How The Internet Developed

In 1969 the Department of Defence in America funded the ARPANET. This was an experiment in reliable networking and a link between the department and its military contractors. The ARPANET quickly grew to span the American continent.

The experiment in reliable networking involved dynamic routing. The network had to keep functioning if one of its links was destroyed. Dynamic routing directed the traffic on a link to another link if the original link was damaged in any way.

Eventually the ARPANET was broken into two parts; one for the military and one for all non-military sites (eg universities, who couldn't get onto it fast enough). The two networks remained connected by a technical scheme called IP (Internet protocol). The IP was designed to allow for tens of thousands of networks. It views all computers on it as equally capable, so each computer on it can communicate with any other machine on the network.

About 1980 universities started to use individual workstations for their hundreds of users rather than having each user connected to a mainframe computer. The users wanted to keep the e-mail and shared file directories they had on the mainframe. Networking the computers did this. Networking the computers was possible because 1) most of the new workstations ran an operating system (UNIX) that included all the software necessary to connect the computers into a network and 2) the necessary network hardware was included by the workstation manufacturers. All the universities had to do was to use student labour to lay the cables to connect the computers. Thus they could get the network for a very low cost.

Every time a university connected to the ARPANET, its site would have hundreds of computers connected to the ARPANET instead of the old one or two users. As well, each new workstation was fast enough to generate enough network traffic to overwhelm the old ARPANET.

The National Science Foundation in America decided to set up five supercomputers for research use. Information would be sent to and from these computers using a network. Using the ARPANET didn't work out, so the National Science Foundation set up its own network - the NSFNET - to connect the supercomputers. This was a lot faster than the ARPANET. It then set up regional networks to connect users in each region with the NSFNET.

The supercomputer centres supported by the NSFNET either didn't work or were overpriced so that use of the NSFNET died out. However, the Internet use of the NSFNET worked very well. By 1990 so much business had moved from the ARPANET to the NSFNET that the ARPANET was shut down. By 1994, several large commercial Internet networks had grown up within the Internet. The NSFNET was wound down, and the people who used it now used the commercial networks instead. Some of these were run by large organisations (eg IBM) and some were run by specialist Internet companies (PSI, Alternet). These commercial networks can be used for any legal purpose, whereas the NSFNET could only be used for research or educational purposes. The commercial networks also connect to regional networks and provide direct connections for customers.

IP networks have appeared in many countries. Almost all countries are connected to other countries via one or more intermediate networks, so international communication is possible.