Windows into Cyberspace

By Håkon W Lie

Information exchange formats
The content of Cyberspace
Telecommunications
Navigating Cyberspace
The inhabitants of Cyberspace
Hackademics
Merchants
The Nintendo generation
Postscript
Further reading
This issue of Telektronikk will describe a world into which many people enter, but few escape. The lion's cave of the 1990s is filled with people like yourself - very much alive and kicking, but still entangled in an addictive web of technological and social relationships.

Reading this journal is probably not addictive in itself. In order to get hooked you need access to a computer which will be your window into cyberspace. Chances are that you already have a computer, and perhaps it is even 'on the net' as cyberspace inhabitants call machines they can reach. If you are reading this on paper, the best thing to do is to throw away the pulp and start exploring cyberspace through your own window. In there somewhere you will find an electronic version of this document (for hints on where to find it, see the text box The electronic Telektronikk).

Whether you are reading it on paper or not, this introduction will present some fundamental concepts used in the articles that follow. Also, it tries to point at some of the current issues in the world of networked computers. These issues are not within a clearly defined field, and 'Cyberspace' is one of the few terms that are general enough to include most of them. We are currently seeing several fields converge in a confusing array of mergers, foundations and strategic alliances. Negroponte's circles from 1978 - in the figure he predicts the coming together of three industries: broadcast, publishing and computers - are coming closer (figure 1a). In addition, at least two more industries are joining the loop: telecommunications and video games (figure 1b).

The common denominator in these fields is information. Information is also the foundation of computer networks, and we start by looking at the vehicle of information distribution: the exchange format.

Information exchange formats

There are many ways to classify information; one well-known system is Dewey's which is used to categorise information into thematic hierarchies. Computers are not good at dealing with thematic issues; from the computer's perspective, the difference between a textual description and a picture (e.g. a picture of a chair and a text describing the chair) is much more significant than the difference between two pictures, no matter how different they are. Therefore, we start by classifying information into monomedia: text, graphics, images, audio, and video. Among these, text is arguably the easiest to handle. Let us take a closer look at text.

Computers work with numbers. They are in fact very good at doing simple things with numbers, like adding two of them together, or remembering them. So, if we want a computer to store, process, transfer, or present information - there are some very good reasons for wanting this - we need to represent the information in terms of numbers.

Let us take a simple example: we want the computer to remember the first sentence of this paragraph so that we can retrieve it later. The first letter is 'L', so we will have to assign a number to represent the letter 'L'. Since we already know the Latin alphabet we start counting and pick the number 12 to represent 'L'. Next letter is 'e'. Again counting - 'e' must be represented by the number 5. Wait! 5 represents 'E', not 'e'. We will have to pick a new number for 'e' - how about 101? Doesn't seem logical? It really is not, but at one point in history, the number 101 was chosen to represent 'e' in a table known as ASCII (see the text box An historical anchor: ASCII).

The ASCII table has become an internationally accepted way of coding text. In addition to 'e' and 'E', all common characters can be found in there. It contains 128 entries which at first glance seems to be sufficient for representing all texts. But, if that was the case, your favourite word processor should have no problems exchanging data files with your neighbour's since they both could use the ASCII table. And, in fact, most word processors know ASCII and can use ASCII when saving or retrieving documents. But, when doing so, information is lost. When your document comes up on your neighbour's screen, chances are that your favourite font has been replaced with gothic, and your carefully prepared layout is lost. Also, your native language's special characters - if they were present - surely have degenerated into some other special interest language's characters. There have been - and still are - attempts to create a common format for text, but no format has yet become universally accepted.

Computers have been processing text for some decades, but text is still - as we have seen - problematic as a data type. In the last decade, computers have evolved into machines that also claim the ability to process and present other data types such as audio, video and graphics. Multimedia is here - at least it will show up tomorrow if we can find the right distribution format (for more on Multimedia, see separate text box).

The introduction of more data types further complicates the task of designing and agreeing upon formats for storage and distribution. Isolated, they each exhibit some of the same problems as text; what properties do we want to store, to what level of detail, and how do we encode these properties? Together, they pose new problems when incorporated into a compound format to handle multimedia publishing. How do the various data segments relate to each other in time, space and content; how should these relationships be encoded; and to what level of detail? Also, it is debatable whether layout information should be part of the document - perhaps the reader should be able to set layout preferences? There exists a tension between the interests of the author and the interests of the readers, and the document format is where the battle will take place.

After having gone through some of the problems of defining a format for multimedia messages, one may welcome the fact that other approaches exist. The fax machine is based on one of them; instead of encoding a document as individual characters, words, sentences, and paragraphs, the fax machine treats each page of the document as a picture. This has several important implications:

First, the fax machine is able to convey anything visible on a sheet of paper, whether it be text, figures or photographs.

Secondly, the computer - for the moment camouflaged as a fax machine - loses the ability to do neat things with text that computers are good at (e.g. searching) since it has no notion of text whatsoever. All it sees is images, and computers cannot perform operations on the content of images. If you want the computer to process the textual content of a fax message, you most likely have to retype it yourself.

Thirdly, due to the strong binding to paper, fax machines will not be able to convey audio or video.

Also, the fax machine has probably wasted both time and paper before it reaches the recipient. A typical path for a faxed document is described in figure 2.

Although a highly successful telecom product in terms of use, the fax machine is a horror example of the importance of taking the changing topology of telephone networks into account when defining information exchange formats; computers with memory and processing capabilities are increasingly terminating our networks, and we should no longer design our protocols for the lowest common denominator - of which the fax machine is a prime example.

The content of Cyberspace

Creating formats for computers to exchange information is an interesting exercise with some thought-provoking implications, but for most people the applications that take advantage of these formats are more relevant. So, let us start at the other end by looking at some applications of distributed multimedia.

Given the freedom of a fresh start, what applications, or more to the point - what data do we ideally want to be presented to, and in what form? The answers to these questions will vary from person to person, and from time to time. People have different backgrounds, interests, skills, handicaps, and preferences. Some of these variables change slowly - learning a new language takes years, while others change frequently. When starting your car, you acquire a new handicap with regard to information systems; suddenly you lose your eyesight. Technology has the potential of compensating for handicaps (your computer can read to you while you drive), and adjust for personal preferences (by e.g. helping you find information items you are seeking).

One metaphor, into which several of the applications discussed in this issue can be put, is the 'personalised newspaper'. The newspaper as we know it on paper is a universally accepted product that over the last centuries have developed into a highly advanced presentation medium. The front page gives readers an instant overview of the most important stories, and the headlines make it possible to scan large amounts of information quickly (see figure 3). Simply by switching from scanning to reading, the user is able to change modality from overview to detail. The front page is your menu in which the selections are available immediately.

However, paperbased newspapers have deficiencies that are much due to the distribution technology currently utilized. All copies of a daily edition contain the same information which the editors have selected based on the perceived interests of the readers. If we replace the printing press, trucks and paper boys with electronic distribution, we suddenly achieve new degrees of freedom:

If we exploit the freedoms offered by the technology we will end up with a medium quite distinct from current newspapers. The new medium will have its own characteristics that surely will be subject to debate in years to come.

Another angle from which we can approach the same idea is labelled interactive TV. TV, in its current linear form offers no user participation, except for switching channels. Some TV broadcasters - eager to extend their domain and fearful of a future without them - have started well-publicised experiments with interactive TV. In Europe, TV stations enclose pages of text that carry news, advertisements and programming information in the TV signal. The user selects information from menus through the remote control. If one includes more memory and processing capabilities into TV sets, one can easily imaging video being selected interactively in a similar fashion. However, true interactive TV will first become a reality when the feedback loop closes and the user gets a chance to respond to and request information. We will come to that shortly.

Whatever angle one takes, huge amounts of information will be available, and computers will assist selection and presentation.

Telecommunications

Today's telecommunication systems use a mixture of digital and analogue transmission (for a description of the difference between analogue and digital, see the text box Digital vs. analogue) When you pick up the handset, chances are that you are connected to your exchange with a pair of copper cables. Most likely, these cables transfer your voice as an analogue signal from your telephone to the telephone company's exchange. In the exchange, your voice is transferred into the digital domain where it stays until it reaches the exchange of your interlocutor. For voice converted, which is analogue by nature, this works fine (see figure 6a). The short distance over which the signal is analogue does not degrade it noticeably.

For computers, however, which start out with a digital signal, the number of conversions double (see figure). Making audible signals out of ones an zeros - which is exactly what a modem does to an outgoing signal - is not a very efficient method of transmitting data over copper cables. A typical modem is able to send 9600 bits per second using an analogue phone line - only a fraction of the potential. See figure 6b.

The logical solution to this problem is to complete the digitisation process, i.e. transmit data digitally all the way to your house or desktop. There, anything that naturally comes in the analogue domain (like your voice) will be converted, and anything that is discrete by nature (e.g. text) will flow more gracefully. The telephone companies have labelled this step ISDN (Integrated Services Digital Network), and they have been planning it for decades. See figure 6c.

ISDN will run over the copper cables that already connect households to the telephone network. At 128kbits/s, ISDN offers a performance increase of a magnitude compared to current modems. Also, this bandwidth can be split into several chunks so that you can continue talking while the computer juggles data on your behalf.

128kbit/s is good news if you are used to current modem technology, but you still need more than 11 hours to transfer the content of a full CD-ROM. Often you do not care since you have no need for 600Mb of archived weather information when all you want to know is if you should bring your umbrella - and that is exactly one bit of information. But there are times when nothing but bandwidth will suffice.

It has long been assumed that broadband services require the introduction of fibre to the home or desktop. This axiom is now being questioned by a copper renaissance movement. One technique that has gained a lot of attention lately is ADSL (Asynchronous Digital Subscriber Line), which will offer up to 6Mbit/s into your home over existing wires (see the text box ADSL: a concept for high-speed transmission on copper lines by Knut C Aas for a description of how normal phone lines can achieve this). Going the other way - from your home to the exchange - the numbers are not that impressive, but we all consume more information than we produce.

The next step on the scale of performance is ATM (Asynchronous Transfer Mode) which will start at 155Mbits/s - enough to transmit several channels of high-quality video and audio. ATM is a rare example of fruitful co-operation between telecom and computer vendors, and it will be used to connect exchanges as well as local area networks. Do not expect this service to be offered into your home yet.

While striving for more bandwidth, one should also try to exploit what is already available. A number of interesting services can be offered through a standard ISDN subscription, and increased bandwidth will not always improve the service. Rather, bandwidth limitations will force us to think more intelligently about the data we are transmitting (see the text box Data compression for more on this).

Navigating Cyberspace

The flow of information through computer networks is in its infancy, but we are already far beyond the point where any human being can claim to have the overview of what resources are available. In order to lower the threshold for accessing information on the net, we need better tools for navigating cyberspace. The interface between humans and computers is by many seen as the bottleneck for taking advantage of modern technology, but the increased functionality of multimedia machines may help us widen the bandwidth of this interface. Virtual reality (VR) is one example of a technology that may help data navigation. Without knowing how to type, users are immersed in synthetically generated worlds where navigation is based on body movements. Head-mounted displays, headphones and motion sensors come closer to the user than traditional computers. Whether it will open new worlds or distort existing ones, virtual reality machines are here and the network will connect them. See figure 7.

Hypertext is another alluring technology that will influence the way we interact with computers. On paper, footnotes, references, and boxes break up the linear structure of text, and the table of contents provide pointers to information. In electronic documents, all these features can be implemented through hyperlinks. By selecting a highlighted text or picture, the computer will take you to the other end of the link. While hypertext documents of the 1980s referenced themselves in closed loops of canned information, some modern systems support network-transparent links.

The inhabitants of Cyberspace

Technology is the foundation of cyberspace. But, it is not only a network of computers - it is also a network of people. As will be stated more than once in this issue, the net is growing quickly. However, merely talking about growth hides some of the conflicts that will arise in the years to come. New groups of people are flocking to the net - some come for adventure, some to make money, and some to escape their past. This is not merely growth, it is immigration.

Hackademics

Up to now, people on the net have been a relatively homogeneous group of academics with an enthusiasm for computers - 'hackademics' (see the text box Hackers for an explanation of the first part of the term). In the 80s they found how useful a personal computer can be for word processing. In the 90s, the same people found how much more interesting computers become when linked in global networks.

To hackademics, information is sacred. Life is used to study information, turn it into knowledge - and convert it back into information by 'publishing'. For example, one writes a scientific article, submits it to a scientific journal and hopes to spread the word in scientific circles. There is seldom any monetary compensation for making information available, academic glamour is often the motivation. It is no coincidence that this group has taken the net into active use; here they find a medium through which they freely can access information and communicate with peers.

Among the new groups that will appear on the net in the years to come are the information merchants and the Nintendo generation.

Merchants

On the net, information has mostly been available free of charge up to now. The merchants come to peddle their information, which means that information comes at a price. Still, many argues, in order to increase the diversity and quality of information, commercial information vendors should be allowed access. Stewart Brand describes the tension that will appear:

The merchants are likely to be organisations that already own and distribute information. For them, the computer is just another medium, but one of increasing importance. They are likely to argue for using the same conventions and statues as cover paper-based publishing.

However, many argue that the characteristics of each medium also should influence the way we handle it:

Moving further away from the traditional view of the information merchants:

These views come from Bruce Sterling, an author who makes a living selling information.

Look at Information is Power.

The Nintendo generation

In a message posted on the net in 1991, Brewster Kahle describes how video games have captured the minds of a generation:

He goes on describing the potential of communication embedded in Nintendo games:

Convinced? Video games will move into the networks and perhaps even become a respectable service. Already, Americans spend more money on video games than they do going to the movies.

Postscript

When designing systems that have the potential of changing the way information is communicated in a society, one should realise the existence of some social responsibilities. The systems that are discussed in the following articles may change the way we work and play, and a minimum of the resources should be available for everyone to use. This responsibility should be undertaken by all groups involved in shaping the future, from programmers and information vendors to bandwidth providers and government.

Further reading

The articles that follow will further elaborate on topics discussed in this introduction. They are written by active researchers in the field, and provide reports from the trenches.

In addition to the articles, the following sources have been inspirational when writing this introduction, and they are hereby recommended:

Robert Metcalfe. Will networks be the bottleneck? Eurographics 92, State of the Art Reports, Cambridge, England, 1992.

All issues of the magazine Wired.

Stewart Brand. The media lab - inventing the future at MIT. New York, Viking Penguin, 1987

Whole Earth Review's special issue on Questioning technology, No. 71, winter 1991 (in it you will find Why multimedia publishing is a crock, by Tim Oren).

Steven Levy. Hackers: heroes of the computer revolution. Garden City, New York, Anchor Press/Doubleday, 1984.

Peter J Denning (ed.) Computers under attack. New York, ACM Press, 1990.

Ithiel de Soola Pool. Technologies of freedom. Cambridge, MA, Harvard University Press, 1983.

Time Magazine, September 27, 1993 (vol 142, no. 13), cover story: The attack of the video games.

New opportunities for publishers in the information services market. Commission of the European Communities, 1993.