Since TELECOM 91 the landscape of on-line services has changed dramatically, fuelled by the rise in popularity of consumer-oriented networks and by the huge growth in business-oriented applications - such as electronic messaging and Electronic Document Interchange (EDI) - over commercial networks.
The consumer-oriented networks, which exist specifically to provide services to individuals (end users), include CompuServe (2.7 million subscribers at the end of 1994), America On-line (AOL; 3 million subscribers in mid-1995), Prodigy (1.3 million subscribers in mid-1995), the Microsoft Network (MSN; limiting 1995 operation to 500,000 subscribers), Apple Computer's eWorld (90,000 subscribers), and a host of regional or local network service providers. In 1995, according to the research group Simba, on-line subscribers in the USA alone total over 11 million, and the major American on-line service providers are expanding to worldwide operation, like CompuServe in the UK and France, and AOL in Germany. In practical terms for users, this means providing telephone access by local call in the country of operation.
Local network providers typically offer a smaller range of services to a more experienced and sophisticated audience. In many cases they take advantage of the economies of offering few value-added services, providing only connection to the Internet and an account with space on a multi-user computer system like those used commercially. Internet Service Providers (ISPs) may require a greater degree of skill from their individual users, but can also reach beyond individuals to provide both consumer and commercial accounts.
A commercial account connects to the Internet via the ISP from its own computers at its own site, and effectively has direct access to Internet-based services in their "raw" form.
In short, we are approaching a marketplace for on-line services that is in some respects much like the marketplace for traditional goods, with a mix of commercial, consumer-oriented, private, public, and tiered service suppliers.
Perhaps the most conspicuous and newsworthy form of networking recently has been the Internet itself. The term "Internet" is used in several ways:
The Internet can therefore be viewed as something similar to an entire continent's system of roads - some private, some public; some large, some small; some rapid, some slow; but all connected together - with each section managed and operated by a relatively independent authority. Because of this distribution of authority, there is no way to know how many people use the Internet. The safest underestimate would be at least one user per known host (system, node), or over 6.6 million, but realistic estimates easily exceed 40 million.
After taking more than eleven years to grow to 1,313,000 systems by January 1993, the number of systems known on the Internet more than quintupled by July 1995 to 6,642,000 (the true number is greater, but no precise method exists for taking the census). Traffic on the backbone segments more than quadrupled in those two and a half years, to over 22 million million bytes (terabytes, TB) per month.
A major contributor to these phenomenal increases is the appearance of a "killer application" on the Internet - the WorldWide Web (WWW, W3, "the Web"). Two factors differentiate the Web: a powerful and flexible underlying protocol able to carry multimedia data, and the ready availability of inexpensive user-friendly Web HyperText "browsers". Browsers are user interface programs capable of presenting Web "pages" (the collection of information transmitted at one time for display and interaction on screen) attractively on many different kinds of computer systems.
The figures speak for themselves: in January 1993 the Web had just become public, and by March 1995 it represented over 4.8 TB of traffic per month - a quarter of everything travelling on the Internet backbone.
The Web and its browsers can support on-line services which once required an individual interface for each data type:
It is significant that in the first few years of the Web, the user interfaces of browsers have been similar to one another irrespective of platform. And the overall quality of user interfaces appears to be a significant factor in the growth of on-line services in general.
An earlier generation of commercial on-line services, typified by medical and legal information retrieval services, was aimed at specialists and accessible only through private connections to the private networks that housed them. Until recently they were used only with text interfaces, and the French Minitel system still does use a text interface, with elementary graphics, videotex.
Since 1983, France Telecom's customers have been able to have a basic Minitel terminal instead of a printed telephone directory. The basic terminal itself and directory information is available free; more elaborate models can be rented for a fee. By the standards of today's personal computers, Minitel is slow, unattractive, and antiquated. It's a small, dull, terminal in an interactive multimedia world. Yet Minitel has been offering thousands of on-line services to more people (14.4 million) longer than any other on-line service provider, and in the process generates US$ 1.2 billion per year in revenue - more than twice as much as CompuServe. Minitel users can check on airline or train schedules, participate in contests, transfer money between their bank accounts, find the book value of a used car, check their children's examination results, or make a date.
As a bellwether for preferences in on-line services, the Minitel experience suggests that today's consumers are interested in home banking, home shopping, and what might be called "useful information" - such as train and airline schedules.
This has not always been the case. Through the 1980s, games and the so-called "pink" (sex-oriented) services represented 22 per cent of revenues. Since 1991 this has fallen to 14 per cent, overtaken by practical services and a more conservative clientele. About a third of the adult population in France uses Minitel, and Minitel terminals are found in most post offices. Although many of the services offered have been profitable, Minitel as a whole has not been profitable for France Telecom, whose investment (to date US$ 11 billion) will not be recovered - so they have been reported to say (Communications International, July 1995) - until "well beyond 2000 AD". And usage per unit is decreasing. Experts do not see a long-term future for Minitel in its present form, especially in competition with other more attractive modes of on-line service, some of them bundled into offerings of consumer-oriented on-line service providers.
Nonetheless, basic services are easy to use, the telephone system is perceived as reasonably secure, and payment for Minitel services is processed via the consumer's regular telephone bill. These are very attractive points in contrast to the uncontrolled diversity of payment schemes now in use on the Internet, where vendors have different methods for billing and payment. And they are in great contrast to the lack of security on the Internet.
Although the notion of a unified payment scheme is attractive, there is as yet no agreement on how to handle monetary payments over networks. There are, however, several promising approaches. One of these is the idea of "digital cash", such as the method proposed by David Chaum (of DigiCash BV in the Netherlands), a respected figure in cryptography. Chaum's digital cash offers immediate verified payment without the exchange of account information, even over insecure network connections, and can be used anonymously and untraceably like traditional cash, making digital cash attractive to those concerned about privacy questions such as surveillance and "data mining" of information collected from on-line transactions.
Another approach is that of Netscape, which uses patented public-key encryption technology to conceal account information transmitted over insecure channels. Experts generally agree that the variant available domestically in the United States is extremely secure because of the length of the keys it uses, but because strong encryption is USA export-controlled technology, the variant approved by the US government for use outside the USA must use shorter keys and is therefore not so secure. (It has already been broken at least twice by knowledgeable amateurs.) Other payment schemes rely to various degrees on encryption or on protocols that use information exchanged off the network between buyer and seller.
The demand for secure transactions over insecure networks contributes to the volatility of a situation with national policy overtones, exposing a conflict of interest between businesses and consumers who want to transact business safely on-line, and governments wishing for reasons of national security to control the use of strong encryption.
An interesting development is the United States Postal Service's plan to act as a trusted authentication agent, guaranteeing the identities of participants in digital transactions.
According to Paul Callahan of Forrester Research (Communications International, August 1995), "the real problem is making sure that you know with certainty who is at the other end of the conversation. Authenticated digital signatures will give users a way to prove who they are electronically, and encryption will make sure that thieves can't steal data." Authentication, of course, is unnecessary with Chaum's type of digital cash...
Services now available on-line can be classed in several ways.
A final major theme is what the telecoms industry calls "convergence", in which the telecommunications media are either already digital or headed to become digital. When they do, the distinction between computer networks and telecommunications networks may entirely disappear.
Further, say industry experts, we can expect to see a blurring or even disappearance of the distinction between smart telecoms devices and telecom-equipped computers. Put simply, there is no distinction between a television equipped to compute and provide network services, and a computer equipped for television.
In Japan today, a significant proportion of personal computers are sold with a television tuner; and interface cards are available for personal computers that can provide a complete telephone interface including signalling, data modem, fax modem, answering machine, and caller number identification (CNID).
On-line services in a converged world may enable an individual to watch a home shopping channel broadcast, or an on-line video-on-demand shopping service, and buy videos, audio recordings by the track, or material goods directly and immediately over the network - and then in the next moment use the same equipment to send electronic mail or surf the Web.
Although the users of on-line services number in the millions, it is only, so far, small numbers of millions. In that sense, on-line services haven't been the same unqualified success as telephones, television, or even the catalogue shopping by mail that originated in the 19th century. Is there any reason to believe that they may succeed now?
There is no certain answer, but there is a strong possibility that it is yes, because several things have changed.
The most obvious of these is the spread of the personal computers which accustom us to interactive digital applications: this is no surprise to children playing computer games or to adults using spreadsheets and on-line applications in the home and at work. Another is the improved usability of on-line services, signalled by the phenomenal growth of the WorldWide Web.
And finally it has become clear that when substantially better ways are found for individuals to use the capacities of networks directly - with excellent user interfaces employing multimedia, and with access to useful services and vast amounts of information - people rapidly adopt them and on-line service vendors equally rapidly provide them.
We can envision a converged on-line environment, in which a computing telecoms device is left permanently on like a telephone, to provide an immediate "network dial tone". And over that device, people will obtain information, conduct business, and be entertained.
Click Here to Return to Telecommunications Features and Press