Television networks and computer networks are almost the opposite of each other. A television network is a distribution hierarchy with a source (where the signal comes from) and many homogeneous sinks (where the signals go to).
Computer networks, on the other hand, are a lattice of heterogeneous processors, any one of which can act both as source and sink. The two are so totally different from each other that their designers don't even speak the same language. The rationale of the one is about as logical to the other as Islamic fundamentalism is to an Italian Catholic.
For example, when you send e-mail over the Internet the message is decomposed into packets and given headers with an address, and pieces are sent over a variety of different paths, through a variety of intermediate processors, which strip off and add other header information and then, quite magically, reorder and assemble the message at the other end. The reason that this works at all is that each packet has those bits-about-bits and each processor has the means to pull out information about the message from the message itself.
When video engineers approached digital television they took no lessons from computer network design. They ignored the flexibility of heterogeneous systems and information-packed headers. Instead, they argued among themselves about resolution, frame rate, aspect ratio, and interlace, rather than let those be variables. Broadcast TV doctrine has all the dogma of the analog world and is almost devoid of digital principles, like open architecture, scalability, and interoperability. This will change, but change has so far been very slow in coming.
The agent of change will be the Internet, both literally and as a model or metaphor. The Internet is interesting not only as a massive and pervasive global network but also as an example of something that has evolved with no apparent designer in charge, keeping its shape very much like the formation of a flock of ducks. Nobody is the boss, and all the pieces are so far scaling admirably.
Nobody knows how many people use the Internet, because, first of all, it is a network of networks. As of October 1994, more than forty-five thousand networks were part of the Internet. There were more than 4 million host processors (growing at more than 20 percent per quarter), but that is not a helpful measure for estimating the number of users. All that needs to happen is that one of those machines serves as a public gateway to, say, France's Minitel system, and all of a sudden you have an additional 8 million potential users on the Internet.
The state of Maryland offers the Internet to all of its residents, as does the city of Bologna, Italy. Obviously all these people don't use it, but in 1994, 20 million to 30 million people seemed to. My guess is that 1 billion people will be connected by the year 2000. This is based in part on the fact that the fastest-growing number of Internet hosts (percent change) in the third quarter of 1994 were Argentina, Iran, Peru, Egypt, the Philippines, the Russian Federation, Slovenia, and Indonesia (in that order). All showed more than 100 percent growth in that three-month period. The Internet, affectionately called the Net, is not North American any more. Thirty-five percent of the hosts are in the rest of the world, and that is the fast-growing part.
Even though I use the Internet every day of the year, people like me are considered wimps on the Net. I use it strictly for e-mail. More savvy users and those who have the time cruise around the Net like walking into and out of shops in a mall. You can literally move from machine to machine and do window-shopping using tools like Mosaic or riding bareback (closer to the metal). You can also join real-time discussion groups, so-called MUDs, a term coined in 1979 meaning "multi-user dungeons" (some people are embarrassed by the name and will claim that it means multi-user domains). A newer form of a MUD is a MOO (MUD object-oriented). In a very real sense, MUDs and MOOs are a "third" place, not home and not work. Some people today spend eight hours a day there.
In the year 2000 more people will be entertaining themselves on the Internet than by looking at what we call the networks today. The Internet will evolve beyond MUDs and MOOs (which sounds a bit too much like Woodstock in the 1960s here in the 1990s in digital form) and start to serve up a broader range of entertainment.
Internet Radio is certainly a bellwether for the future. But even Internet Radio is the tip of the iceberg, because it is, so far, not much more than narrowcasting to a special kind of computer hacker, as evidenced by one of its major talk shows, called "Geek of the Week."
The user community of the Internet will be in the mainstream of everyday life. Its demographics will look more and more like the demographics of the world itself. As both Minitel in France and Prodigy in the United States have learned, the single biggest application of networks is e-mail. The true value of a network is less about information and more about community. The information superhighway is more than a short cut to every book in the Library of Congress. It is creating a totally new, global social fabric.
We are looking forward to your suggestions for other editorial links or commercial links relevant to this section.
Post your comments Follow our "Being Digital" discussion forum
Read more of the electronic version of "Being Digital":
Next: Teaching Disabled Previous: Signals With A Sense Of Themselves