← Before: The Rise of the Internet, Part 1: Exponential Growth
In the summer of 1986, Senator Al Gore ( Albert Arnold Gore Jr. ) of Tennessee introduced an amendment to a Congressional Act to provide a budget for the US National Science Foundation (NSF). He called on the federal government to explore the possibilities of "communication networks to link supercomputers at universities and federal research centers." To explain the purpose of the bill Gore led a striking analogy:
One of the promising technologies is the development of fiber-optic systems for voice and data transmission. As a result, we will have a system consisting of many fiber optic systems scattered throughout the country. American highways transport people and materials around the country. Federal highways connect to state roads, which in turn connect to county and city roads. To transfer data and ideas, we will need a telecommunications highway connecting users from both coasts, from all states, from all cities. The study commissioned by this amendment will identify the challenges and opportunities that our country will face in creating such a highway.
In the years that followed, Gore and his allies would call for the creation of an "information superhighway," or, technically speaking, a national information infrastructure (NII). As Gore wanted, his analogy with the interstate highway system allows one to imagine something like a central switch connecting local and regional networks together, allowing all Americans to communicate with each other. On the other hand, the analogy is a little confusing - Gore did not suggest creating a data network that would be paid for and supported by the government. He imagined that the information superhighway would not look like its concrete-tarmac namesake. It will appear thanks to the efforts of market forces and within the framework of certain rules... The rules will guarantee free enterprise, free and open access to any service provider (later this concept will be called "net neutrality"). They should also provide subsidies or other mechanisms that open access even to the least fortunate sectors of society, in order to prevent a gap between information-rich and information-poor people from emerging.
Over the ensuing decade, Congress, responding to the increasing importance of computer networks for American researchers, education, and society in general, slowly developed rules for such a network. However, this slow process did not keep pace with the rapid growth of NSFNET, which was followed by the bureaucracy of executive directors attached to the network. Despite the sclerotic reputation of the bureaucracy, it was originally created for the ability to instantly and without hesitation react to what is happening - in contrast to the legislature. Therefore, from 1988 to 1993, the NSF developed the rules according to which the Internet would later become private, and then public. Each year, the foundation has had to deal with new demands and expectations from users of NSFNET and related networks. As a result, the foundation made decisions on the fly,and they quickly overtook congressional plans to manage the development of the information superhighway. And for the most part, all decisions depended on one person -Stephen Wolff .
Acceptable use
Wolf received his Ph.D. in electrical engineering from Princeton in 1961 (around the same time that Bob Kahn[Robert Elliot Kahn is the inventor of the TCP protocol; together with Vinton Cerf - the inventor of the IP protocol]). The beginning of his work could result in a comfortable academic career. He started out as a post-doc at Imperial College, after which he taught for several years at Johns Hopkins University. But then he suddenly changed direction and took a job at a ballistic research laboratory in Aberdeen, Maryland. He spent the entire 1970s and early 1980s there, researching communications and computer systems for the US Army. He introduced Unix to the lab's offices, and connected Aberdeen to the ARPANET. Unfortunately, his biographical data before joining the NSF is very little - I could not even find his date and place of birth.
In 1986, he was hired by NSF to work with the backbone for supercomputers. He was ideally suited for this position, given his experience connecting military supercomputers to the ARPANET. He became the chief architect of the evolution of NSFNET from the beginning until he left the foundation in 1994, when he went into the private sector and became a manager at Cisco Systems. The original job of the network that Wolf was hired to run was to connect researchers from all over the United States with NSF-funded supercomputing centers. However, as we saw last time, once Wolf and the other network managers realized the demand generated by the emerging highways, they quickly developed another NSFNET option: a communications network for all US researchers and academics.
However, Wolf did not want the government to be constantly involved in providing network services. From his point of view, NSF just needed to start the process by creating the initial demand that could generate a market for commercial networking services. After that, Wolf believed, the government should no longer compete with commercial enterprises. Therefore, he intended to remove NSF from the equation by privatizing the network and handing over the management of the backbone to private companies not financed from the budget - so that the market would take everything into its own hands.
This idea was consistent with the spirit of the time. In the West and in nearly all political regimes, government leaders of the 1980s advertised privatization and deregulationas the best ways to spur economic growth and innovation after the stagnation of the 1970s. One example from many: Around the same time that NSFNET began to develop, the FCC lifted several decades-old restrictions on broadcast corporations. In 1985, she removed the restriction on print and broadcast media ownership in the same locality. Two years later, the doctrine of impartiality, which required the media to present different currents of view on debates related to state policy, was canceled.
Through his position at the NSF, Wolf had several leverage to help him achieve his goal. The first was to interpret and enforce the acceptable use policy (AUP). In line with NSF's mission, the NSFNET backbone policy prior to June 1990 stipulated that the network would be used for "research and other academic purposes." The limitation was serious, preventing the network from being used commercially. However, Wolf was very flexible about the interpretation of this rule. Aren't there regular mailing lists for scientific research with new product lists from corporations selling data processing software? Or, for example, permission to connect the MCI postal service to the mainline at the request of Vinton Cerffrom government to MCI to manage the development of MCI Mail. Wolff attributed this move to research support by allowing researchers to connect with a broader range of people that they might need to work. Later, this rationalization justified connecting other commercial e-mail systems like CompuServe. Far-fetched? Maybe. However, Wolf believed that by allowing some commercial traffic on the same infrastructure used for NSF's public traffic, he would attract private investment, which would then continually support research and education.
Wolff's strategy to open the doors of NSFNET to commercial organizations, as far as possible, was supported in 1992 by Congressman Rick Boucher. He helped oversee the work of the NSF as Chairman of the Science Subcommittee. He also pushed for an NSF amendment that allowed NSFNET to be used in a way that "enhances the overall capabilities of networks to support research and education." It was a justification for Wolf’s approach to commercial traffic after the fact, and it allowed just about anything to do — as long as it generated profits that spurred investment in NSFNET and its connected networks.
Dual-purpose networks
Also, Wolf, taking care of the commercial development of the networks, supported the lease of the capacity of the regional networks for commercial traffic. The NSF backbone brought together a variety of non-profit regional networks, from NYSERNet in New York to Sesquinet in Texas and BARRNet in Northern California. NSF did not sponsor these networks directly, but it did support them indirectly, with money for laboratories and universities, which helped them reduce the cost of connecting to local networks. Some regional grids began to use the same infrastructure, supported from the budget, to create commercial organizations. They began to sell access to the network over the same wires that were used for scientific and research purposes with money from the foundation. And Wolf supported them in these endeavors,considering this as another way to accelerate the transition of the national research and educational infrastructure under private management.
This was in keeping with the political spirit of the 1980s, which encouraged entrepreneurs to capitalize on budget generosity in the hope that society would indirectly benefit from economic growth. You can see here parallels with the double use of regional networks under the Baye-Doyle Act of 1980, according to which patents obtained during publicly funded research were transferred to the ordering organizations, and not to the state.
The most prominent example of dual-use networks was PSINet, a commercial company founded in 1988 as Performance Systems International. It was created by NYSERNet founder William Schroeder and its vice president Martin Schofstall. Schofstall, a former BBN engineer and co-author of the Simple Network Management Protocol (SNMP), which manages devices on an IP network, was the main technical leader. The business was run by Schroeder, an ambitious Cornell University biologist and MBA who helped his institute open a supercomputer center and connect it to NSFNET. He was confident that NYSERNet should sell services not only to educational institutions, but also to commercial enterprises. When other board members disagreed with him, he quit and started his own company. He first bought services from NYSERNet,and then he made enough money to organize his own network. PSINet was one of the first commercial ISPs, while continuing to provide non-profit services to colleges and universities wishing to connect to the NSFNET backbone.
In the late 90s, Schroeder and PSINet made good progress in the Internet bubble, aggressively buying up other businesses. The most extravagant purchase of the company was the right to change the name of the stadium for the American football team “Baltimore Ravens”. Schroeder tried his luck when he published in 1997 an article “Why the Internet will never collapse”. Unfortunately for him, this happened, leading to his dismissal from the company in 2001 and the bankruptcy of PSINet the following year.
Wolf's last leverage to encourage the development of the commercial Internet was his role as contract manager for the Merit-IBM-MCI consortium that operated the backbone. However, initially, it was not Wolf who expressed the desire to change something in this area, but the network operators themselves.
Commercial highway
MCI and her colleagues in the telecommunications industry have had strong incentives to seek or create demand for the exchange of computer data. In the 1980s, they were busy upgrading their long-distance networks, moving from coaxial cable and microwaves - which were already higher throughput than old copper wires - to fiber optic cables. These cables, which carried laser beams along glass fibers, had enormous bandwidth, no longer limited by the cable itself, but by the technology of the transmitters and receivers at both ends. And this ability was far from saturation. By the early 1990s, many companies had deployed OC-48 equipmentwith a bandwidth of 2.5 Gbit / s - a decade ago it was impossible to imagine anything like that. As a result, the explosive growth in traffic should have attracted new revenues with little to no overhead - which meant near-net profit.
To better imagine the rate at which bandwidth costs are falling: In the mid-1980s, renting a T1 line from New York to Los Angeles would have cost $ 60,000 a month. 20 years later, the OC-3 channel with a bandwidth of 100 times more cost $ 5,000 - the unit cost of the channel dropped more than a thousand times.
The desire to gain the knowledge needed for the growing data communications market prompted MCI to join the NSFNET proposal proposed by Merit. The latter arranged a serious dumping (offering $ 14 million for five years compared to the proposals of competitors - $ 40 and $ 25 million), which would undoubtedly result in short-term financial losses for MCI and IBM. However, they hoped to recoup investments and start making money by 1989. The existing backbone network was approaching maximum throughput, transmitting 500 million packets a month - each year their number grew by 500%. So when NSF asked Merit to upgrade its backbone from 1.5 Mbps T1 links to 45 Mbps T3 links, they took the opportunity to offer Wolf a new contract.
T3 was a new frontier for networks - previously digital providers did not have such equipment for digital networks and experience with it. Therefore, companies insisted on the need to increase private investment. This would require a restructuring, and would allow IBM and Merit to use the new infrastructure for transporting commercial traffic, making it a dual-use backbone. To this end, the consortium was to form a new non-profit company, Advanced Network & Services, Inc. (ANS), which will provide NSF T3 network services. Its subsidiary, ANS CO + RE, will sell the same services to any paying clients. Wolf agreed to this proposal, considering it another step towards the transfer of the network to commercial management. Moreover, he feared that if the commercial use of the highway continued to be blocked,eventually the providers will decide to split from NSFNET and create their own commercial Internet.
Until that moment, Wolf's plan to gradually withdraw NSF from control of the network had no specific schedule or planned milestones. A working meeting on this issue took place at Harvard in March 1990. There, Wolff and many other early Internet leaders considered many options without making specific plans. It was the ANS stratagem that set off the cascade of events that led directly to the complete privatization and commercialization of NSFNET.
It all started with a negative reaction to events. Despite Wolff's good intentions, IBM and MCI's ANS maneuver caused irritation among networkers. And the problems arose because of the very commercial networks connected to the backbone that Wolf promoted. Until that moment, they got along at the very least, because they worked on the same conditions, as equal. However, after the emergence of the commercial company ANS, it took de facto a monopoly on the backbone located at the center of the Internet. Moreover, despite Wolf's attempts to interpret the network acceptable use policy loosely, the ANS chose to interpret it quite strictly. It refused to link the non-commercial portions of the backbone (dedicated to NSF traffic) to commercial networks such as PSI, as this would mix non-commercial and commercial traffic.When this caused discontent, the company backtracked and adopted a new policy that allowed communication with a commission based on traffic volume.
PSINet didn't like it. In the summer of 1991, they merged with two other providers - UUNET, which began selling commercial access to Usenet even before Internet access, and CERFNet (California Federal Education and Research Network) operated by General Atomics. The trinity formed their own traffic exchange point, bypassing the ANS highway. The Commercial Internet Exchange (CIX) initially consisted of a single routing center in Washington DC, passing traffic between three networks. They agreed to share traffic for free, regardless of its volume, so that each of the networks pays for the operation of the center equally. Then new centers opened, in Chicago and Silicon Valley, and other networks joined in trying to avoid paying ANS fees.
Exposure
Rick Boucher, a congressman we know who supports the commercialization of the NSF, nevertheless demanded an investigation into the legitimacy of the ANS-related actions. An investigation ruled that the NSF's actions could be considered hasty, but not malicious or corrupt. However, Wolf decided it was time to relinquish control of the highway. With the advent of ANS + CORE and CIX, privatization and commercialization went full blast, but the process developed in such a way that it threatened to split the Internet into many unrelated fragments, as CIX and ANS refused to communicate. In the summer of 1992, NSF drafted a plan for a new, privatized network architecture, published it for public comment, and completed it in May 1993. In the spring of 1995, it was planned to close NSFNET, and transfer its assets to IBM and MCI.Regional networks were to continue to operate, contracting with private providers for network access, and financial support from NSF would gradually diminish over four years.
But what could replace the backbone in a world with many competing ISPs? What mechanism will bring these competing private interests together into a coherent whole? Wolf's answer was inspired by existing traffic exchange centers set up by cooperatives like the CIX. NSF will order the creation of four network access points (NAP), traffic exchange centers that can be used by various providers. Four separate contracts will help avoid the problems encountered with ANS and prevent the formation of a monopoly of ownership of these centers. One NAP will be located at the existing Metropolitan Area Ethernet East (MAE-East) center in Vienna, Virginia, operated by Metropolitan Fiber Systems (MFS). MAE-West, operated by Pacific Bell, opened in San Jose, California.Another NAP was placed under Sprint in Pensoken, New Jersey, and the last was taken over by Ameritech in Chicago. The transition went smoothly and the NSF shut down its mainline on schedule on April 30, 1995.
The most difficult part of privatization was not at all the issues with equipment and network infrastructure. Problems arose with the transfer of control of the Domain Name System (DNS). For most of her history, she was ruled by a single person - John Postel. However, this situation did not suit businesses investing millions in the commercial Internet. Therefore, the government turned over the management of the domain name system to a contractor, Network Solutions. The NSF did not have a real mechanism to oversee DNS activities (perhaps it should have divided control of different top-level domain names among different contractors), and Congress failed to intervene in time to create any kind of regulatory regime. Once again control passed from hand to hand in 1998, and ended up in the non-profit organization ICANN (Internet Corporation for Assigned Names and Numbers - "Corporation for the management of domain names and IP addresses"), however, DNS management is still controversial.
Separation of monopoly
Gore and others have often used the phrase "information superhighway" as a metaphor for digital networks. But Congress has never seriously considered the idea of using the interstate highway system as a model for designing a regulatory policy network. The US government paid for the construction and maintenance of interstate highways to provide a reliable transportation network for the entire country. But in an era where deregulation and privatization were seen as a good idea by the mainstream parties, a French-style government network and information services system like Transpac and Minitel was simply not considered.
The only minor exception to the rule was Senate Daniel Inouye's proposal to reserve 20% of Internet traffic for public use. No wonder it didn't pass.
The long-distance market, which emerged from the split of the Bell System from 1982 to 1984, was the most attractive political model for the future-planning Congress. In 1974, the US Department of Justice sued AT&T for its first lawsuit against the organization since the 1950s. In the lawsuit, the company was accused of non-competitive behavior and violation of the Sherman antitrust law... Specifically, the company was accused of using its market opportunities to prevent new innovative enterprises from entering the market. Among them are operators of mobile radio communications, digital networks, satellite communications, manufacturers of special terminals, etc. The lawsuit was certainly based on disputes dating back to the 1960s between AT&T and companies such as MCI and Carterfone (these disputes were described in the previous articles of the series).
When it became clear that the Department of Justice was not joking and was going to crush AT&T, the company first appealed to the US Congress. John de Butts, chairman and director of the company since 1972, tried to push through Congress the Bell Bill, formally called the Consumer Communications Reform Act. He was supposed to legally support AT&T's claims that the benefits of a single universal telephone network outweigh the risks of monopoly abuse, which the FCC would anyway suppress. However, the proposal met with strong opposition from the Communications Subcommittee, and never got to a vote.
In 1979, the board of directors changed tactics, replacing the belligerent de Butts (he once told telecom policymakers that he was opposed to competition and supported monopolization) with the more accommodating Charles Brown. By then, however, it was too late to try to stop the growing antitrust case. It became more and more obvious to the directors that they could not win. Therefore, in January 1982, Brown agreed to take the Bell System apart under the supervision of Judge Harold Green.
The various Bell companies that hauled the cable to customers' homes, and usually operated within state lines (New Jersey Bell, Indiana Bell, and so on), were cut into seven pieces called Bell Operating Companies (RBOC). Clockwise, these were: NYNEX in the northeast, Bell Atlantic, Bell South, Southwestern Bell, Pacific Telesis, US West, and Ameritech. Each of them still owned a monopoly in their area, but they were prohibited from entering other telecommunications markets.
AT & T itself remained with long distance communication lines. Unlike local telephone operators, the agreement opened the market to free competition from any players willing and able to pay commissions for the transfer of calls to and from RBOC. An Indiana, for example, always had an Ameritech local telephone company, but could contract long distance services with anyone.
However, splitting the local and long-distance market meant eliminating the subsidies that AT&T was redirecting to village phone users. She took less money from them, making up for rich users using long distance. The sudden spike in the cost of telephony in rural areas across the country could not be tolerated, so the deal continued. A new, non-profit organization, National Exchange Carrier Association, was founded to collect commission from long distance carriers and distribute funds between RBOCS.
The new structure is working. In the 1980s, two big competitors, MCI and Sprint, entered the market and dug deep into AT&T's market share. Long-distance prices began to fall quickly. It is difficult to say what role the emergence of competition played in this, and what role was played by the emergence of high-bandwidth fiber-optic networks. However, the agreement is generally considered a highly successful example of regulatory relaxation and an argument for the market's ability to modernize stagnant industries.
This market structure, although created by court order solely for a specific case, but proved to be apparently successful, became the template by which Congress resolved the telecommunications policy of the Internet era in the mid-1990s.
Second time out of luck
But before the main event, another, smaller one happened. The High Speed Computing Act of 1991 was tactically important, but not strategically important. He did not promote any broad initiatives. Basically, he provided additional funding and support to Congress for the process that Wolf and the NSF were already engaged in and planned to continue. This included providing networking services to the research community, subsidizing scientific organizations to connect to NSFNET, and continuing to upgrade the backbone infrastructure.
And then in January 1995, the 104th Congress took office. Republicans gained control of the Senate and House of Representatives in the US Congress - for the first time in 50 years. And on their agenda was the fight against crime, tax cuts, government cuts and reforms, and support for righteousness and virtue. Gore and his allies have long touted universal access to the web as a key component of the nation's information infrastructure. But after these changes in power, the chances of a reform introducing universal telecommunications access services have dropped to zero. Instead, encouraging competition in telecommunications and Internet access, with a touch of censorship, was taken as the main course.
Market conditions looked promising. Around 1992, the major players in the telecommunications industry were already full. The traditional phone industry was dominated by seven RBOCs, GTEs and three large long distance companies - AT&T, MCI and Sprint, as well as many smaller ones. Newcomers included Internet access providers UUNET, PSINET, and the IBM / MCI backbone-born ANS company. Other companies, such as Metropolitan Fiber Systems (MFS), have also tried to pull their fiber networks. BBN, a contractor for ARPANET, began building a small internet empire of its own, buying up some of the regional networks in orbit around NSFNET - Nearnet in New England, BARRNet in San Francisco Bay, SURANet in the southeastern United States.
The main goal of the 1996 Telecommunications Act was to maintain and expand the competitive landscape. This was the first major change to the communications law since 1934. It was supposed to tweak the telecommunications law for the digital age. In the original law, operators were divided according to the physical methods of data transmission - telephony, radio broadcasting and television, cable TV. Each of them existed in their own sandbox, with their own rules, and usually they were forbidden to interfere with each other's business. As we've seen, sometimes separate sandboxes were created inside the sandboxes to separate the long distance and local phone markets. As all of these media were reduced to the same bits of data, carried over the same fiber optic links, satellites, or Ethernet cables, this separation made less and less sense.
The goal of the 1996 law, supported by both Republicans and Democrats, was to break down these barriers, these “regulatory Berlin walls,” as Gore called them in the law's summary. It is simply impossible to give the full content of this tome here, but several typical examples can be mentioned. Among other things, it includes:
- Competition with RBOC in the long distance telephone markets was allowed;
- The restrictions that prohibited one company from providing both cable and broadcast services were lifted;
- The rules that prevented concentration of radio station ownership were abolished.
However, by removing all the rules, opening gateways and allowing any participant to enter any market, regulators risked simply re-creating AT&T on a larger scale - a monopolistic mega-corporation that would dominate all communications and strangle all competitors. Most of the problems were expected from the control over the so-called. "Last mile" - communication lines from the local switch to the client's home or office. Establishing a long-distance network connecting major US cities was expensive, but not too expensive. Several companies have dealt with this for several decades, from Sprint to UUNET. However, repeating the pulling of all cables to every house in at least one city was already a task of a completely different level. From the earliest days of wild telephony, when tangled coils of iron wire crossed the city streets,competition in the field of wire communication at the regional level was almost non-existent. As for the Internet, the problem was especially acute in the case of high-speed data transmission directly to the destination - what later became known as broadband access. Competition between dial-up Internet providers flourished for many years - after all, all a user needed to gain access to the provider's computer was a telephone line. However, in this case it would no longer work - the new services will not use telephone networks.Competition between dial-up Internet providers flourished for years - after all, all a user needed to gain access to the provider's computer was a telephone line. However, in this case it would no longer work - the new services will not use telephone networks.Competition between dial-up Internet providers flourished for many years - after all, all a user needed to gain access to the provider's computer was a telephone line. However, in this case it would no longer work - the new services will not use telephone networks.
Legally, this problem was solved through the creation of the concept of CLEC - a competitive local exchange operator. Now, RBOC members, renamed ILEC (acting local exchange operators), were granted unlimited access to the long-distance market only after they allowed CLEC to use their infrastructure. The latter were engaged in the delivery of telecommunications services to homes and offices. This allowed competing providers of the Internet and other new services to maintain access to local networks even after the dial-up service disappeared. It was like a new type of telephone network for broadband access. In this model, CLECs played the same role as long-distance providers in the telephone market after the monopoly split.With the ability to freely connect to existing local telephone networks for a modest fee, they created a competitive environment in a market previously dominated by the problem of natural monopoly.
In addition to the creation of the CLEC, another important piece of law affecting the Internet had more to do with the virtuous attitude of the Republicans than with economic considerations. Chapter V, known as the "Communication Decency Law", prohibited the transmission of indecent or offensive content - describing or depicting "acts of sex or excretion, or related organs" - to any part of the Internet accessible to minors. It was an extension of the rules regarding the prohibition of broadcasting obscene content, carried over to the world of interactive computer services.
To what extent has this large-scale act coped with its tasks? By most standards, almost failed. The easiest thing was to reject the law on decency in communications, which was repealed in 1997 as violating the First Amendment [to the US Constitution, which guarantees, among other things, freedom of speech / approx. transl.]. However, several parts of Chapter V have gone through revision - including section 230, the most important for the future of the Internet. It allowed websites hosting user-generated content to survive without fear of constant lawsuits, and protected the continuity of everything from giants like Facebook and Twitter to small amateur forums.
Attempts to encourage local competition have lasted longer, but have achieved no more than controlling obscene content. What happened to the CLEC, which was given access to infrastructure so that they could compete on price and service mix? The law required the Federal Communications Commission in its rule-making to describe in detail what specific network separation was to be proposed. Incumbent companies required the courts to overturn any decisions that would open their networks to competitors. They have repeatedly won cases against the Federal Communications Commission, while also threatening that the emergence of competitors will slow down their plans to deliver fiber to consumer homes.
In 2001, with the arrival of the Bush administration and new chairman of the communications commission, Michael Powell, the commission began to actively oppose the original goals of the telecommunications act. Powell believed the need for alternative broadband access would be met by competition between different types of communications — cable, telephone, cellular, wireless, and power-line networks. They were not going to accept any new rules in favor of CLEC. At some point in 2000, it was possible to subscribe to high-speed Internet access using the infrastructure of a local telephone or cable provider. But after that, most of the main messages of the telecommunications act aimed at promoting competition simply stopped working.The well-publicized project for pulling fiber to every house actually connected a significant number of users after 2010. As historian Fred Goldstein wrote, existing providers “covered themselves with a fig leaf of competition without losing significant market shares” [Goldstein, The Great Telecom Meltdown, 145].
For most of the 20th century, networked enterprises in the United States first grew rapidly, fueled by entrepreneurial energy, and when they became important enough in terms of public interest, they were crammed into a matrix of regulatory platforms. Broadcasting and cable television developed along this pattern. Cargo transportation by road and air traffic. However, when the CLEC practically died in the early 2000s, the law on decency in communications was withdrawn, and other attempts to control the Internet, such as the Clipper law [it was proposed to install a special chip in telecommunications equipment that would allow the US government to bypass software encryption of data], stalled , the internet went the opposite way.
By nurturing the Internet under the strict guidance of the state, it was allowed to develop with almost complete non-interference. Opened by the NSF, NAPs were the latest major government intervention in the Internet. It provided both the transport layer — for transporting raw data on networks like Verizon and AT&T — and the application layer. The latter was used by software service providers from portals such as Yahoo! and search sites like Google to online stores like Amazon. In the final chapter, we look at the implications of this development pattern and briefly describe the evolution of the Internet in the United States from the mid-1990s to the present day.
What else to read
- Janet Abatte, Inventing the Internet (1999)
- Karen D. Fraser “NSFNET: A Partnership for High-Speed Networking, Final Report” (1996)
- Shane Greenstein, How the Internet Became Commercial (2015)
- Yasha Levine, Surveillance Valley: The Secret Military History of the Internet (2018)
- Rajiv Shah and Jay P. Kesan, “The Privatization of the Internet's Backbone Network,” Journal of Broadcasting & Electronic Media (2007)