<< Before this:
In 1990 , a network technology consultant and UNIX expert, published a comprehensive review of the state of computer networks at that time. In a small section dedicated to the future of computing systems, he predicted the emergence of a single global network for "email, conferencing, file transfer, remote login â just like the global telephone network and global mail exist today." However, he did not assign a significant role to the internet. He suggested that this global network "will most likely be managed by government communications agencies," except for the United States, "where it will be managed by the regional subsidiaries of Bell Operating Companies and long-distance carriers."
The aim of this article will be to explain how the internet, with its sudden explosive exponential growth, has so thoroughly refuted completely natural assumptions.
Handing Off the Baton
The first critical event that led to the emergence of the modern internet occurred in the early 1980s when the Defense Communication Agency (DCA) [now DISA] decided to split ARPANET into two parts. DCA took control of the network in 1975. By that time, it was clear that for the Information Processing Techniques Office (IPTO) of ARPA â the organization involved in exploring theoretical ideas â it made no sense to participate in the development of a network being used not for research communications but for everyday communication. ARPA unsuccessfully attempted to hand over control of the network to the private company AT&T. DCA, responsible for military communication systems, seemed to be the best second option.
In the first few years of this new situation, ARPANET flourished in a state of benign neglect. However, by the early 1980s, the aging communications infrastructure of the Department of Defense sorely needed updating. The proposed replacement project, AUTODIN II, whose contractor was Western Union, appeared to be failing. So, the heads of DCA appointed Colonel Heidi Hayden to be responsible for choosing an alternative. He proposed to use packet-switching technology, which was already available to DCA in the form of ARPANET, as the basis for the new defense data transmission network.
However, there was an obvious issue with transferring military data via ARPANET â the network was filled with long-haired scientists, some of whom actively opposed computer security or secrecy â for example, with his hacker colleagues from the MIT Artificial Intelligence Laboratory. Hayden suggested splitting the network into two parts. He decided to keep the research scientists funded by ARPA on ARPANET and to separate the computers working at defense contractors into a new network called MILNET. This kind of mitosis had two important consequences. Firstly, the division of military and non-military parts of the network was the first step toward transferring the internet to civilian and, subsequently, private control. Secondly, it proved the viability of fruitful internet technology â the TCP/IP protocols, which had been conceived about five years earlier. The DCA needed all ARPANET nodes to transition from the old protocols to TCP/IP support by the beginning of 1983. At that time, few networks used TCP/IP, but afterward, this process united the two proto-internet networks, allowing message traffic to connect research and military enterprises when necessary. To ensure the long-term viability of the TCP/IP protocol in military networks, Hayden established a fund of $20 million to support computer manufacturers who would write software to implement TCP/IP in their systems.
The first step in the gradual transition of the internet from military to private control also gives us a decent opportunity to say goodbye to ARPA and IPTO. Its funding and influence, which were under the management of Joseph Carl Robnett Licklider, Ivan Sutherland, and Robert Taylor, directly and indirectly led to the emergence of all early developments in interactive computing and computer networks. However, by establishing the TCP/IP standard in the mid-1970s, it played a crucial role in the history of computers for the last time.
The next major computing project organized by DARPA will be the autonomous vehicle competition of 2004-2005. The most famous project prior to this will be the strategic computer initiative based on AI from the 1980s, which cost a billion dollars and generated several useful military applications but had virtually no impact on civil society.
A decisive catalyst in the organization's loss of influence was . Most research scientists believed they were fighting for a just cause and defending democracy when Cold War-era research was funded by the military. However, those who grew up in the 1950s and 1960s lost faith in the military and its objectives after the latter became mired in the quagmire of the Vietnam War. Among the first was Taylor himself, who left IPTO in 1969, taking his ideas and connections to Xerox PARC. The Democrat-controlled Congress, concerned about the detrimental influence of military money on fundamental scientific research, passed amendments stipulating that defense funds should be spent solely on military research. ARPA reflected this shift in the funding culture in 1972 by renaming itself DARPA â .
. Thus, the baton passed to the civilian (NSF). By 1980, with a budget of $20 million, NSF was responsible for funding about half of the federal computer research programs in the U.S. And a large portion of these funds would soon be directed toward a new national computer network, .
NSFNET
In the early 1980s, Larry Smarr, a physicist from the University of Illinois, visited the Max Planck Institute in Munich, where the "Cray" supercomputer was available to European researchers. Frustrated by the lack of similar resources for scientists in the United States, he proposed that the NSF fund the creation of several supercomputing centers across the country. The organization responded to Smarr's claims and those of other researchers with similar complaints by establishing a division for advanced scientific computing in 1984, which led to the funding of five such centers with a five-year budget of $42 million. They stretched from Cornell University in the northeast to San Diego in the southwest. The University of Illinois, where Smarr worked, received its own center, the National Center for Supercomputing Applications (NCSA).
However, the centers' ability to improve access to computing power was limited. For users not living near one of the five centers, utilizing their computers would be difficult and would require funding for scientific trips lasting a semester or summer. Therefore, the NSF decided to build a computer network as well. History repeated itself â Taylor promoted the creation of ARPANET in the late 1960s precisely to give the research community access to powerful computing resources. The NSF would provide a backbone that would connect key supercomputing centers across the continent and then link regional networks, allowing access to these centers for other universities and research laboratories. The NSF would take advantage of internet protocols that Hayden promoted, delegating the responsibility for creating local networks to local scientific communities.
Initially, the NSF assigned the tasks of creating and maintaining the NCSA network from the University of Illinois as the source of the original proposal for a national supercomputing program. In turn, the NCSA leased the same 56 Kbps lines used by ARPANET since 1969 and launched the network in 1986. However, these lines quickly became congested with traffic (details of this process can be found in David Mills' work.Once again, the story of ARPANET repeated itself â it quickly became clear that the primary purpose of the network should not be scientists' access to computing power, but communication between people who had access to it. We can forgive the authors of ARPANET for not knowing that such a thing could happen â but how could the same mistake occur almost twenty years later? One possible explanation is that it is much easier to justify a seven-figure grant for using computing resources that cost eight figures than to justify spending such sums on seemingly trivial goals like the ability to exchange emails. It cannot be said that NSF was deliberately misleading anyone. But as the anthropic principle states that the physical constants of the universe are what they are, because otherwise we simply wouldn't exist and couldn't observe them, I wouldn't have to write about a government-funded computer network if there weren't such somewhat fictitious justifications for its existence.
Convincing itself that the network itself was at least as valuable as the supercomputers justifying its existence, NSF turned to external help to update the backbone of the network by installing T1 (1.5 Mbps) lines. The T1 standard was established by AT&T in the 1960s and was designed to handle up to 24 phone calls, each coded into a digital stream of 64 kbps.
The contract was awarded to Merit Network, Inc. in partnership with MCI and IBM, and over the first five years, it received a $58 million grant from NSF for the construction and maintenance of the network. MCI provided the communication infrastructure, IBM supplied the computing power and software for the routers. The non-profit company Merit, which managed the computer network connecting the campuses of the University of Michigan, brought with it experience in supporting a scientific computer network and gave the entire partnership a university spirit, which made it easier for NSF and the scientists using NSFNET to relate to it. However, the transfer of maintenance from NCSA to Merit was an obvious first step towards privatization.
Initially, MERIT stood for Michigan Educational Research Information Triad [Michigan Educational Research Information Triad]. The state of Michigan contributed $5 million to help this home network expand on T1.
Traffic flowed through the Merit backbone from over ten regional networks, from New York's NYSERNet research and educational network linked to Cornell University in Ithaca, to California's CERFNet federated research and educational network connected in San Diego. Each of these regional networks connected to countless local campus networks, as hundreds of Unix machines operated in college labs and faculty offices. This federal network of networks became the seed crystal of the modern internet. ARPANET only connected researchers in computing with substantial funding working at elite academic institutions. By 1990, almost any university student or faculty member could go online. By routing packets from node to nodeâfirst over local Ethernet, then into regional networks, and finally over long distances at the speed of light via the NSFNET backboneâthey could exchange emails or engage in polite discussions on Usenet with colleagues from the other side of the country.
After many more scientific organizations gained access through NSFNET than through ARPANET, in 1990, the DCA decommissioned the outdated network and completely removed the Department of Defense from the development of civilian networks.
Rise
During this period, the number of computers connected to NSFNET and its associated networksâwhich we can now collectively call the internetâgrew each year by about double. 28,000 in December 1987, 56,000 in October 1988, 159,000 in October 1989, and so on. This trend continued until the mid-1990s, after which growth . How, one wonders, considering this trend, could Quaterman not have noticed that the internet was destined to rule the world? If the recent pandemic has taught us anything, itâs that itâs very difficult for people to envision exponential growth, as it doesnât correspond to anything we encounter in everyday life.
Of course, the name and concept of the internet emerged even before NSFNET. The internet protocol was invented in 1974, and networks communicating via IP existed before NSFNET. We already mentioned ARPANET and MILNET. However, I could not find any references to 'the internet' â a single, worldwide network of networks â until the emergence of the three-tier NSFNET.
The number of networks within the internet grew at a similar pace â from 170 in July 1988 to 3,500 in the fall of 1991. Since the scientific community knows no borders, many of them were located abroad, starting with connections to France and Canada established in 1988. By 1995, it was possible to access the internet from nearly 100 countries, from Algeria to Vietnam. While it is much easier to count machines and networks than to count actual users, reasonable estimates suggest that by the end of 1994, there were 10-20 million users. In the absence of detailed data on who used the internet, for what purpose, and at what time, it is quite challenging to substantiate any historical explanation for such an incredible growth. A small collection of stories and anecdotes is unlikely to explain how, from January 1991 to January 1992, 350,000 computers connected to the internet, and the following year saw an additional 600,000, with another 1.1 million the year after that.
However, I will venture onto this epistemically shaky ground and assert that the three overlapping waves of users responsible for the explosive growth of the internet, each with their own reasons for connecting, were driven by an inexorable logic , which states that the value (and thus the attractiveness) of a network increases as the square of the number of its participants.
Initially, scientists arrived. The NSF intentionally distributed computing across as many universities as possible. After that, every scientist wanted to join the project because everyone else was already there. If the emails hadn't reached you, or if you didn't see or participate in the latest discussions on Usenet, you risked missing the announcement of an important conference, the chance to find a mentor, or not noticing cutting-edge research before its publication, and so on. Feeling pressured to join online scientific discussions, universities quickly connected to regional networks that could link them to the NSFNET backbone. For example, NEARNET, covering six states in the New England region, had over 200 participants by the early 1990s.
At the same time, access began to seep from faculty and graduate students into a much larger community of students. By 1993, about 70% of Harvard freshmen had created their email addresses. By that time, the internet had physically reached every corner of Harvard and its affiliated institutions. to run Ethernet not just into every building on campus but also into all student dormitories. Surely, it was only a matter of time before a student stumbled into their room after a wild night, plopped down in a chair, and struggled to type out an email they would regret sending the next morning - whether it was a confession of love or a fierce rebuttal to an enemy.
In the next wave, around 1990, commercial users began to arrive. That year, 1,151 .com domains were registered. The first commercial participants were the research departments of tech companies (Bell Labs, Xerox, IBM, etc.). They were essentially using the network for scientific purposes. Business communications for their executives were handled on other networks. However, by 1994, over 60,000 names in the .com domain, and making money on the internet truly began.
By the end of the 1980s, computers began to become a part of everyday work and home life for citizens of the United States, and the importance of having a digital presence for any serious business became evident. Email provided a way for easy and incredibly fast communication with colleagues, clients, and suppliers. Mailing lists and Usenet offered not only new ways to stay informed about professional developments but also new forms of very inexpensive advertising for a wide range of users. The internet provided access to a vast variety of free databasesâlegal, medical, financial, and political. Recently hired graduates, living in connected dorms, embraced the internet just as much as their employers did. It offered access to a much larger user base than any individual commercial service (once again invoking Metcalfe's Law). After paying for monthly internet access, practically everything else was available for free, unlike the significant costs for usage hours or sent messages that services like CompuServe and others charged. Among the early internet market participants were companies mailing softwareâsuch as The Corner Store from Litchfield, Connecticut, which advertised in Usenet groups, and The Online Bookstore, an e-book store founded by a former editor at Little, Brown and Company, more than a decade ahead of Kindle.
Then came the third wave of growth, bringing regular consumers who began to flock to the internet in large numbers in the mid-1990s. By this time, Metcalfe's Law was already operating at full throttle. Increasingly, 'being online' meant 'being on the internet.' Consumers could not afford to bring dedicated T1 lines home, so they almost always accessed the internet via We have already encountered part of this story, where commercial BBS gradually transformed into internet service providers. This change benefited both users (whose digital pool suddenly expanded to an ocean) and the BBS themselves, which transitioned to a much simpler business model as intermediaries between the telephone system and the 'leaving for the highway' of the internet with T1 bandwidth, without needing to maintain their own services.
Larger online services evolved in the same way. By 1993, all national-scale services in the U.S. â Prodigy, CompuServe, GEnie, and the young company America Online (AOL) â offered users, totaling 3.5 million, the ability to send emails to internet addresses. Only the lagging Delphi (with 100,000 subscribers) provided full internet access. However, in the following years, the value of internet access, which continued to grow at an exponential rate, quickly outweighed the access to proprietary forums, games, stores, and other content from the commercial services themselves. 1996 became a turning point â by October, 73% of online users were utilizing the WWW, compared to 21% a year prior. A new term, 'portal', was coined to describe the rudimentary remnants of services provided by AOL, Prodigy, and other companies, for which people paid simply to gain internet access.
Secret ingredient
So, we have a rough idea of how the internet grew at such explosive rates, but we don't fully understand why this occurred. Why did it become so dominant despite the multitude of other services trying to thrive in the preceding ?
Of course, government subsidies played their part. In addition to funding backbone communication lines, when the NSF decided to seriously invest in network development regardless of its supercomputer initiative, it didnât hold back. The conceptual leaders of the NSFNET program, Steve Wolf and Jane Kaveness, aimed to build not just a network of supercomputers, but a new information infrastructure for American colleges and universities. They established the Connections program, which took on part of the costs for connecting universities to the network in exchange for ensuring access to as many people as possible on their campuses. This accelerated the spread of the internet both directly and indirectly. Indirectly, as many regional networks spawned commercial enterprises that utilized the same infrastructure built on subsidies to sell internet access to commercial organizations.
But Minitel had subsidies too. However, what set the internet apart most was its layered decentralized structure and inherent flexibility. IP allowed networks with vastly different physical properties to operate with the same addressing system, while TCP ensured packet delivery to the recipient. That was it. The simplicity of the basic network scheme allowed almost any application to be built on top of it. Importantly, any user could contribute new functionality if they could convince others to use their program. For example, file transfer via FTP was one of the most popular ways to use the internet in its early years, but finding servers offering the files you were interested in, other than through word of mouth, was impossible. Therefore, enterprising users created various protocols for cataloging and listing FTP serversâlike Gopher, Archie, and Veronica.
Theoretically, it had the same flexibility, as well as official endorsement from international organizations and telecom giants as a standard for inter-network communication. However, in practice, the field remained dominated by TCP/IP, and its decisive advantage became the code that first ran on thousands and then millions of machines. It had the same flexibility, as well as official endorsement from international organizations and telecommunications giants for networking standards.
Delegating application-level control to the far reaches of the network has led to significant consequences. It meant that large organizations, accustomed to managing their own domain, could feel comfortable. Organizations could set up their own email servers and send and receive emails without having all their content stored on someone else's computer. They could register their own domain names, launch their own websites accessible to anyone on the internet, while keeping them fully under their control.
Naturally, the most vivid example of layered structure and decentralization has been the World Wide Web. Two decades of systems, from time-sharing computers in the 1960s to services like CompuServe and Minitel, revolved around a small set of core information exchange services â email, forums, and chats. The web became something fundamentally new. The early years of the web, which consisted entirely of unique hand-crafted pages, are nothing like its current state. However, the ability to jump from link to link already had a strange allure, providing businesses with a means of extremely cheap advertising and user support. None of the internet architects planned for the web's emergence. It was the brainchild of Tim Berners-Lee, a British engineer at the European Organization for Nuclear Research (CERN), who created it in 1990 to facilitate the convenient sharing of information among laboratory researchers. Yet, it easily lived on the TCP/IP base and used the domain name system, created for other purposes, for the universally widespread URLs. Anyone with internet access could create a website, and by the mid-90s, it seemed that everyone did â city halls, local newspapers, small businesses, and hobbyists of all sorts.
Privatization
In this narrative about the growth of the internet, I omitted several important events, and you may have some questions left unanswered. For instance, how exactly did businesses and consumers gain access to the internet, which was initially centered around NSFNETâa network funded by the U.S. government and seemingly intended to serve the community of researchers? To answer this question, in the following article, we will revisit certain key events that I have not yet covered; events that gradually but inevitably transformed the government scientific internet into a private and commercial one.
Further reading
- Janet Abatte, Inventing the Internet (1999)
- Karen D. Fraser âNSFNET: A Partnership for High-Speed Networking, Final Reportâ (1996)
- John S. Quarterman, The Matrix (1990)
- Peter H. Salus, Casting the Net (1995)
Source: habr.com
