I P Sharp Associates and the Telephone Monopolies

by Ian Sharp

In 1968 we made a decision to install a computer and go into the time-sharing business. In 1968 the FCC in Washington made a decision to override objections from AT&T and allow non-telephone-company-provided equipment to be attached to the telephone switched network. That case arose in the first place when a court ordered FCC to overturn an earlier decision in which they (the FCC) sided with AT&T and banned the attachment to the telephone network of anything not provided by the telephone company. This 1968 ruling was known as the Carterfone decision1 and its effect on subsequent computer and communications developments was profound.

Bell Canada was not directly affected by the Carterfone decision, but the winds of change were picking up and Bell was well aware that foot-dragging on its part could incur the wrath of the Canadian regulator, CRTC. The role of the Government regulators, FCC and CRTC was largely one of monitoring the behaviour of private communications companies, so that the public could be protected from the worst abuses by companies who had been granted a monopoly. In most of Canada, and the rest of the world outside the United States, the national government itself was the owner and provider of the communications network. Thus any government regulatory authority was in fact regulating another government department, or if you prefer the regulator and the regulated were on the same side..

This atmosphere of protected monopolies did little to provide an incentive for technical development, since by definition the monopoly had no competitors. While the computer industry made huge strides from its start in the 1940s, the communications industry was at a virtual standstill from the 1940s to the 1970s. It is ironic that a lot of important technical pioneering work during this period was carried out at Bell Labs, which was owned by AT&T, and to a lesser extent by Northern Electric which was owned by Bell Canada, yet so little of this output found its way into the monopoly services.

The situation outside North America was much worse. The government in each country was the self-proclaimed monopoly provider of communications services, which were lumped underthe heading of Postal, Telephone and Telegraph services or PTTs. The word "regulation" came to have a different meaning on each side of the Atlantic. In the United States, British Columbia, Ontario and Quebec it meant regulation of the common carriers by a body with legal powers to protect the interests of users. In Europe and elsewhere in the world it meant regulation of the user so that he or she could do nothing that the carrier did not like.

All of the worldwide monopoly providers formed an international gentlemen's club, in which they agreed never to do any business outside their own territory. Since countries had to connect to each other, each country owned exactly half of the connecting link. The club also had committees to formulate standard technical protocols, so that they could in fact communicate with each other, and finance committees to ensure the equitable division of revenue, regardless of who actually collected it. Statutes existed in most if not all countries to protect the rights of the monopoly provider, which included the right to regulate the behaviour of the users with the force of law.

The Carterfone decision provided the first chink in this well established and finely honed armour. One of the larger members of the club, AT&T was now vulnerable to attack by competitors for the first time in the club's history, and attack they did. In retrospect it seems that the club members were not prepared for the newly emerging phenomenon of data transmission and their typical reaction to it was one of denial that such an idea had any merit. They seemed to hope that it was a passing fad and would self-destruct quickly. They had to hurriedly learn about multiplexors, modems, acoustic couplers etc, all of which they had to somehow fit into their kit of products. This led to some odd tariffs such as an installation charge for an acoustic coupler which they mailed in a cardboard box. Since most available computer terminals operated at a low enough speed that many could theoretically use the same voice grade circuit, phone companies such as Bell Canada wanted desperately to preserve their right to attach equipment to the switched network. Therefore, on a very arbitrary basis, they approved the use of some manufacturer's equipment and disapproved of others.This equipment was available commercially, but Bell insisted it must be obtained from them, at a much higher price of course. The situation in Canada was complicated by the fact that Bell Canada operated in Ontario and Quebec only, and GTE operated in British Columbia. Bell and GTE were regulated by CRTC2. In all other provinces the common carrier was owned by the provincial government and was not regulated by CRTC, in fact was not regulated at all. Each province had a different idea about how to deal with their customers, and each was a separate fiefdom operating like the individual fiefdoms in Europe.

Our 1968 decision resulted in the installation of an IBM System 360 model 50 in Toronto in the fall of 1969. We had local dial service in Toronto and Bell provided local dial service in Ottawa and Montreal through specially tariffed foreign exchange lines. In addition we took delivery of a time-division multiplexor and 4800 bps modem from an associated company in Washington DC. Scientific Time Sharing Corporation (courtesy of the FCC) constructed a dedicated network covering their various offices in New York, Washington, Chicago and Los Angeles. This network terminated in Washington and from there was connected via a 4800 bps circuit to our computer room in Toronto. Bell Canada was quite picky about us having one, and only one, border crossing point between Canada and the United States. They had previously introduced that rule to prevent Canadian companies from saving money by communicating with Western Canada through the United States. In the United States interstate tariffs were controlled by the FCC, whereas intrastate tariffs were not. This in turn led to some interesting anomalies like a leased circuit from Los Angeles to New York City being cheaper than the same circuit from Buffalo to New York City.

The effects of the Carterfone decision in the United States were clear to everyone, and in the early 1970s Bell Canada started to hire people who understood that data transmission was here to stay, and it represented a challenge and an opportunity rather than a threat and a nuisance. Relations became more cordial and less confrontational and by 1973 we had a fairly reliable trans-Canada network.

As the result of a successful business relationship with Xerox in North America, we opened an office in London, England and installed a trans-Atlantic line in the fall of 1973. This brought us into contact for the first time with Teleglobe and the British Post Office. Both of these organizations had been fairly well insulated from the effects of the Carterfone decision and both were seasoned members of the afore-mentioned club. Hence both of them regarded us with great suspicion, and unfortunately neither organization was endowed with much in the way of technical excellence. Local dial service in London worked after a fashion, but the error rates on the trans-Atlantic circuit were too high for user comfort. The time-division multiplexing technology that we were using had limited error detection and no error correction. It was this shortcoming that precipitated the development of a homegrown intelligent network, capable of disguising the shortcomings of the long-haul circuits and optimized for interactive terminal use. Time division multiplexing is an inefficient way to allocate bandwidth when data traffic is sporadic as in time-sharing. Michael Harbinson at our Amsterdam subsidiary realized that packet-switching solved the high error rate problem while improving the use of bandwidth. Without any encouragement from head office he began programming using Computer Automation hardware with which he was familiar.

This development allowed for a subsequent rapid expansion of the network throughout most capital cities in Europe, followed by Hong Kong, Singapore, Tokyo and 3 cities in Australia. Most of this had been accomplished by the end of the 1970s and this included the replacement of all the installed North American equipment. Each individual country had its own bizarre rules and regulations, and each one treated us initially as some kind of evil presence. In Belgium it was illegal for a customer to connect to our Brussels centre by means of a leased line, the customer could only use dial-up facilities. In Spain it was illegal for a customer to connect to our Madrid centre by dial-up, he could only use a leased line.

In Australia the only equipment approved by Australian Telecom for attachment to the network had to be manufactured by Digital Equipment Corporation (DEC). Since our node did not meet that requirement, it was necessary for our Sydney branch manager to obtain a DEC sticker and affix it over the offending words 'Computer Automation'. We subsequently passed the inspection of the local Telecom police.

A constant source of irritation was the universal attitude of all telephone companies to the idea of electronic mail. It was regarded as a huge threat to the considerable income that they all derived from Telex traffic. Telex was an ultra-low speed communication system among Teletype terminals. The service was introduced in the mid 1940s as the solution to all corporate data transmission requirements. No significant improvements or upgrades were made to that service over the next 30 to 40 years. The first real threat came with the advent of electronic mail, or at least that was how electronic mail was perceived. The first reaction of any monopoly to a perceived threat is to ban it outright, and that is more or less what the telephone companies of the world did. As a company we had no serious intention of offering a general inter-company communications facility that might have competed with Telex. We were mainly interested in an internal communication system among employees and a mechanism to communicate with our customers. Assuaging paranoia was almost a full time job and was probably the subject of more meetings around the globe than any other single issue.

We never did succeed completely in allaying suspicion, and many phone companies were sure that we were eating their Telex lunch, even though they had no evidence of it. As a gesture of goodwill, we modified our Email software so that it was not possible for one customer to communicate with another, but that was largely a waste of time since they did not really believe that we had done that. In Germany the Bundespost invented a tariff which they said we must apply to all Emails and send them money each month. Since we did not have revenue from electronic mail in Germany, we never sent them "their share" and eventually the issue atrophied.

It was of course the proliferation of computers and private corporate networks which caused the erosion of Telex use. More and more people started to use electronic mail and the telephone companies were simply over-powered by the tidal surge of individual as well as corporate usage. It is doubtful that as many as 1% of the present population of Email users has any idea that just a few short years ago the whole thing was illegal.

In the late 1970s some of our customers had grown to the point that their time-sharing bills were an embarrassment. This provided us with a business opportunity to charge the customer for the use of our time-sharing software which we would install at his data centre.The architecture of the network allowed for some software modifications which would permit a number of time-sharing hosts to co-exist on the network. This offered the considerable advantage of virtual transparency to the remote end user, who could continue to do what he had always done, once he had selected his appropriate host. The only problem was that now we were selling communications on an international basis. Some of our larger law-abiding customers asked their local phone company if they could buy their national and international communications from us. The answer of course was in the negative. We made every attempt to justify the legality of the arrangement, relying on an argument that the network was only connected to software systems owned by us, and who really cared about who owned the hardware. Some customers considered our argument had merit, but they did not wish to risk the possibility of legal action or worse from their traditional communications provider, so they went ahead and installed their own private network more or less in parallel with ours, which of course was a very expensive approach. Other customers thought that our arguments were completely sound and hence not worth discussing with anyone else. In the absence of any discussion, there was no awareness of a worldwide network terminating in the customers' data centre.

Probably the most interesting case involved a large customer in Copenhagen, who approached the national phone company with a request that they needed local dial connections in all the locations where they had subsidiary companies, which included North America, the Far East, Australia, Japan and numerous locations in Europe. They needed such a network in order to remain competitive in their very competitive world. The phone company told them that such a network was currently beyond their capability, but could well be available at some time in the future. So with only a minor push from the Danish government, the phone company agreed to the temporary installation of our full software system including communications, pending the availability of a locally-provided solution. .

As a face saver the national PTT would like to be paid for all traffic in each direction on the line between the customer's site and our office in Copenhagen. We agreed to that, and since the PTT had no means of measuring that traffic, we offered to do it for them and provide the monthly information that they needed in order to bill us. We did that, but no invoice was ever received.

The conspicuous success of the Internet and the ubiquity of Email has effectively removed the telephone companies from any position of influence. Their traditional businesses are all eroding quickly and they are scrambling to maintain some relevance in the domain in which they enjoyed a monopoly for about a century.

[1] Further information on the Carterfone (aka Carterphone sic) can be found at:
Short, breezy history of AT&T breakup
Short biography of Thomas F. Carter
Full text of FCC Carterfone decision

[2] Attachment rules were guided by: "An Act respecting the Bell Telephone Company of Canada" (Chapter 48), In Acts of Parliament, 27th Parliament, Second Session 1968, p 43