Get best of the week

Internet History: Era of Fragmentation; part 1: load factor



Other articles in the series:


By the beginning of the 1980s, the foundation had been laid for what we know today as the “Internet” - its basic protocols were developed and tested in combat conditions - however, this system remained closed and was almost completely controlled by the only entity, the US Department of Defense. Soon this will have to change - the system will expand to all departments of computer science of different institutes using CSNET. The network will continue to grow in academia before finally fully opening up for general commercial use in the 1990s.

But the fact that the Internet will become the very center of the advancing digital world, the widely advertised “information community”, was not at all obvious in the 1980s. Even for people who heard about him, he remained only a promising scientific experiment. But the rest of the world did not stand still, holding his breath, awaiting his arrival. Instead, many options for providing access to online services for the mass consumer competed for money and attention.

Personal computing


Around 1975, breakthroughs in semiconductor manufacturing led to the emergence of a new type of computer. A few years before, engineers figured out how to shove the main logic of data processing onto a single microchip - a microprocessor. Companies such as Intel have begun offering high-speed short-term memory on chips to replace the memory on the magnetic cores of the previous generation of computers. As a result, the most important and expensive parts of the computer came under the influence of Moore's Law, which in the following decades constantly reduced the cost of processor chips and memory. By the middle of the decade, this process had already so cheapened these components that a representative of the middle class of America could already consider the opportunity to buy and assemble your own computer.Such machines began to be called microcomputers (or sometimes personal computers).

There was a fierce struggle for the right to be called the first personal computer. Some considered LINC of Wes Clark or TX-0 from Lincoln's laboratories as such - after all, only one person could use it interactively. If you drop the questions of primacy, then any candidate for first place, if you evaluate the historical sequence of events, is obliged to give it to one obvious champion. No other machine has been able to achieve the catalytic effect that the MITS Altair 8800 spawned in the explosive growth in the popularity of microcomputers in the late 1970s.


Altair 8800 standing on an additional module with an 8 "drive

Altair has become a seed crystal for the electronics community. He convinced amateurs that a person could assemble their own computer at a reasonable price, and these amateurs began to gather in communities discussing their new machines - for example, the Homebrew Computer Club ["club of home-made computers"] in Menlo Park. These amateur cells launched a much more powerful wave of commercial microcomputers, based on mass production machines that did not require electronic skills - for example, Apple II and Radio Shack TRS-80.

By 1984, 8% of US households had their own computer, which amounted to about seven million cars. Meanwhile, enterprises acquired their own fleets of personal computers at a speed of hundreds of thousands of units per year - mainly, these were the IBM 5150 and their clones. In the more expensive single-user computer segment, the market for workstations grew from Silicon Graphics and Sun Microsystems, more powerful computers with advanced graphic displays and networking equipment intended for use by scientists, engineers, and other technical experts.

Such machines could not be invited to the exquisite world of ARPANET. However, many of their users wanted access to the promised merger of computers and communications, which theorists have been broadcasting in the popular press since Taylor and Liklider’s 1968 article, “Computer as a Communication Device,” and some even earlier. Back in 1966, scientist John McCarthy promised in Scientific American that "the technology already demonstrated to us is enough to imagine how computer consoles appear in every home, connected to public computers by telephone." He stated that the range of services offered by such a system is simply impossible to list, but he cited several examples: “Everyone will have access to the Library of Congress, and better quality than the librarians now have.Full reports on current events will be available, be it baseball results, the smog index in Los Angeles, or a description of the 178th meeting of the armistice commission in Korea. Income taxes will be automatically calculated due to constantly accumulating records of income, deductions, deposits and expenses. ”

Articles in popular literature described the possibilities of e-mail, digital games, and various services from legal and medical consultations to online purchases. But how exactly will all this look? Many of the answers were far from the truth. Looking back, that era looks like a broken mirror. All the services and concepts that characterized the commercial Internet of the 1990s - and many others - appeared in the 1980s, but fragmented, scattered across dozens of different systems. With some exceptions, these systems did not intersect, they stood apart. The users of one system had no way to interact or communicate with users of the other, so the attempts to attract more users to any of them were mostly a zero-sum game .

In this article, we will consider one subset of participants in this new digital territory capture - companies that sold shared access, tried to enter a new market with attractive conditions.

Load factor


In 1892, Samuel Insal, Thomas Edison’s protege, traveled west to head Edison’s new electric empire division, the Chicago Edison Company. In this post, he consolidated many key principles of modern utility management, in particular, the concept of load factor - it is considered as the average load on the electrical system divided by the highest load. The larger the load factor, the better, since any deviation from the ideal ratio of 1/1 means waste, which is unnecessary funds that are needed to process peak downloads, but are idle in schedule failures. Insal decided to fill the gaps in the query curve by developing new classes of consumers who would use electricity at different times of the day (or even at different seasons), even if they had to sell them electricity at a discount.In the early days, the supply of electricity was mainly used for lighting houses, and for the most part, in the evening. Therefore, Insal began to advertise the use of electricity in industrial production, increasing its daily consumption. This left failures in the mornings and evenings, so he convinced the Chicago transportation system to convert the trams to electric traction. In this way, Insal maximized the value of invested capital, even though sometimes it was necessary to sell electricity at a discount.so he persuaded the Chicago transportation system to convert trams to electric traction. In this way, Insal maximized the value of invested capital, even though sometimes it was necessary to sell electricity at a discount.so he persuaded the Chicago transportation system to convert trams to electric traction. In this way, Insal maximized the value of invested capital, even though sometimes it was necessary to sell electricity at a discount.


Insal in 1926, when his photo was placed on the cover of Time magazine.

The same principles apply to investing capital in computers almost a century later - and it was the desire for a balanced load, which led to the offer of discounts at a time when there were no peak downloads that allowed two new online microcomputer services, launched almost simultaneously in the summer of 1979: CompuServe and The Source.

CompuServe


In 1969, the recently established Golden United Life Insurance Company in Columbus, Ohio, registered a subsidiary of Compu-Serv Network. The founder of Golden United wanted to create the most advanced, high-tech company with computerized record keeping, so he hired a young graduate student in computer science, John Goltz, to lead this project. However, DEC's sales manager talked to Goltz to buy the PDP-10, an expensive car whose computing capabilities significantly exceeded Golden United's current needs. The idea of ​​creating Compu-Serv was to turn this error into an opportunity - it was planned to sell surplus computing power to customers who could connect to the PDP-10 via a telephone line from a remote terminal.In the late 1960s, such a model with the separation of time and sales of computing services was quite actively distributed, and Golden United wanted to snatch its piece of cake. In the 1970s, a division of the company split into an independent unit, renamed CompuServe and created its own packet-switched network to offer low-cost and nationwide access to computer centers in Columbus.

The national market not only gave the company access to more potential customers, it also expanded the demand curve for computer time, spreading it across four time zones. However, there was still a big gap between the end of the day in California and the start of the day on the East Coast, not to mention the weekend. CompuServe CEO Jeff Wilkins saw an opportunity to solve this problem with a growing fleet of home computers, as many of their owners spent evenings and weekends on their electronic hobbies. What if we offer them access to e-mail, bulletin boards, and games on CompuServe computers at a discount in the evening hours and weekends ($ 5 / hour, versus $ 12 / hour during working hours)? [in current money it is $ 24 and $ 58 respectively].

Wilkins launched a trial service, calling it MicroNET (specially distanced from the main CompuServe brand), and after a slow start it gradually turned into an incredibly successful project. Thanks to CompuServe having a national data network, most users could simply call their local number to get to MicroNET, and thereby avoid long-distance bills, even though the real computers they contacted were in Ohio. When the experiment was successful, Wilkins abandoned the MicroNET brand and transferred it to the CompuServe brand. Soon, the company began offering services specifically designed for microcomputer users, such as games and other software that could be bought online.

However, the most popular services by a wide margin are communication platforms. For long-term discussions and posting content, there were forums whose topics varied from literature to medicine, from working with wood to pop music. CompuServe usually gave forums to the users themselves, and some of them, who took on the role of "sysops", were engaged in moderation and administration. Another major messaging platform was CB Simulator, which Sandy Trevor, one of CompuServe's directors, sketched overnight. It was named after the then popular hobby of amateur radio (citizen band, CB), and allowed users to sit in real-time text chat rooms on dedicated channels - this model was similar to talk programs available on many time-sharing systems.Many users hung around for hours in CB Simulator, chatting, making friends, and even finding lovers.

The Source


On the heels of MicroNET, another online service for microcomputers was launched, launched just eight days after it, in July 1979. In fact, it was aimed at almost the same audience as the Jeff Wilkins service, despite the fact that it developed completely according to another pattern. William von Meister, the son of German emigrants, whose father helped organize airship flights between Germany and the USA, was a serial entrepreneur. He began to engage in a new enterprise as soon as he lost interest in the old one or as soon as disappointed investors ceased to support him. It was hard to imagine a person more unlike Wilkins. In the mid-1970s, his greatest successes were: the Telepost electronic messaging system, which sent them electronically across the country to the switch closest to the recipient,and overcoming the last mile in the form of mail delivered the next day; TDX system, which used computers to optimize the routing of telephone calls, which reduced the cost of long-distance calls for large enterprises.

Predictably losing interest in TDX, von Meister in the late 1970s was enthusiastic about the new project, Infocast, which he wanted to launch in McClean, Virginia. It was, in fact, an extension of the Telepost concept, but instead of using mail to deliver the message at the last mile, he had to use the FM sideband (according to this technology, the station name, musician name and song name are transmitted to modern radios) to deliver digital data to computer terminals. In particular, he planned to offer this to highly geographically distributed enterprises, which had many points that needed regular updating of information from the central office - banks, insurance companies, grocery stores.


Bill von Meister

However, what von Meister really wanted to create was a nationwide network for delivering data to homes through terminals for millions, not thousands. However, it is one thing to convince a business to spend $ 1,000 on a special FM radio and terminal, and another to ask private consumers to do the same. Therefore, von Meister went in search of other opportunities to deliver news, weather information and other things to homes; and he found such a method in hundreds of thousands of microcomputers that crawled through American offices and homes like mushrooms and appeared in houses already equipped with telephone lines. He took in partnership with Jack Taub, a wealthy businessman with connections who liked the idea so much that he wanted to invest in it. Taub and von Meister first named their new CompuCom service,in a characteristic way for computer companies of those days, trimming and composing words, but then came to a more abstract and ideological name - The Source.

The main problem they encountered was the lack of a technical infrastructure capable of serving the realization of this idea. To get it, they entered into an agreement with two companies whose resources in total were comparable to those of CompuServe. They had computers suitable for working on a time-sharing system, and a national data network. Both of these resources were almost idle in the evenings and weekends. Computer power was provided by Dialcom, headquartered on the banks of the Potomac River in Silver Spring, Maryland. Like CompuServe, it began operations in 1970 as a time-sharing computer services provider, although by the end of the decade it had already offered many other services. By the way, it was thanks to the Dialcom terminal that I first got to know computersEric Emerson Schmidt , future chairman and CEO of Google. The communications infrastructure was provided by Telenet, a packet-switched network that spun off at the start of the decade from Bolt, Beranek and Newman , BBN. Paying access to Dialcom and Telenet services at a discount during minimum hours, Taub and von Meister were able to offer access to The Source at a price of $ 2.75 per hour at night and on weekends with a down payment of $ 100 (this is $ 13 per hour and $ 480 down payment in today's dollars).

In addition to the payment system, the main difference between The Source and CompuServe was the expectation of users using their system. The earliest services from CompuServe included email, forums, CB, and program exchange. It was assumed that users would independently create their own communities and build their superstructures on top of base iron and programs - just as corporate users of time-sharing systems do. Taub and von Meister did not have experience with such systems. Their business plan was based on providing a lot of information for higher-end professional consumers: the New York Times database, news from United Press International, stock information from Dow Jones, airfare, reviews of local restaurants, and wine prices. Maybe,the most characteristic feature was that The Source users were greeted by an on-screen menu from the available options, while CompuServe users were welcomed by the command line.

Consistent with the personal differences between Wilkins and von Meister, the launch of The Source turned out to be as grand an event as MicroNET was quietly launched. Isaac Asimov was invited to the first event so that he could personally announce how the arrival of science fiction became a scientific fact. And, typical of von Meister, his work at The Source did not last long. The company immediately faced financial difficulties due to the serious superiority of expenses over revenues. Taub and his brother had a large enough share in the business to oust von Meister from it, and in October 1979, just a few months after the launch party, they did so.

Decline in time-sharing systems


The last company to enter the microcomputer market due to the load factor logic was General Electric Information Services (GEIS), a division of the manufacturing giant in the field of electrical engineering. GEIS was founded in the mid-1960s, when GE was still trying to compete with others in computer manufacturing as part of an effort to push IBM out of its dominant position in computer sales. GE tried to convince customers that instead of buying computers from IBM, it was easier for them to rent computers from GE. This attempt had virtually no effect on IBM's market share, but the company earned enough money to keep investing in it until the 1980s, and by then GEIS already owned a worldwide data network and two large data centers in Cleveland, Ohio, and in Europe.

In 1984, someone from GEIS noticed how well The Source and CompuServe were growing (the latter at that time already had more than 100,000 users), and figured out how to make computer centers work outside of the main load hours. To create their own offer for users, they hired CompuServe veteran Bill Lauden. Lauden, annoyed by how corporate sales managers started trying to get into an increasingly attractive consumer business, left the company with a group of colleagues to try to create their own online service in Atlanta, calling it Georgia OnLine. They tried to turn the lack of access to the national data network into an advantage by offering services specifically designed for the local market, such as special advertising and information on various events, but the company went out of business.therefore, Lauden was delighted with the proposal from GEIS.

Lauden called the new service GEnie genie - genie] - it was the backronym of the General Electric Network for Information Exchange [GE information exchange network]. She offered all the services developed at that time in The Source and CompuServe - chat (CB simulator), message boards, news, weather and sports information.

GEnie was the latest personal computer service born out of the computing access industry with time sharing and load factor logic. With the increase in the number of small computers to millions of pieces, digital services for the mass market by themselves began to gradually become an attractive business, and ceased to be just a way to optimize available capital. In the early days, The Source and CompuServe were tiny companies that served several thousand subscribers in the 1980s. Ten years later, millions of subscribers already paid a monthly fee to the United States - and CompuServe was at the forefront of this market, absorbing its former rival, The Source.The same process made time-sharing access less attractive for business - why pay for communication and access to someone else's remote computer when it became so easy to equip your own office with powerful machines? And until the advent of fiber-optic channels, which dramatically dropped the cost of communications, this logic did not change its direction to the opposite.

However, this market could be accessed not only for companies offering access with time sharing. Instead of starting with large mainframes and looking for ways to download them to the fullest, other companies started with the technology that was already in the homes of millions of people and began to look for ways to connect it to a computer.

What else to read
  • Michael A. Banks, On the Way to the Web (2008)
  • Jimmy Maher, “A Net Before the Web,” filfre.net (2017)

More articles:


All Articles