocol.

In order to show off this diversity of media, we suspended the cabling fromthe ceiling, where it was in plain view. Unfortunately, the diversity wasn't asapparent as we would have liked, since most of the kinds of media use cables

57

that are thin and black ... A number of people complained to me that thetransceivers, which were hung on loops in the cable about a dozen feet abovethe floor, looked "messy." (Almquist 1989, 2)

From these uncertain beginnings, the show network would mature into a spectacle

illustrating the technical prowess of its engineers and their practices of bootstrapping

between standard and implementation that had emerged with the development of

computing technologies in America. The focus on interoperability and connectivity would

help drive the binding visions that unified engineers. Engineers associated with the Interop

network, like Karl Auerbach, used the show network to illustrate their vision that

interoperable networks would allow the seamless flow of information:

The whole idea here, you've got to make your equipment talk to one another.Because consider a telephone. What value would be the fanciest telephone inthe world if it couldn't talk to another telephone on the other side of thecountry? (1993)71

As a prototype, the INTEROPnet provided a mechanism to illustrate both the technical

viability of the Internet protocols, but also to promote those technical standards' ability to

"interoperate" with a number of competing protocols. Such flexibility underlined the

importance of integrating innovation and implementation, a technique that resembled the

IETF aesthetic of "rough consensus and running code." The network design of the show

network at Inteorp93, for example, was altered more than a dozen times in the weeks

leading up to the event. This sentiment was echoed by Interop Manager of International

Engineering and Design Bo Pitsker:

Networks should be implemented very quickly. Because if the deployment ofthe network is stretched out over a period of years, which frequently happensin the corporate setting, the requirements change. By the time that networksbuilt, it's already obsolete. We design the network and implement it in aperiod of less than six months. And then we tear it down. We re-design it,and re-deploy it in less than six months again. (1993) (see Figure 6)72

These tactics also helped the Internet leadership retain and extend their influence over the

network system as it extended into the commercial domain, at a time when more explicit

efforts to exert oversight over network standards had proven inadequate.

71 Auerbach, Karl. 1993. Interop93. Interop93. San Jose, CA: Interop Company promotional material.Video. http://www.lazy-booklet.org/-atzko/interop93full.mp4.72 Pitsker, Bo. 1993. Interop93. Interop93. San Jose, CA: Interop Company promotional material.Video. http://www.lazy-booklet.org/-atzko/interop93full.mp4.

Figure 6: Screen grab of the construction of the Interop trade show's INTEROPnet (or ShowNet). Therouting and bridging equipment used to construct the show network was estimated to have been asmuch as a major corporation would use to supply offices in fifteen or twenty cities.73

Coordinating Collaboration Through the Interop Trade Show INTEROPnet

The Interop trade show network functioned as a kind of hybrid production network

research lab that, if it had been privately funded, might have cost millions of dollars (see

Figure 7). Interop's show network was built on participation from academic researchers as

well as their counterparts in networks and enterprise information technology. A number of

companies, like Cisco Systems and Sun Microsystems, donated technical expertise and

hardware. The trade show fostered collaborative research in an environment where

competitors (entrepreneurs and companies alike) could work out challenging

interoperability issues more efficiently and with relatively less risk than they would have

faced on the open market.

73 Interop93. San Jose, CA: Interop Company promotional material. Video. httpi://www.lazy-booklet.org/-latzko/interop93full.mp4.

59

Figure 7: Diagram of the Interop9O exhibition INTEROPnet (or ShowNet). Note connections to theInternet via NASA Ames and to the OSI-supported networks via AT&T Accunet. Courtesy of theComputer History Museum.

In his account of the Interop show network, Malamud wrote:

I spoke to one engineer who says he gets more bugs worked out in one week

at INTEROP than he can in six months in the lab. By testing his

implementations with those of other vendors, we can quickly hone in on

ambiguities in the standards and figure out what to do to make the standard

an interoperable reality. (1993a, 33)

In their analysis of network development in a British aerospace project, John Law and

Michel Callon (1992) have argued that what is required to successfully join distinct

networks is a "negotiation space" that affords project builders the autonomy and privacy to

make mistakes, experiment, and arrive at solutions (1992, 21-52). They have also written

that the negotiation space needs to be able to "impose itself as an obligatory point of

passage" (46). Thus, even as "interoperability [became] increasingly difficult to achieve," the

nature of the INTEROPnet and the Interop leadership itself forced engineers and firms to

address difficult internetworking issues caused by competing standards and test new

products to ensure that they neither failed to work with competitors' products nor "broke

the Internet" (Pitsker 1993, 3). Interop's focus on bootstrapping between consensus around

standards and rapid implementation helped drive the construction of products according to

TCP/IP protocol specifications by entrepreneurs as well as by more established computer

companies. This was helped by the practicalities as much as by loyalty to Internet

practices. Thus, although the OSI model was far more comprehensive, it was generally still

not yet built, making rapid implementation difficult. In a matter of years, the Internet

protocol had become one of the most widely employed standards. By the mid-1990s, the ISO

model, which had once seemed untouchable, had been officially retired.

Depicting the Global Network

Who was in this world that network developers created? To the extent that it is possible to

assemble a partial image of the Internet as it was imagined by one group of engineers

affiliated with Interop, it would be largely American, white, likely affiliated with a

university, entrepreneurial, and overwhelmingly male. In this way, it resembles many

other histories of the Internet. As Turner has noted in his account of the Whole Earth

network, "it would turn away from questions of gender, race, and class, and toward a

rhetoric of individual and small-group empowerment" (2006, 97).

The spectacle of the INTEROPnet had unleashed visions of boundless connectivity,

ideas so compelling that, as Ole Jacobsen wrote in an issue of the trade show journal

ConneXions - The Interoperability Report, attendees had considered aloud whether they

might soon be connected to outer space: "On Thursday morning, September 29, the space

shuttle Discovery lifted off, and I heard a few attendees wondering if they'll be able to

contact the shuttle from next years show floor. I can just see it now:

%pingdiscovery.shuttle.nasa.gov" (1988, 7). Yet for all of the attention paid to connectivity

and markets, there was almost no attention paid to regions of the world that were not

already industrialized. For the Interop engineers, theirs was a world in which they were the

rightful heirs to the global networked computing infrastructure that they had assembled

into a unified entity. They were the ones tasked with ensuring that users behaved in the

proper way. This claim recalls a story Karl Auerbach told me. During one early Interop

event, a hacker based in Italy kept breaking the Interop show network. He and Carl

Malamud responded by "turning off' the Internet in Italy (interview 2009).

By the mid-1990s, Interop's grand era of influence came to a close. Lynch had sold

Interop to Ziff-Davis for 160 million dollars in late 1990, but had continued to run the

61

business until late 1994.74 By this time, many engineers felt that the largest computer-

networking issues had been resolved. Lynch recalled that he left when the trade show

started to become what he called "overrun with marketers" (interview 2009). Interop

continues to operate to this day - and is in fact one of the largest enterprise networking

trade shows in the world. Still, Lynch's point is well taken, for the energies that had

invigorated the production of physical networks had largely given way to excitement over

the World Wide Web. Although the core membership of the Interop network seemed to be

fraying, the visions around the network forum remained intact for a time. Interop affiliate

Carl Malamud was still focused on the global need for connectivity. His Technical

Travelogue had offered a detailed representation of a disorganized and heterogeneous

emerging Internet, a half-formed vision populated by a range of individuals with distinct

problems and goals that differed from the carefully manicured heterogeneity of Interop. In

short, Malamud's publication presented a different aspect of the difficulties of scaling a

technology. He had just begun to scheme how to pull off his most far-reaching endeavor yet

- an international exposition organized on the Internet - but also for the Internet.

74 After Ziff-Davis acquired another information technologies trade show, Networld, Interop wasrenamed Interop+Networld trade show. Despite these institutional changes, however, this study willretain the term "Interop" throughout to describe the trade show.

62

Chapter 4: In Truth, All the World was There: The Internet 1996 Exposition 7 5

By the mid-1990s, dramatic increases in public computer networking as well as the

expansion and privatization of computer networks had helped facilitate the growth of a

series of commercial and alternative networks. For Carl Malamud, faster networks brought

with them the promise of new services and products, and with that, the possibility of

additional consumers. Malamud had authored numerous technical resource manuals but by

1993, he began to actively explore the communications applications in the online space. He

started a non-profit organization, Internet Multicasting Service (or IMS).76 Malamud's

choice of the term "multicast" (sometimes called Multicast Backbone, or "Mbone") referred

to an experimental method for sending audio and video over existing Internet

infrastructure that would cut down on the expense of sending large data files that also

tended to overload existing bandwidth.77 It also revealed his intention to become a "desktop

broadcaster" (Malamud, 1993b).

Through IMS, Malamud developed projects that explored the possibilities he

envisioned for the new medium, including a 1993 effort to integrate fax and e-mail. This

initiative, which Malamud developed with Interop affiliate Marshall Rose and debuted at

Interop, was conceived as a kind of "community library" that would service the public "over

a portion of the telephone address space." 78 Malamud also launched one of the first Internet

radio stations, an initiative he called "Internet Talk Radio." 79 His online-only service, which

offered recordings of National Press Club luncheons as well as a "Geek of the Week"

75 Frederick Ward Putnam (Chief of Department of Ethnology, World's Columbian Exposition) 1891,quoted in Griffiths 2002, 46.76 Christophe Diot et al, 1997. "Multipoint Communication: A Survey of Protocols, Functions, andMechanisms." IEEE JSAC.Brown, Ian, Jon Crowcroft, Mark Handley, and Brad Cain. 2002. "Internet Multicast Tomorrow."The Internet Protocol Journal. http://www.isoc.org/pubs/int/cisco-1-6.html.77 I find Andy Lippman's example of the video streams available on airplanes to be a handy way ofthinking about multicasting, versus broadcasting, which assumes similarly assumes a one-to-manymodel but is considerably less accommodating to the notion of temporal consumer demand.78 In my interview with him in 2009, Malamud suggested that this "hack" upset several of theInterop inner circle, who feared that Malamud had strayed too far into establishedtelecommunications territory. Drawing the attention of the FCC, the Interop organizers believed wastrouble they were anxious to avoid. Malamud recalled that he had been asked to cancel the faxproject at Interop. He went ahead with his plan, apparently with few ill effects. RFC 1529 isassociated with this project: http://www.fags.org/ftp/rfc/pdf/rfc1529.txt.pdf.79 In addition, Malamud launched of one of the first live streaming "cyberstations" at Interop94.

program featuring recorded interviews with Internet pioneers, had an estimated 100,000

listeners in about 30 countries. 80

For Malamud, digital publishing offered an opportunity to develop the Internet in a

manner analogous to the development of networks. That is, he saw the creation of online

information spaces as well as the growth of a "variety of ways of interacting" online,

through electronic mail as well as "real-time video connection," to be a project undertaken

by a group of experts who would guide the development of quality material online. A 1993

interview suggested that Malamud took his role as a "desktop broadcaster" seriously,

basing his IMS company in the National Press Building among traditional media

representatives.

It is the global village, and we need people producing real information... Wewant to see professional production on the Internet. We want to show NPR,and CNN, and these other groups, here's how you, who produce information,after all ... here's a new medium you can send your information out onto ...We are next to the Kansas Star Gazette and the Arkansas Gazette. And onthis door down a long hall, you'll see Internet Multicasting Service. ... a roomthat's kind of half-radio station, half-TV studio. And a whole bunch ofcomputers. We have the fastest link in Washington DC to the Internet. ...[W]hat we're ready to do there is pump large amounts of data into thenetwork ... It is the global village, and we need people producing realinformation. (Malamud 1993b)

Malamud's statement also reveals the intimate connection he saw between connectivity,

which in this case meant a faster network, and, in turn, new services and new consumers.

By 1994, this conviction would help drive Malamud's decision to undertake one of

the most ambitious Internet projects of its time. Nearly a decade after Dan Lynch had first

assembled the former ARPAnet researchers to brainstorm about the future of the Internet,

Malamud employed a version of the Interop trade show as a model for the Internet 1996

Exposition, a year-long international trade fair and exhibition that set out to drive greater

connectivity. It comprised a website that employed flashy graphics and midi audio files, 81 a

series of online exhibits, and geographically located events permitting face-to-face meetings,

as well as network structures such as a global network "backbone" and multiple computer

80 This audio was originally made available through FTP. These recordings are still available today,now through the World Wide Web. The "Geek of the Week" program is available here:http://town.hall.org/radio/Geek/ and 1993-1995 recordings of National Press Club luncheons areavailable here: http://town.hall.org/radio/Club/.81 The visual style of the Internet Expo website looks quite similar to O'Reilly's Global NetworkNa viga tor.

'34

"libraries." In its membership and implementation strategies, the initiative resembled the

networks that had first formed around Interop trade shows. For the Internet Expo,

Malamud brought together network developers and computer-networking firms as well as

international university researchers-some of the same groups that were already working

in partnership through Interop forums-and employing similar frames, to assemble a

prototype of a global Internet, albeit this time on a worldwide scale and in real time. The

exhibition diverged from Interop in other significant ways, both in the management

structure and the project's framing of the solution. The next few pages provide an overview

of the Internet 1996 Exposition and its components, which the remainder of the chapter

will examine more closely.

The Internet Expo got its start when Carl Malamud approached Vint Cerf - Internet

pioneer and MCI senior vice president as well as board member of Malamud's non-profit

Internet Multicasting Service (IMS) - with the notion of putting on a world's fair. Malamud

was looking for a way to continue funding his company, and he believed that the world's

fair metaphor presented an ideal opportunity. He had been working for several years to

build a communications business on the Internet. For Malamud, communications on the

Internet was still in its infancy and many metaphors seemed appropriate: "We can easily

call ourselves a global schoolhouse, a telephone company, or a radio station." Malamud had

chosen the radio metaphor, framing the collection of projects that IMS comprised-the free

international fax program, the collection of online audio recordings of people building the

Internet, and the online databases of telecommunications and SEC standards-as "Internet

Talk Radio." When the radio metaphor hadn't yielded sufficient funding for his efforts,

Malamud thought about what might prove more appealing to corporate interests,

considering a "global schoolhouse" and a "telephone company" before settling on a "world's

fair" (1997, xv-xvi), a framing that has long proven evocative for technologists. This

metaphor evoked the spectacular displays that emerged in Industrial Era

expositions-from the Crystal Palace in London, to the spectacular lighting displays of the

World's Columbian Exposition in Chicago, to wireless telegraphy displays at the Louisiana

Purchase Exposition, to the vernacular architectures of the 1939 New York World's

Fair-that captured the public's imagination and faith in a idealized technological future.

In less than a year, Malamud assembled an array of supporters. Putting on an

international exposition for the Internet allowed him to mix Interop's hybrid production

635

strategies and cybernetic ideals with the countercultural idealism of the MIT Media Lab

and the status seeking jockeying for position and status of global trade shows. Interop

provided early institutional support for the exposition, and the fair became a keynote of the

Interop trade show gatherings throughout 1996. Jun Murai from Keio University in Japan,

and Dr. Rob Blokzijl, a physicist with the National Institute for Nuclear and High Energy

Physics (or NIKHEF)8 2 in the Netherlands were collaborators. Other supporters included

network developers affiliated with Interop, including Simon Hackett, Joichi Ito, Paul Vixie,

and Mike Millikin. In addition, publisher Tim O'Reilly of O'Reilly & Associates and a

number of MIT Media Lab professors and students helped, including then-student Deb Roy

and Glorianna Davenport's Interactive Cinema Group. Corporate support totaled more

than $100 million in resources and included support from U.S.-based companies like Sun

Microsystems, MCI, IBM, Bay Networks, and UUNET Communication Services as well as

from the Dutch Ministry of Economic Affairs, Deutsche Telekom AG, Korea Telecom,

Samsung, AT&T Jens Corporation, IBM Korea, NEC Corporation, Sony Corporation, and

Keio University. As we see, the international sponsors were concentrated in Korea and

Japan but also included representatives from the Netherlands and Germany. More than 80

"regions" of the world created online pavilions, including Japan, Tibet, Singapore, Egypt,

and the Netherlands. U.S. government presence in the lead-up to the fair was minimal,

although President Clinton sent a letter of support; after the project launched, there would

be a number of exhibits sponsored by the United States.

The initiative had a number of core design elements that defined the scope and look

of the Internet Expo. As we shall see, the fair metaphor - in its reliance on the vernacular

and its sense of the spectacular - deeply shaped the event. Instead of merely trying to

attract the attention of traditional media (as he had with earlier projects), Malamud would

use the exhibition to illustrate the most ambitious aspects of the networked society he

envisioned. These included "pavilions," (online websites that would be open for anybody to

develop),s online "events," and geographic places where the public could interface with the

82Nationaal Instituut voor Kernfysica en Hoge-Energiefysica, or National Institute for Nuclear andHigh Energy Physics has since changed its name to Nationaal instituut voor subatomaire fysica, orNational Institute for Subatatomic Physics). It is one of seven locations of the Amsterdam InternetExchange, an Internet exchange point, that began in 1994 as a collection of Internet serviceproviders83 These pavilions were almost exclusively sponsored by national governments or corporations. Therewere a few exceptions, including "Randyland," which was designed by an individual.

fair. Two final technical components of the fair, which Malamud dubbed the "Internet

Railroad" and "Central Park," were envisioned as its infrastructural legacy. Malamud also

hoped that they would help mobilize corporate (and even national) interests to improve the

expensive yet sluggish connections that were "holding up" the development of the Internet

and keeping it from functioning as it could.

The international links were so overloaded that many were losing 70 percentof all packets by trying to put the equivalent of a grand piano through a mailchute. Using the world's fair as an excuse, we set about trying to beg andwheedle bandwidth out of carriers. (1997, 144)

This additional bandwidth would be required in order to make the audio, video, and real-

time streams "flow" around the world as rapidly as had been envisioned. To achieve this,

Malamud needed telecommunications carriers to partner with one another for the duration

of the event and allocate additional bandwidth or assemble faster connections between

countries and regions. The project received its earliest, and most substantial, support from

the telecommunications firm MCI, 84 which donated backbone resources for a year. Other

carriers and exchange sites would also support the project. Malamud hoped that the

Internet Expo would demonstrate what he saw as a critical need for a global network

"backbone." A final aspect of the event would consist of large computers staged at "key

Internet exchange points" that would mirror data, and thus provide a measure of

redundancy and alternative routes through which Expo pavilions could be accessed,

avoiding "traffic jams" and expensive international connections (Malamud, 1995). These

technical components (the "Internet Railroad" and "Central Park," respectively) will be

examined in more depth later in this study.

The fair launched on January 1, 1996. Over the months, the exhibition enlisted the

support of additional sponsors, and affiliated with additional projects and offline events. As

this occurred, additional pavilions were added to the fair website. Online, the Internet Expo

functioned as a web directory that aimed to be encyclopedic-in the words of co-organizer

Rob Blokzijl, -to "take all aspects of world and society and make it visible to the world"

(1995).85 These various initiatives were organized on the Expo website in several ways: by

location (regions and continents), themes (such as "Cities and Districts," "Food and

Markets," and sub-themes such as "World Art Treasures" and "Mimi's Cyber Kitchen").

84 MCI, like AT&T, was part of the international construction boom in the 1990s.85 Blokzijl, Rob. 1995. A world's fair. Presentation at the semi-annual Networld + Interop, July 21, inTokyo, Japan. http://www.scribd.com/doc/2576764/A-Worlds-Fair.

637

These included initiatives created explicitly for the Internet Expo, and included country-

specific projects. At launch, Japan committed twenty-two corporate projects, including the

multimedia "Sensorium" - more than any other country. Additional exhibits included the

Netherlands' simulated cow pavilion, art and technology exhibits, and an IBM-sponsored

Mongolian road race. Some of these events were created specifically for the exhibition, but

many more, including Malamud's "Congressional Memory Project," 86 Ted Machover's "Brain

Opera,"8 7 the "CyberFair96,"8 and the chess match between world champion Garry

Kasparov and IBM's "Deep Blue" computer,8 9 were events that the exposition "linked to"

from elsewhere on the Internet.

At the end of 1996, the Internet Expo event officially came to a close, celebrated with

a closing ceremony in Kobe, Japan, that included blessing an exhibition "time capsule" - a

digital videodisc of the main portion of "Central Park" that had grown to over 10 gigabytes

- that would be stored in the City of Kobe Museum. The exhibition website had received

some fifty million "hits," with an estimated five million unique "visitors" (Malamud 1997,

172-173X 90 In the months following the fair's closure, Malamud set about transforming the

website "fairgrounds" into a "public park." He added an online map to the exposition

website, depicting the event as an enclosed outpost of pavilions connected by rail in an

86 This was largely a repackaging of Malamud's work through the Internet Multicasting Service. Init, he recorded nearly ten months of U.S Congressional feeds to a database, where the audio wassearchable by member of Congress, date range, location, or political party affiliation.87 MIT's Ted Machover debuted his Brain Opera at the MIT Media Lab, where it was recorded for theInternet Expo website (and streamed to a convention center in Japan), as well as at New York'sLincoln Center Festival. The performance was based on Marvin Minsky's book, The Society of Mind,included a set of "hyperinstruments" designed by Machover and his Media Lab students.88 The Cisco/MCI Global Schoolhouse CyberFair96 was an initiative to help schools get online. Morethan 350 schools signed up, posting information about their school as well as various photos anddesigns from students.89 To celebrate the fiftieth anniversary of the electronic computer, IBM and the Association forComputing Machinery (ACM) sponsored a chess match between Kasparov and the corporation's"Deep Blue" computer. The speculation about "thinking" machines and the ways in which the gameof chess explored the bounds of machine "intelligence" is a rich topic, which has been explored at TheComputer History Museum in Mountain View, California. More information is available here:http://www.computerhistory.org/chess/.9 The term "hit" refers to the number of times a file is sent to a browser by a web server. Since awebsite is comprised of many different files, a single request to view a single webpage can generatenumerous hits. In contrast, a "visitor," suggests an individual accessing a website, although onevisitor can make multiple visits to a site. By comparison, according to a May 1995 article inInteractive Age, Netscape.com reported 30 million hits and 3 million users daily and HotWired.com,a online spin-off of Wired magazine, reported 3 million hits and more than 400,000 users daily.http://www.cs.columbia.edu/-hgs/internet/notes.html.

68

otherwise unpopulated frontier. Also in 1997, Malamud published a coffee table book

detailing the planning and execution of the event. The website (www.park.org) remains

online, an archive of the Internet 1996 Expo but also as a record of the Internet as it

appeared in 1996, and intended to be a "pristine structure that will remain forever present"

(1997, 253) (see Figure 8).91

A WORLD'S FAIR FOR book 'b- 1ThE INFOir

N ew! Full-color book about th. fair!

Figure 8: Screen shot of the Internet 1996 Expo website, www.park.org.

91 Malamud envisioned that he would maintain an archival "snapshot" of the project. The notion that

the Internet might function as a "library" was a common metaphor in the 1990s. Today, many sites

and events affiliated with the Internet Expo have since been moved or taken offline. The website

remains an archive of the web capabilities at a particular point in history. Unlike a site that is

constantly updated, it can seem incongruous today, prompting a blogger who recently came across

www.park.org to complain that "instead of leaving a recyclable graveyard of architectural oddities,

what is left is frighteningly static ... Clicking through to view the exhibits leads either to shell sites,

diversions to vast telecom conglomerate promotions sites, or the familiar old 404 not found

tombstones ... despite over 100 million US dollars from diverse governments and corporations

funnelled [sic] into it." Everything2 blog, http://everything2.com/title/Internet%25201996%2520World%2520Exposition. More files are available on the CD-ROM that accompanied the book

publication, although some require "vintage" plugins to run properly.

Today, however, the many links on the site itself are no longer working, an artifact of the

early Internet, and a time when many believed that the Internet might store permanent

records of knowledge and events. Let us turn now to the closer examination previously

promised.

Reviving the Spectacles of the Industrial Age

By the mid-1990s, the dramatic increase in public computer networking as well as

expansion and privatization of computer networks had facilitated the growth of a series of

commercial and alternative networks that had sprung up, promising to usher in the

digitized meritocratic marketplace communications medium. Interop founder Dan Lynch

had accomplished much that he had set out to do with the trade show. He had helped

ensure the success of the Internet protocol. He had also been a critical force in the assembly

of heterogeneous networks into a global Internet that had become intertwined with the

global economy.

For Carl Malamud, in contrast, the physical expansion of computer networks as well

as the growing commercialization of the Internet had led him to conclude that network

infrastructures had advanced sufficiently for him to articulate his own visions of

connectivity. On one hand, the Internet Expo modulated Interop's collaborative practices,

persuading partners to work together to assemble a functioning exhibition network that

would simultaneously demonstrate connectivity while also offering a near-future experience

of what the Internet could become, and of the role that corporations would play in this

future. At the same time, the Internet Expo diverged from the Interop trade show in a

number of ways. Whereas Interop emphasized the critical importance of successfully inter-

linking heterogeneous networks to one another, the Internet Expo emphasized the speed

and bandwidth of the connection. Both stages, of course, are critical to the "seamless

integration" of computer networking technologies, yet they operate at different stages in the

expansion of a technology. Malamud believed that a critical next step in bringing the

Internet to scale would require a global public demonstration that would include

governments as users. He also believed that it would require an additional driver:

consumers, who, presumably, would bring with them the promise of the commercial

viability of products and services.

70

In order to accomplish his goals, Malamud enlisted the support of a range of

interests. In July 1995, Malamud stood before a crowded conference audience at the Interop

trade show in Tokyo and formally unveiled the Internet 1996 Exposition: A World's Fair for

the Information Age. As he had previously, Malamud relied on a usable past, imagining

himself in intellectual connection with the actors of previous technological systems, and

actively sought to make this very connection to his international audience.

It is tempting to say that we are living in unique times, but if we look athistory - if we learn from history - we will see many parallels between ourinformation age of this century and the industrial age of the past. (1995)

The past that most interested Malamud was the succession of global public expositions that

ushered in the industrial age, spectacles that simultaneously astonished and subdued

audiences - from the mechanization of factories (1876) to the standardizing influence of the

railroad (1893) to the managerial efficiency of assembly lines (1915) to the commoditized

futures of the "World of Tomorrow" (1939) - events that integrated the technological

processes and the corporation ever more completely into society. Recalling these

expositions allowed Malamud to frame the introduction of the commercial Internet through

the lens of inventions that were both technological and territorial.

As he had in the Technical Travelogue, Malamud depicted standardization as a

critical battle in which progress and a better future were at stake. An exposition, Malamud

offered to his audience, functioned as a critical site in the success of a technology. To prove

his point, Malamud relied on the usable past of George Westinghouse and his struggles

with Thomas Edison over the future of electricity. The battle, Malamud suggested, was

resolved at the 1893 World's Columbian Exposition in Chicago.

This was 1893 ... This was the birth of electricity and there was a big fightgoing on. A guy you may have heard of, Thomas Edison, had got into thepower business. He was championing a power distribution technology calleddirect current, DC. ... but DC had problems. Most of the power got lost in thedistribution network.

A bunch of young Turks had come up with a radical new technology calledAlternating Current, AC. They claimed AC would allow efficient powerdistribution over long distances, but Edison ... waged a bitter publiccampaign, telling people how AC would harm their health, how the

technology was unstable, was untested, that AC was nonstandard and wecouldn't allow every group to come up with their own standard.

One of the leaders of the young Turks was an engineer named George

Westinghouse. He got the contract to build the show network for the 1893

71

Chicago Columbian Exposition. He put in 22,000 horsepower of generatingcapacity. Chicago was a great success. Soon after, George Westinghousereceived a contract to place his equipment at Niagara and the modern powerindustry was born. (1995, 11)

In his recounting of the story, Malamud promoted a version that emphasized the conflict as

a "battle of the currents," the version of the struggle between two competing technological

systems that was popularized in the press at the time. This downplayed the degree to which

the controversy played out on technical, economic, and political levels, and the degree to

which it was resolved on these levels (Hughes 1983, 106-140). The specifics of electricity

aside, Malamud's message was clear: he was celebrating the work of scaling a technological

system and transforming society, emphasizing it over the invention of the technology itself.

By describing the wiring of the Chicago Exposition as a "show network" (and later, the

control of this show network as a "network operating center"), Malamud expressed this past

success in terms familiar to his Interop audience (1995, 8, 10-11). In the process, he

suggested that the exhibitions functioned as critical sites of technological change. In other

words, Malamud enticed his audience to become part of the reorganization of society and

the economy by supporting his proposed Internet exhibition.

Malamud's story likely operated on another level as well. In my interview with him,

Malamud suggested that a new generation had driven the Internet's expansion. Although

this study argues that original ARPAnet researchers were, in fact, deeply involved in the

expansion and commercialization of the Internet, Malamud's statement implies something

about the multiple stages of standardization that a technological system undergoes. In the

1980s and early 1990s, Interop trade shows defined their user base largely as businesses

and consumers. However, by the mid-1990s, the increasing availability of personal

computers and the advent of the World Wide Web had greatly increased the number of

users - and hence not only their perceived commercial importance but also their "claim" in

the network. At the same time, the Internet was still far from a developed technology. The

network risked fracturing into multiple competing models. Significant gaps in the

networking infrastructures remained. As Internet traffic grew exponentially, outages

proliferated. In a 1995 InfoWorld column,9 2 network engineer Bob Metcalfe predicted that

the Internet would collapse by the end of 1996. Malamud had come to believe that what

92 Metcalfe, Bob. 1995. "Wireless computing will flop - permanently." Infoworld 15(33): 48.

72

was needed was to drive development that would result in greater connectivity and faster

networks.

In a manner reminiscent of the Interop tactics, the Internet Expo offered a prototype

for how collaborative partnership could drive improved connectivity. In a 1995 promotional

video, Vint Cerf suggested that an Internet railroad would link "various cities together, and

expose the various populations to the wonders of the Internet" without "any freeway

congestion." Their efforts, Cerf suggested, would be accomplished through the traditional

Internet leadership, which included Cerf, and the telecommunication corporation that

employed him.

We'll be able to "deliver the goods" just as we did with the railroads of the1800s. So please join with The Internet Society and with MCI to help buildthe Internet railroad for the 1990s. (1995)93

Malamud gambled that this tactic of driving collaborative partnerships that had worked so

successfully for Interop would help to generate the political and corporate will to assemble

fast enough networks so that the kinds of services that would attract consumers would be

easily accessible. As with electricity's rapid integration into society, Malamud imagined

that the seamless integration of computer networks into daily life was critical if the

technology were going to become "useful" for the larger public.

The computer must disappear, becoming part of the facilities instead of ashowcase on stage. In the early days of electricity, there were no electricaloutlets. Wires ran all over the place and homeowners became adept atstringing new appliances directly into the mains. Over time, we learned howto make the infrastructure disappear, to become a natural part of buildings.(1997, 31)

Just as world's fairs had left "lasting impressions on the landscape ... and on the minds of

their visitors" (1997, 27), Malamud imagined that the Internet Expo would help address

what he saw as a critical danger for the commercially operated Internet. Worrying that

financial interests would leave little "public space" for citizens, Malamud would create two

"architectural legacies": the "Internet railroad" to drive connectivity, and the "Central Park"

as a series of global repositories of data. Malamud suggested, in a manner reminiscent of

Disney's Spaceship Earth, "We are trying to get consumers to move to the global village, to

bring this technology into their homes and businesses, to bring this technology into their

93 Cerf, Vint. 1995. World's fair promo tape. Internet Multicasting Service presentation at the semi-annual Networld + Interop, July 21, in Tokyo, Japan. http://www.archive.org/details/org.park.expo promo.

72

daily lives. ... This Internet's worlds fair is about public parks, but it is also about building

the infrastructure that will allow our information economy to succeed" (1996, 25-27).

Malamud had helped build numerous Interop show networks. For Malamud, the Internet

1996 Expo would be the ultimate ShowNet.

Exploring the "Global Village"

The 1990s were driven by market populism and an enthusiasm for the "new economy" that

celebrated private investment, entrepreneurs, and deregulation. Many pundits also rejected

any role that the government might play in the development of the Internet. Yet the

Internet Expo not only relied heavily on corporate sponsors, it also engaged numerous

government and other bureaucratic organizations.

The Internet 1996 Expo officially went online in January 1, 1996. In actuality, the

pavilions and the infrastructures designed for the exhibition came online gradually

throughout the year. The exhibition was produced around the same time as two other

projects that explored the affordances of the Internet as a global communications medium.

All had some connection to the MIT Media Lab, though all but Malamud's Internet Expo

were generally conceived as online artistic exhibitions. The first, a book/web project

produced to celebrate the research group's 10th anniversary, was A Day in the Life of

Cyberspace.94 For ten days in October in 1995, the site's organizers pulled in stories and

other materials covering a number of themes, including Privacy, Place, Expression, Wealth,

and Environment. In her thesis on the project, co-designer Judith Donath suggested that

the virtual event was intended to offer a "Portrait of the Net, 1995" that would "encourage

people to think about how cyberspace is developing and its impact on their own lives and to

send in writings and pictures about their experience with this new world" (1996).

A similar hybrid book/web project, photographer Rich Smolan's 24 Hours in

Cyberspace: Painting on the Walls of the Digital Cave,95 was timed to launch on the same

day as the grand opening of the Internet Expo. Smolan's project pulled together a team of

150 photojournalists who, on February 8, 1996, "fanned out across the world to document

how the Internet and online communication are changing people's lives." For twenty-four

94 Donath, Judith. 1996. A Day in the Life of Cyberspace. Cambridge, MA: MIT Media Lab.Multimedia project. http://www.media.mit.edu/events/1010/1010 intro.html.95 Smolan, Rich. 1996. 24 Hours in Cyberspace. Cambridge, MA.http://undertow.arch.gatech.edu/homepages/virtualopera/cyber24/SITE/htm3/toc.htm?new.

74

hours, a team of computer programmers and editors worked in real time to download the

photographs sent from the field and to put the best ones online. The site contained about

200,000 images and allowed users to add their own pictures and stories. An estimated four

million people visited the site over the 24 hours the site was active. The project was later

released as part of an exhibit at the Smithsonian Institution and published as a book and

CD-ROM. By far the most ambitious, both in scope and scale, was the Internet 1996 World

Expo, which, unlike the other exhibits, was explicitly focused on the infrastructures of the

Internet, and therefore, on technical barriers to connectivity.

Although the fair was produced, in part, while Malamud was at the MIT Media Lab,

the visual style of the fair had the most in common with O'Reilly & Associates' Global

Network Navigator (or GNN),96 which had been the first commercial website (and the first

online advertising) on the World Wide Web. 97 In a similar manner, the Internet Expo site

employed digital interfaces that drew on elements employed on the GNN site. This included

a Whole Earth countercultural vernacular that drew on visual elements like balloons. Like

the GNN, the Expo highlighted sponsors and other commercial elements on the site.

However, the critical aspects of the Expo were not its visual style, but rather its treatment

of infrastructures.

Prototypes and Corporate Infrastructures

Malamud used the metaphor of the world's fair in one final way to draw attention to the

most critical aspects of his project: the construction of large public infrastructures. These

two main elements consisted of the "Central Park," comprising a dozen donated servers

located at key Internet exchange points around the world and an Internet Railroad

(originally conceived as a "globe-girdling" T3 line) that would "supercharge" Central Park

(1995). Both were intended to improve connectivity as well as the quality of content online,

and would be funded largely through corporate support.

96 O'Reilly and Associates, Inc. launched GNN after an early prototype of the site was firstdemonstrated at the Interop 92 trade show. At this point, the Internet was overseen by the NSF,which had rules against commercial activity online. O'Reilly obtained a special dispensation to putonline advertisements on his site. The GNN home page as it looked when it launched in 1993 can befound here: http://oreilly.com/gnn/.97 Also in 1993, the graphical browser, Mosaic, was made available to the public for the first time,quickly becoming popular enough to drive growth in the World Wide Web itself. This growth wasfollowed by the commercialization of services on the Web.

7.5

The linked machines of "Central Park" functioned as storehouses of web sites,

multimedia, and other data that was amassed and systematically distributed to other key

machines around the world. They helped compensate for the technical difficulties and

expense of sending large multimedia files over long distances. They also reflected the

degree to which Malamud conceived of his project as a web directory that was not only

assembling but also recording all of the content available online. Curiously, once data had

been collected, Malamud noted that a provider would be free to take this information and

"sell it or give it to their users."

Despite the contemporary popularity of the "information highway" metaphor, for the

second technical element of the Internet Expo, Malamud instead chose the metaphor of an

"Internet Railroad." Alluding to the railroad as a mass (and not individualized) transit

mode that helped to industrialize the U.S., Malamud noted:

The backbones are carefully managed infrastructures that aggregate trafficfrom thousands of simultaneous users. These key transit links are intenselymonitored and planned. The term information highway implies a wide-openspace that people wander about in. A transit backbone is more like a train,where packets arrive at a router, queue up until a slot becomes available, andare injected into the long-distance links. If the current Internet is a set ofunpaved country roads that may someday lead to the information highway,our backbones are truly the narrow-gauge rails of the beginning of thenineteenth century. (1997, 143)

In this way, the prototype served the larger political goal of trying to elicit the proper

support that would serve as "the first step toward a real global infrastructure" (Malamud

1995). Even in countries with Internet access, links between countries were limited (1997,

144). The numerous links between the U.S. and Europe were very slow, but four times

faster than Japan's connectivity to the rest of the world. In part because Asia had the least

developed infrastructure, the Internet Railroad had the largest impact there. In Japan,

NTT donated fourteen T3 lines to connect all of the regions together. JCSAT committed two

full transponders off two satellites, bringing connectivity to Japan as well as to the entire

Pacific Rim. This line would also connect various locations, providing data exchange and

real-time audio/video streams. On a larger structural level, the emphasis on physical

infrastructure and technology transfer as well as connectivity extends the railroad

metaphor to include its deep connections to the imperial age and colonialism.

At the end, however, the Internet Railroad was temporary. Almost all of these

infrastructures were donated only for the duration of the Expo, and at the end of the trade

76

show, many of the links were returned to their original purposes. However, by pulling off

such an ambitious infrastructural prototype, Malamud had hoped to encourage providers to

build more robust networks.

We weren't network operators and we didn't see any point in competing withthe commercial providers. The whole point of the world's fair was to go onestep ahead and provide a spur to accelerate the development of the Internet.

(1997, 154-155)

To do so, he had assembled these infrastructures (and the whole exhibition in fact) using

strategies employed at Interop. In an account of the event, Malamud noted:

The bottom line for us was that we were able to build the infrastructure forour world's fair, just as the engineers in 1893 installed lights and trains andthe other networks that they used for theirs... For one year, the InternetRailroad was an international service provider with operations in a half-dozen countries, 24-hour network operating centers, and a host of users atuniversities, special event sites, and Central Park sites. The railroadprovided an ideal customer story for the contributing companies becauseeverything we did was out in the open ... [M]ore importantly, the regionalbackbones, national backbones, and special event sites provided idealtraining for the engineers participating. (1997, 155)

Malamud had once envisioned himself as an explorer. With the Internet Expo he functioned

as a salesman exporting technology and the technological practices of Interop. The

exhibition had drawn sufficient attention through the year that it was active, and even

achieved some success in encouraging Japan to improve its internal national networks as

well as its connectivity to external networks. Yet once the fair closed, it was largely

forgotten. The Internet Expo was never as successful as Interop. It was never able to foster

the kinds of long-term partnerships that occurred through the Interop trade show, nor did

the exhibition model truly offer the right environment to foster collaborations between

fierce competitors. Perhaps most critically, by venturing into collaborations with national

governments and a wider array of commercial sponsors, the Internet Expo ventured

squarely into territory that Interop organizers had always adroitly avoided: namely, the

degree to which the Internet not only challenged the traditional territories of

telecommunications industries, but also of numerous national governments that had

historically seen telecommunications as the realm of the state. The model of a world's fair

might have had the right mix of corporate and national competitive qualities to convince

numerous entities to participate, yet it was unlikely that such an exhibition would ever

have led to the kind of collaborations and flexible partnerships that emerged through the

77

military-industrial research world. Notwithstanding, at the same time, the success of

Malamud's project was less critical because the exhibition represented an early vision for

how a truly global network infrastructure might have emerged.

78

Conclusion

This study ends with a final 1990s-era example that helps more precisely define the degree

to which physical computer networks embody the politics of network infrastructures. In

1996, a multi-national conglomerate was laying a 17,000-mile transoceanic fiber-optic

cable98 that, when completed, would circumnavigate the world, running from England

across Egypt and through Dubai before crossing the Indian Ocean to reach the Asia Pacific

rim. The Fiber-Optic Link Around the Globe, or FLAG, was part of a frenzy of laying fiber-

optic cable in this period, spurred on by expectations about high-bandwidth real-time

applications on the Internet (Chun 2006, 27).

In his account of FLAG's construction, author Neal Stephenson wrote that in the

"deregulated telecom environment in the United States," the Internet had grown like an

"exotic weed ... thriving, colorful, wildly diverse, essentially peaceful, and plagued only by

the congestion of its own success" (Stephenson 1996). Yet Stephenson also noted that the

FLAG initiative exposed a critical weakness in the so-called "nethead" narrative that

computer networks were ushering in an ideal society. Suggesting that such a view

"overlooks much history and totally misconstrues the technology," Stephenson noted that

many of the same corporations and their affiliates that had built the wires, cables, and

other transmission media wiring the world together for a century and a half were now

laying the fiber that would make up the Internet. The global telecom business was "so

tangled that no pure competition exists," Stephenson wrote. "Most of the companies ... have

their fingers in pies in dozens of countries all around the globe" (1996). Many of these

companies had helped spawn the Industrial Era as well as an earlier wave of globalization

defined by global telecommunications systems built for economic and military purposes

(Hugill, 1999). Put another way, the snarl of physical cables and hardware of information

technologies exposed the degree to which the distributed, networked forms of management

in fact co-existed with apparently contrasting systems and, in fact, overlaid them in ways

that were not distributed uniformly across time, space, and cultures.

Despite these inherent contradictions, Stephenson concluded that projects like

FLAG would "help blow open bandwidth and weaken the telecom monopolies." Over time,

98 At the time, fiber-optic technology was widely heralded as a dramatic advance for transatlanticcommunication because, unlike technologies like coaxial cable, fiber optic promised far greaterinformation capacity with less of the distortion and degradation of the data that have alwaysplagued telecommunication carriers. For a more on fiber-optics, see Jeff Hecht. 1999. City of Light:The Story of Fiber Optics New York: Oxford University Press.

79

they could substantially reduce the physical challenges of global networked computing,

leaving only "the cultural barriers that have always hindered cooperation." Stephenson's

closing comment is revealing because he forecasts that people, not technologies, will tend to

"fail"-through resistance or error, thereby limiting the promise of universal connectivity.

Infrastructures are generally invisible, functioning seamlessly within society until they

break down. This moment of "willful disconnectivity"99 becomes powerful because it offers

an account of networks at the periphery (a zone that is decidedly less inviting than the

interface of personal computers) that exposes the "modernizing" logics of networks. It also

exposes the allure and the implications of distributed, networked forms of management, an

assumption that Stephenson was less willing to interrogate. Networked computing has

indeed functioned as a technology that is not only democratizing, empowering individuals

by allowing them greater market privileges, but also one that, despite the centralized and

hierarchical structures of previous systems, operates according to its own managerial

structures.

Infrastructures are similarly visible when they are under construction, as they were

in Silicon Valley in the 1980s and 1990s. These networks, conceptually and physically half-

formed, revealed the visions of the engineers themselves, which have been the focus of this

study. As evidenced by their willingness to "disconnect" users from the Internet for various

infractions, these engineers understood that the global information technology system they

were constructing required widespread adherence to be effective. In this way, the half-

formed nature of these networks also directly confronts how the technology - an invention

first developed to address the U.S. military's need to promote a flexible, heterogeneous

system able to string together a diverse range of command and control systems (Abbate

1999, 144) - was "normalized" in order to become a commercially viable communications

medium. These infrastructures, or at least traces of their presence, are discernible across

Disney's simulated landscapes of the emergence of the networked age.

This study began with Spaceship Earth because, unlike the visions of networked

computing extolled through the Whole Earth or Interop networks, the instrumentality of

the exhibit's narrative is never in question. In particular, the theme park has always

99 Galloway (2004) has defined "disconnectivity" in technical terms, which might include a Denial ofService (DoS) attack (that either involves overwhelming the targeted machine with externalcommunications requests, rendering the device unable to respond to legitimate traffic, or respondingtoo slowly to be effectively available), or an instance when an Internet Service Provider (ISP)controls or cuts off a user because of a time limit.

80

explicitly marketed the notion of progress itself, integrating artifacts and iconic moments

into "coherent ensembles" from which visitors could glimpse a future that is at once

computational and corporate (Nye 1994, 205). Although Epcot was directly inspired by the

corporate futurism of the 1939 New York World's Fair, the Disney theme park envisioned

the proper role of corporate forces as one infused with cybernetic rhetoric that viewed

human beings (and their histories) and technological systems as interconnected. This tone

has shifted and softened over the years; now, the exhibit extols the myth of the "guy

tinkering in the garage," downplaying the Internet's origin as an experimental, U.S.

military-funded solution to the tactical problem of connecting dissimilar networks in a

polarized Cold-War era. Yet the interconnectedness of systems theory infuses the entire

corporate exhibit, suggesting that networks are built upon powerful but nonpublic market-

driven decisions and military-subsidized research and development. For this reason

Spaceship Earth functioned as an ideal site from which to begin an exploration of the

military-industrial research world's role in the physical construction of networks and

networking hardware that led to the commercial success of the Internet in the 1990s.

Aspects of this question have been approached by a number of scholars. Fred

Turner, for example, suggests that cybernetic discourse and the collaborative,

interdisciplinary work styles of the military-industrial research world intertwined with the

American counterculture to help fuel what would become a widespread utopian vision that

computer networks would usher in an ideal society. By the 1990s, descendants of this

research world - organizations like the Stanford Research Institute (SRI), the MIT Media

Lab, and the Santa Fe Institute - became "models of a collaborative world ... in which

technologies were rendering information systems visible, material production processes

irrelevant, and bureaucracy obsolete" (Turner 2006, 178). These models, and the

relationships they supported, helped blend countercultural and cybernetic rhetoric and

practice in ways that helped corporate executives model and manage their work in the post-

industrial networked economy. Yet this analysis offers less insight into the physical

construction of networks.

The role of the military-industrial world in the commercialization of the Internet has

also been addressed on the technical side. Some analyses have focused on the groups tasked

with overseeing the standard-setting and architectural design processes - the Internet

Activities Board (IAB) and the Internet Engineering Task Force (IETF). Yet these

81

organizations were less suited to respond to the practicalities of implementing these

standards, particularly at scale. As Janet Abbate suggests, "perhaps the key to the

Internet's later commercial success was that the project internalized the competitive forces

of the market by bringing representatives of diverse interest groups together and allowing

them to argue through design issues" (1999, 144), a collaborative tactic formed in the

military-industrial research labs. The present study contributes to this existing body of

research by suggesting that the Interop trade show network, as a series of forums where

former ARPAnet researchers partnered with commercial interests, functioned as one of the

systems that Abbate describes.

The present study has focused on the network of individuals and activities around

the Interop trade show, suggesting that, unlike other mechanisms of standardization, the

show network not only offered a manner of ensuring partnerships among a set of diverse

and often competing interests, but also offered a mechanism for testing standards in a

technically complicated and commercially competitive environment. This suggests that

Interop functioned in tandem with the established RFC documentation process (and the

organizational logics comprised),1 00 addressing the practicalities of implementing these

standards across domains. Such strategies ensured that the Internet's core organizational

logics would be adaptable enough to transform into a private commercialized infrastructure

and survive the resulting fragmentation of authority.

The figures most closely associated with Interop were actively involved in securing

the Internet's future and explicitly integrating the Internet into the emerging global

economy. For most of the network engineers affiliated with Interop, this expansion was

driven by an attention to interoperability, a goal that envisioned interconnecting machines

that were, at least ideally, interchangeable, openly sharing and processing information.

This imperative drove toward "open systems" that, according to Chris Kelty, amounted to

"openness through privatization," a formulation that equated the marketplace with the free

exchange of knowledge and fought against the proprietary solutions that threatened

monopoly control by corporations over products.

In particular, then, Interop was driven by the practical need to ensure that the

flexible TCP/IP standards, first built to satisfy military conditions, thrived in the global

100 In his work on protocols, Alexander Galloway has referred to these logics as the"governmentality" of information systems (2004, xviii).

82

open market, and thus become the de facto standard for global networked computing. To

accommodate the NSFNET, which oversaw control of the Internet in the early 1990s and

banned commercial activity on the network, as well as leverage the widespread support

that Internet protocols and practices had in the computer science community at large,

Interop organized the expansion of the Internet through universities around the world. By

doing so, they avoided engaging with national governments, and the attendant flood of

difficulties, including competing protocol standards that already had the support of many

governments as well as competing claims of ownership from nationalized telecom

industries.

By contrast, Carl Malamud, focused on connectivity (and openness) as a means to an

end, more interested in what faster networks could mean for increased services and new

communities. His attitude was perhaps most clearly articulated by his Marshall McLuhan-

inspired mantra, "the medium is not the message" - a technological vision, but one that

focused on the delivery of the content and not the physical networks themselves to deliver

on the promises of an ideal society. Although Malamud was deeply immersed in Interop's

goal to transform the Internet into a vehicle of global enterprise, he also tended to advocate

for an articulation of the commons, expressing ambivalence about what a wholly

commercial turn would mean for more civic-minded activities on the Internet. These efforts

revealed a crucial difference between Malamud and many Interop engineers with deep ties

to the military-industrial research world. Individuals like conference founder Dan Lynch as

well as Vint Cerf and David Clark belonged to a close-knit group of former ARPAnet

researchers working to retain substantial authority over the Internet, most explicitly as

representatives of the IAB. In contrast, Malamud was not only a generation younger than

these Internet pioneers, but he came from a different "user community" that shared more in

common with early commercial publishers like Tim O'Reilly of O'Reilly and Associates and

other technical groups. These distinctions would become even more apparent as the most

substantial challenges of routing information between computer networks were solved-and

the Internet moved into a new phase of standardization and expansion. In other words,

spectacles like the Internet Expo focused on technical aspects of networks, and on the need

for greater connectivity, in order to allow the affordances of built networks to flourish.

Perhaps the greatest point of departure in this regard was Malamud's Internet 1996

Exposition. Employing many of the same strategies and figures involved with Interop,

Malamud produced an event that not only successfully demonstrated the viability of faster

networks, but also explicitly highlighted the role of governments and other state actors

that, until this point, had largely been excluded from Interop-style network expansions. To

further appeal to them, he even touted the potential consumer appeal of a massive

spectacle that traded on the nostalgia and excitement of a world's fair. These tactics, more

than any of his other provocations, likely annoyed the Internet leadership. They studiously

worked to define the Internet for its technical attributes, not for its communities; and had

fought even more powerfully to work outside of the regulatory and political boundaries of

international law and of national governments and commercial enterprises. Malamud, in

contrast, invited these parties to the table.

Next Steps

This study has approached the Internet's commercial transition from the particular

perspectives of a relatively small group of network engineers, most with direct ties to

ARPAnet, who were physically based in Silicon Valley and involved with the Interop trade

show. It has suggested that Interop played a critical role in the implementation of the

RFCs, the technical standards that define the core operations of the Internet, and as such

should be considered alongside this well-researched technical standards-setting effort.

Beyond the particular contributions of the network engineers affiliated with Interop, a

much broader story remains to be told about the trade show network. This next stage of

research might be generally conceived in two ways.

First, more research needs to be conducted on the role of additional technical

publications and conferences in the commercialization of the Internet in the 1980s and

1990s. Of greatest interest is O'Reilly Media, which organized technical publications and

conferences, helping make technical aspects of the Internet accessible for a wider audience.

For example, O'Reilly Media was not only actively involved in publishing programming

handbooks that continue to be the definitive works in this field, but also published one of

the first guides to the Internet, Ed Krol's Whole Internet User's Guide and Catalog (1992)

and launched the first commercial website, Global Network Navigator (1993). Although it is

clear that organizations such as O'Reilly Media were influenced by the Whole Earth

network and its steady stream of publications, they don't seem to have overlapped

significantly.

Another arena beyond the scope of the current research would more deeply examine

Interop's role in response to explicit efforts to build global infrastructures amenable to U.S.

national and corporate interests at a time when global economic and technological forces

sweeping the industrialized world were the source of considerable anxiety over the United

States' ability to retain global technological and economic leadership. In the mid-1980s, for

example, Japan was perceived as a particular threat because it had launched a joint

government-industry-university research effort focused on high technology, namely,

artificial intelligence, parallel processing, and microprocessing technologies. Of particular

interest are the mechanisms by which the Internet expanded on a global level, namely the

relationship between Silicon Valley and the various nodes of the emerging Internet, such as

the major trading nations of Japan and the Netherlands (where, incidentally, the Internet

1996 Expo was far more popular than in the U.S.) as well as locales "outside" the

industrialized world. In other words, this research would benefit from a multi-sited history

that better reflects the complexities of assembling a network infrastructure.

In addition, further analysis might more substantially understand Interop's role

within the Silicon Valley culture of forums, partnerships, and demonstrations 10 1 as well as

within the larger economic and social reorganizations underway in the early 1990s. This

would include archival and primary research that could include individuals who

contributed in critical ways to Interop's history, such as David Brandin and Douglas

Engelbart, both affiliated with the military-industrial research firm SRI. It might also

include gathering research from the Defense Communications Agency (now known as the

Defense Information Systems Agency, or DISA) as well as from Defense National

Intelligence, organizations that were involved in facilitating the success of the Interop trade

show overseas, and finally corporate figures from firms such as AT&T, MCI, IBM, Cisco,

and Sun Microsystems. Such research would deepen our understanding of the impact of

trade shows like Interop, which likely shepherded military-industrial concerns into the

global field while integrating networked information technologies into the global economy.

101 Sun Microsystems organized a "Connectathon," in 1986 (http://www.connectathon.ora/), thatappears to have resembled the INTEROPnet. Given Sun's enormous influence on Silicon Valley inthis era, their events might well have been one of the inspirations for the trade show's functionalnetwork. For more on Silicon Valley and Sun Microsystems in the early 1990s, see Saxenian,AnnaLee. 1994. Regional advantage: Culture and competition in Silicon Valley and Route 128.Cambridge, MA: Harvard University Press.

Interwoven through the avenues of research outlined above might be an effort to

examine the technical development of current networking technologies, (re)considering

their relationship to the range of experimental technologies, such as the multicasting (and

the MBONE) that was designed (at least in part) to address perceived bandwidth and

connectivity issues. Recall that numerous scholars have stated that technologies are shaped

by the strategies of social groups in power who then tend to create "technologies [that]

mirror our societies" (Bijker and Law 1992, 3), reproducing the assumptions and

preferences of the engineers who crafted them. This avenue of research around

technological change could also reveal the complex relationship between the adoption of

networked information technologies (so-called "technology transfer") and the growing

complexity of how we might understand agency and the relationship between technical

design and proximity to power in an age that puts forth the potential of each individual

over the capacities of a social or cultural group.

Yet, as much as further studies might interrogate the degree to which the tools of

networked information technologies (driven by the cybernetic logics of protocols) and their

companion market-oriented reforms 0 2 infuse contemporary development strategies, these

political and economic practices might be examined far closer to home. It is easy to note the

continuing role that figures such as Dan Lynch and Vint Cerf, as well as many descendants

of the military-industrial research world, continue to have on the Internet today. Yet it is

perhaps more compelling to consider how the "mobilizing visions" that spawned Lynch's

"interoperability" trade show, and the imperative to expand the organizational logics of

Internet protocols worldwide through market-oriented partnerships as well as through the

policies of deregulation and the democratic free market, continue to critically inform the

concerns expressed by figures like Carl Malamud.

In fact, lest this study appear to be merely an historical account of the relationship

between mobilizing utopias and the managerial demands of commercialization, Carl

Malamud's own career suggests that much of the same operational logic that drove Internet

commercialization in the early 1990s - the work of reconstituting society to conform to the

logics of network protocols - has not simply been a chapter in the history of the early

Internet but rather a utopian effort that is constantly underway. In 2009, Malamud

102 In her research on efforts in Peru to both modernize the government and prepare citizens for theglobal, information-based economy, Anita Chan (2008) has employed the term "neoliberal networks"to describe the integration of the regulatory logic of protocols in global capital.

843

discussed how he and a small group of dedicated open-government activists "liberated" the

U.S. federal courts' record database from the privately-managed paywalls,1 0 3 making it

what he believed it should be - free and widely accessible - by publishing millions of pages

of the cases on the Internet. Malamud also shared his principles of open data-that

includes data that is widely accessible, "machine processable," and available in a primary

and non-proprietary format (Malamud 2007).104 These changes, Malamud suggested, were

leading to the next "wave" of governance.

We are now witnessing a third wave of change-an Internet wave-where theunderpinnings and machinery of government are used not only bybureaucrats and civil servants, but by the people. (2009, 18)

This transformation, Malamud has suggested, results in "government as platform"

(O'Reilly, 2009), a term that conceives of government systems as the basis for private

enterprise as well as for the traditional tasks of governance. Malamud has further argued

that, in this view, the traditional tools of government become critical elements of the

architecture of the network itself.

Government information-patents, corporate filings, agriculture research,maps, weather, medical research-is the raw material of innovation, creating

a wealth of business opportunities that drive our economy forward.Government information is a form of infrastructure, no less important to ourmodern life than our roads, electrical grid, or water systems. (2009, 21-22)105

By proposing new expectations about the accessibility of government data, a subject that

has preoccupied him for nearly his entire career, Malamud promotes new channels of

connectivity between citizens and the state while at the same time advocating for the

"reformation" of traditional government structures to conform to the managerial logics of

protocols. More precisely, for Malamud the "mobilizing visions" that so engaged his

imagination in the 1990s continue to critically inform his work today. Governance and the

control of the production and distribution of knowledge, as it relates to the Internet, have

changed considerably in the intervening decades. Malamud's attention has turned away

from the construction of "big networks." Instead, his focus has turned toward the far more

103 This database is known as PACER, the government-run Public Access to Court ElectronicRecords. It is only accessible for a charge, is not searchable, and not user-friendly for the generalpublic.104 In 2007, Carl Malamud and Tim O'Reilly of O'Reilly Media held an invitation-only "OpenGovernment Working Group" to generate principles for open government data. The 8 principles areavailable here: http://resource.org/8 principles.html.105 Here, Malamud cites Alfred Chandler's Strategy and Structure (1962) as the defining work on"the intertwined nature of government, infrastructure, and industry."

87

intimate project of incorporating the logics of networks into individuals themselves. He

seeks to cultivate individuals who possess the capacity to self-govern, distributing the

responsibilities once assumed by modern states to citizens themselves. In this way,

Malamud is an example of the enduring impact of the Interop trade show and the politics of

Internet commercialization on individuals.

88

Appendix A: List of Interviewees

Auerbach, Karl (former ARPAnet engineer and key member of the Interop trade showINTEROPnet team. 2009. Interview with author in San Jose, California, June 30.

Brandin, David K. (former vice president and director of SRI International and vicepresident of programs at Interop). 2009. Phone interview with author, July 28.

Clark, David (senior research scientist at MIT Computer Science and Artificial IntelligenceLaboratory as well as chief protocol architect from 1981-1989 and chair of IAB). 2009.Phone interview with author, June 25.

Crocker, David (former ARPAnet engineer who contributed to the development ofinternetworking capabilities in the research and commercial sectors). 2009. Interview withthe author in Palo Alto, CA, June 25 and June 29.

Davenport, Glorianna (founding member of the MIT Media Lab and former director of theInteractive Cinema group). 2009. Interview with the author in Cambridge, Massachusetts,March-April.

Jacobsen, Ole (editor and publisher of The Internet Protocol Journal and long-time editorand publisher of Interop Company's ConneXions-The Interoperability Report). 2009.Interview with the author in San Francisco, July 2.

Lucas, Marty (directed audio and web production for the Internet 1996 Expo). 2009. Phoneinterview with author, July 7.

Lynch, Daniel (former computing manager at SRI, long-time member of the IAB [1983-19931, and founder of Interop Company). 2009. Interview with author, June 24.

Malamud, Carl (former document resource author, founder of the Internet MulticastingService and Public.Resource.Org). 2009. Interview with the author, May 13.

Rheingold, Howard (member of the Whole Earth network as well as author of VirtualCommunity: Homesteading on the Electronic Frontier and former executive editor ofHotWired, one of the first commercial content web sites). 2009. Interview with the author,June 23.

Rose, Marshall (network protocol and software engineer who contributed to thedevelopment of network management and distributed systems management and foundedDover Beach Consulting). 2009. Phone interview with author, June 27.

89

References

The Charles Babbage Institute, housed at the Center for the History ofInformation Technology at the University of Minnesota, hosts copies of Interop'smonthly "ConneXions - The Interoperability Report" publications. To accesstheir online archives, see http://www.cbi.umn.edu/. The Computer HistoryMuseum, in Mountain View, California, holds a partial record of publishedmaterials and other documents related to Interop.

Additional Interop documents were accessed from the Computer HistoryMuseum's archival collection.

The sources designated as RFC (Request for Comments) are technicaldocuments that define technical standards or network procedures. They can befound online at www.rfc-editor.org. Conversations about the technical history ofthe Internet can be found at the USC-Information Sciences Institute's PostelCenter [www.postel.org].

Abbate, Janet. 1999. Inventing the Internet. Cambridge, MA: MIT Press.

Agre, Phil. 1998. Yesterday's Tomorrow. Times Literary Supplement, July 3, 3-4.

Almquist, Phil. 1989. The INTEROP 88 Network-behind the scenes. ConneXions - TheInteroperability Report 3(2): 2. www.cbi.umn.edu/hostedpublications/Connexions/ConneXionsO3 1989/ConneXions3-02 Feb1989.pdf.

Bijker, Wiebe E. 1997. Introduction to Of bicycles, bakelites, and bulbs: toward a theory ofsociotechnical change, 1-19. Cambridge, MA: MIT Press.

-- -. 1997. King of the road: The social construction of the safety bicycle. Of bicycles,bakelites, and bulbs: Toward a theory of sociotechnical change, 19-101. Cambridge, MA:MIT Press.

Bijker, Wiebe E. and John Law. 1992. Introduction to Shaping technology / building society:Studies in sociotechnical change, ed. Wiebe E. Bijker and John Law, 1-21. Cambridge,MA: MIT Press.

Brandin, David R., and Michael A. Harrison. 1987. The Technology War: A Case ofCompetitiveness. New York: John Wiley & Sons.

Bukatman, Scott. 1991. There's always Tomorrowland: Disney and the hypercinematicexperience. October 57: 55-78.

Carlson, W. Bernard. 1992. Artifacts and frames of meaning: Thomas A. Edison, hismanagers, and the cultural construction of motion pictures," In ShapingTechnology/Building Society: Studies in Sociotechnical Change, ed. Wiebe E. Bijker andJohn Law, 175-198. Cambridge, MA: MIT Press.

Chan, A.S. 2008. The Promiscuity of Freedom: Development and Governance in the Age ofNeoliberal Networks. PhD dissertation, Massachusetts Institute of Technology.

90

Chun, Wendy Hui Kyong. 2006. Control and freedom: Power and paranoia in the age of

fiber optics. Cambridge, MA: MIT Press.

Clark, David D. 1992. A Cloudy Crystal Ball: Visions of the Future. Lecture slides fromplenary presentation at 24th meeting of the Internet Engineering Task Force, July 13-17, in Cambridge, Massachusetts. http://ietf20.isoc.org/videos/future ietf 92.pdf.

Deese, R. S. 2009. The artifact of nature: 'Spaceship Earth' and the dawn of globalenvironmentalism. Endeavour 33(2): 70-75.

DeNardis, Laura. 2009. Protocol politics: The globalization of Internet governance.Cambridge, MA: MIT Press.

Donath, Judith. 1997. Inhabiting the Virtual City. PhD dissertation, Massachusetts Institute ofTechnology. http://smg.media.mit.edu/people/judith/Thesis/Inhabitants.frame.html.

Edwards, Paul. N. 1996. Preface to The closed world: Computers and the politics ofdiscourse in cold war America, Cambridge, MA: MIT Press.

---. 1998. "We defend every place": Building the cold war world. The closed world:Computers and the politics of discourse in cold war America, Cambridge, MA: MITPress.

-- -. 1998. Y2K: Millennial reflections on computers as infrastructure. History andTechnology 15: 7-29.

-- -. 2004. "A vast machine": Standards as social technology. Science 304(5672): 827-828.

-- -. 2004. Infrastructure and modernity: Force, time, and social organization in thehistory of sociotechnical systems. Modernity and Technology, ed. Thomas J. Misa,Philip Brey, and Andrew Feenberg, 185-225. Cambridge, MA: MIT Press.http://www.si.umich.edu/-pne/PDF/infrastructure.pdf.

Fischer, Michael J. 1995. Worlding cyberspace: Toward a critical ethnography in time,space, and theory. In Critical anthropology now: Unexpected contexts, shiftingconstituencies, changing agendas, ed. George Marcus. 245-304. Santa Fe, New Mexico:School for American Research Press.

Flichy, Patrice. 2007. The Internet imaginaire. Cambridge, MA: MIT Press.

Galloway, Alexander R. 2004. Protocol: Control after decentralization. Cambridge, MA: MITPress.

Griffiths, Alison. 2002. Introduction to Wondrous difference: cinema, anthropology, & turn-of-the-century visual culture. New York: Columbia University Press.

Gunning. Tom. 2003. Re-Newing old technologies: Astonishment, second nature, and theuncanny in technology from the previous turn-of-the-century. In Rethinking MediaChange: The Aesthetics of Transition, ed. David Thorburn and Henry Jenkins, 39-59.Cambridge, MA: MIT Press.

Hafner, Katie, and Matthew Lyon. 1996. Where wizards stay up late: The origins of theInternet. New York: Simon & Schuster.

Harvey, David. 1990. Between space and time: reflections on the geographical imagination.Annals of the Association of American Geographers 80: 418-34.

91

Hughes, Thomas P. 1987. The Evolution of Large Technological Systems. In The SocialConstruction of Technological Systems, ed. Wiebe E. Bijker, Thomas P. Hughes, andTrevor Pinch, 51-82. Cambridge, MA: MIT Press.

Interop Company. 1987. Solutions Showcases Demonstrations Conference, SRI NetworkInformation Center Records, Lot X3578.2006, Interoperability Materials, ComputerHistory Museum, Mountain View, California.

Interop Company. 1991. News Release: Interop 91 Fall to Feature Major Announcement onStandards Availability via International Video Link. Mountain View, California:September 16. http://www.scribd.com/doc/2571592/INTEROP-91-Fall-to-Feature-Major-Announcement.

Jacobsen, Ole. 1988. INTEROP 88 report. ConneXions - The Interoperability Report 2: 7-9.www.cbi.umn.edu/hostedpublications/Connexions/ConneXions02 1988/ConneXions2-

11 Nov1988.pdf.

Jackson, Steven J., Paul N. Edwards, Geoffrey C. Bowker, and Cory P. Knobel. 2007.Understanding infrastructure: History, heuristics, and cyberinfrastructure policy. FirstMonday: June 12. http://firstmonday.org/htbin/cgiwrapi/bin/ois/index.php/fm/article/view/1904/1786.

Jenkins, Henry, and Fuller, Mary. 1995. Nintendo and New World Travel Writing. InCybersociety: Computer-Mediated Communication and Community, edited by Steve G.Jones, 57-72. Thousand Oaks, CA: Sage Publications.

Kehoe, Brendan P. 1992. Zen and the art of the Internet: A beginner's guide to the Internet.Englewood Cliffs, New Jersey: Prentice Hall. http://www.gutenberg.org/etext/34.

Kelly, Kevin. 1994. Out of control: The rise of neo-biological civilization. Reading, MA:Addison-Wesley. http://www.kk.org/outofcontrol/.

-- -. 1998. New rules for the new economy: 10 radical strategies for the connected world.New York: Penguin Group. http://www.kk.org/newrules/.

Kelty, Christopher M. 2008. Two Bits: The cultural significance of Free Software. Durham,NC: Duke University Press.

Krol, Ed. 1992. The whole Internet user's guide and catalog. Sebastopol, CA: O'ReillyMedia. http://www.archive.org/details/wholeinternet0Okrolmiss.

Larkin, Brian. 2008. Introduction and Conclusion to Signal and noise: Media,infrastructure, and urban culture in Nigeria. Durham, NC: Duke University Press.

Law, John and Michel Callon. 1992. The life and death of an aircraft: A network snalysis oftechnical change. In Shaping technology / building society: Studies in sociotechnicalchange, ed. Wiebe E. Bijker & John Law, 21-52. Cambridge, MA: MIT Press.

Lessig, Lawrence. 1999. Code and Other Laws of Cyberspace. New York: Basic Books.

Levy, Steven. 1994. Insanely great. Wired: (2)02. http://www.wired.com/wired/archive/2.02/macintosh pr.html.

Lynch, Daniel C., and Marshall T. Rose, ed. 1993. Internet System Handbook. Reading,MA: Addison-Wesley Publishing Company, Inc.

92

Malamud, Carl. 1991. ITU Adopts a New Meta-Standard: Open Access. ConneXions - TheInteroperability Report 5:21. http://www.cbi.umn.edu/hostedpublications/Connexions/ConneXions05 1991/ConneXions5-12 Dec1991.pdf.

-- -. 1993a. Exploring the Internet: A technical travelogue. Englewood Cliffs, NJ: Prentice-Hall.

-- -. 1993b. Cyberjockying in the 21st Century. Sunergy 7. Television program hosted byJohn Gage. Silicon Valley, CA. http://www.archive.org/details/sunergy 7.

-- -. 1995. A world's fair. Presentation at the semi-annual Networld + Interop, July 21, inTokyo, Japan. http://www.scribd.com/doc/2576764/A-Worlds-Fair.

- - -. 1997. A World's fair for the global village. Cambridge, MA: The MIT Press.

-- -. 2009. By the People. Presentation at the Gov2.0 Summit, September 10, inWashington, D.C. http://www.public.resource.org/people/.

Markoff, John. 1993. Building the Electronic Superhighway. New York Times. January 24.

Marvin, Carolyn. 1990. When old technologies were new: Thinking about electriccommunications in the late nineteenth century. Oxford: Oxford University Press.

Miller, Daniel, and Don Slater. 2001. The Internet: An ethnographic approach. Oxford, UK:Berg Publishers.

Negroponte, Nicholas. 1995. Being digital: A book (p)review. Wired Magazine: (3)02.

Nye, David E. 1992. The Electrifying Future. In Electrifying America: Social meanings of anew technology, 1880-1940, 339-380. Cambridge, MA: MIT Press.

- - -. 1994. American Technological Sublime. Cambridge, MA: MIT Press.

Pinch, Trevor, and Wiebe E. Bijker. 1987. The social construction of facts and artifacts: Orhow the sociology of science and the sociology of technology might benefit each other.In The social construction of technological systems, ed. Wiebe E. Bijker, Thomas P.Hughes, and Trevor Pinch, 17-50. Cambridge, MA: MIT Press.

Pitsker, Bo. 1993. Insights into the INTEROPnet. ConneXions - The InteroperabilityReport (7)3: 2-8.

Russell, Andrew L. 2006. 'Rough consensus and running code' and the Internet-OSIstandards war. IEEE Annals of the History of Computing July-September: 48-61.

Rydell, Robert W., and Rob Kroes. 2005. Buffalo Bill in Bologna: The Americanization ofthe world, 1869-1922. Chicago: University of Chicago Press.

Saxenian, AnnaLee. 1994. Regional advantage: Culture and competition in Silicon Valleyand Route 128. Cambridge, MA: Harvard University Press.

Star, Susan Leigh, and Karen Ruhleder. 1996. Steps toward an ecology of infrastructure:Design and access for large information spaces. Information Systems Research 7(1):111-134.

Stefik, Mark. 1996. Internet dreams: Archetypes, myths, and metaphors. Cambridge, MA:MIT Press.

Stephenson, Neal. 1996. Mother Earth mother board. Wired 4(12). http://www.wired.com/wired/archive/4.12/ffglass.html.

92

Streeter, Thomas. 1999. "That deep romantic chasm": Libertarianism, neoliberalism, andthe computer culture. In Communication, citizenship, and social policy: Re-thinking thelimits of the welfare state, ed. Andrew Calabrese and Jean-Claude Burgelman, 49-64,Lanham, Maryland: Rowman & Littlefield.

-- -. 2003. The romantic self and the politics of Internet commercialization. CulturalStudies 17(5): 648-668.

Sunshine, Carl. 1989. Computer network architectures and protocols. New York: SpringerPublishing.

Taylor, Charles. 2004. Modern social imaginaries. Dunham, N.C.: Duke University Press.

Thrift, Nigel J. 2005. Knowing capitalism. London: Sage Publications Ltd.

Turchi, Peter. 2008. Maps of the imagination: The writer as cartographer. San Antonio,Texas: Trinity University Press.

Turner, Fred. 2006. From counterculture to cyberculture: Stewart Brand, the WholeEarth network, and the rise of digital utopianism. Chicago: University of ChicagoPress.

Warf, Barney. 1998. Reach out and touch someone: AT&T's global operations in the 1990s.Professional Geographer 50(2): 255-267.

Williams, Rosalind H. 1982. The dream world of mass consumption. In Dream worlds: Massconsumption in late nineteenth century France, 56-106. Berkeley: University ofCalifornia Press.

Winseck, Dwayne R., and Robert M. Pike. 2007. Wireless, war, and networks. InCommunication and empire: Media, markets, and globalization, 1860-1930. Durham,Duke University Press.

Yates, JoAnne. 1993. Control through communication: The rise of system in Americanmanagement. Baltimore: The Johns Hopkins University Press.

Zittrain, Jonathan. 2008. The future of the Internet-and how to stop it. New Haven: YaleUniversity Press.

94