Back

In this excerpt, the author discusses the challenges faced by the engineers at Honeywell and BBN in the early stages of building the first Interface Message Processor (IMP) for ARPA’s packet-switching network. The first IMP that was delivered to BBN had not been properly modified and wired, so Bob Barker had to fix it at BBN. However, even after fixing it, the IMP still experienced intermittent crashes. Barker suspected that the problem was in the machine’s timing chain and wired the central timing chain to fix the issue. This fixed the crashes and IMP Number One was almost ready for delivery.

The author then shifts focus to Vint Cerf and Steve Crocker, who were friends since high school and had a passion for science and mathematics. Cerf was known for his unique personality and his love for fantasy literature. He attended Stanford on a scholarship and majored in math, but soon became interested in computing and programming. After graduating, Cerf worked at IBM before joining UCLA’s computer science department, where Crocker was pursuing his graduate studies. Both Cerf and Crocker became interested in computer networking in 1968 and joined Len Kleinrock’s Network Measurement Center at UCLA.

Kleinrock was in charge of gathering data to evaluate the performance of the ARPA network and had a team of forty students, including Cerf and Crocker. They worked on building the host-to-IMP interface and the host-to-host protocol that would allow computers in the network to communicate with each other. Crocker, Cerf, and another important member of the team, Jon Postel, had different personalities and approaches to work. Cerf was more conservative, Crocker was the unofficial leader, and Postel, who had a checkered academic career, had strong opinions.

The UCLA team faced challenges with their Sigma-7 computer, which was unreliable and difficult to program. They were excited about the ARPA network because it would allow them to connect different host machines and explore various applications. They had to build the interface between the Sigma-7 and the IMP, as well as develop the host-to-host protocol, which needed to be versatile and universal.

Overall, this excerpt highlights the challenges and the dedication of the engineers and researchers involved in building the first IMP and developing the early stages of the ARPA network.

Words: 375