Back

In the mid-1970s, the National Science Foundation (NSF) became a new potential source of funding for the computer science community. The NSF was interested in creating a network for the academic computer science community, as it saw the benefits of advanced communication and collaboration between geographically separated researchers. However, implementing this network proved to be a challenge.

In 1979, a group of computer scientists, led by Larry Landweber, proposed the creation of the Computer Science Research Network (CSNET). CSNET would be open to computer science researchers in academia, government, and industry, and would be run by a consortium of eleven universities. The network would use slower links and have a lower cost compared to the ARPANET. The proposal received enthusiastic praise, but it was rejected by the NSF after peer review.

Undeterred, Landweber and his team revised their proposal and presented a new plan in 1980. The new proposal involved a three-tiered structure using ARPANET, a commercial service provider like TELENET, and an email-only service called PhoneNet. The NSF agreed to support CSNET for a five-year startup period, after which it would be funded by user fees. The network would be managed by the University Corporation for Atmospheric Research (UCAR).

By June 1983, over seventy sites were connected to CSNET, and by the end of the five-year period, almost all computer science departments and many private research sites were connected. CSNET had become financially stable and self-sufficient. The success of CSNET paved the way for more NSF ventures in computer networking.

In 1985, the NSF created the NSFNET backbone, which provided a national network connecting supercomputer centers around the country. Regional networks were formed and given access to the NSFNET backbone, allowing for easy communication and collaboration between different networks. The funding for the regional networks came from member companies, while the NSF provided the backbone for free.

The battle between TCP/IP and OSI for network interconnection standards continued through the 1980s. TCP/IP, developed by Vint Cerf and others, had the advantage of being implemented and tested in real-world networks, while OSI was still a theoretical model. The popularity of UNIX, particularly the Berkeley version with TCP/IP, played a crucial role in the growth of the Internet. Ethernet, developed by Bob Metcalfe, also contributed to the expansion of local area networks and the demand for network interconnection.

In 1983, the ARPANET made its official transition to TCP/IP. The Defense Communications Agency split the network into two parts: the MILNET for nonclassified military information, and the ARPANET for the computer research community. The transition to TCP/IP was a significant milestone that allowed for the branching of the network and easy transmission of data between different networks.

In 1988, the ISO released standards for the OSI protocols, which were adopted by the U.S. Government as its official standard. However, the American culture of the Internet, built on TCP/IP, continued to grow. Ethernet and local area networks became increasingly popular, creating a large networking constituency. TCP/IP’s openness and flexibility, as well as its real-world implementation, made it the preferred choice for many computer scientists.

Overall, the development of computer networking in the 1980s was marked by the growth of various networks, the transition to TCP/IP, and the competition between TCP/IP and OSI for network standards. The success of CSNET and the establishment of the NSFNET backbone were significant milestones in the expansion of the Internet.

Words: 562