It all started with the U.S. Department of Defense through its research arm, the Advanced Research Projects Agency, or ARPA (later, the acronym would be changed to “DARPA”). Under the increasing tensions of the Cold War and the stockpiling of weapons, the agency sought to create a national computer network that would continue to function even if a major portion of it were wiped out in a nuclear war.

According to Mills, ARPA was looking for a way to talk across three main communications technologies: landlines (ARPANet), satellites (SATNet, which linked the U.S. and Europe) and packet radio, a device that transmits and receives data via short-range radio. It was 1977.

“Networks were expensive central processors; we wanted to take a dumb terminal and access a smart computer mainframe. That’s how the Internet was born,” he recalled.

Mills served on several national panels and advisory groups. He chaired the Gateway Algorithms and Data Structures Task Force (GADS), the precursor to the Internet Engineering Task Force which develops and promotes Internet standards, and was the first chairman of the Internet Architecture Task Force.

A self-professed “Internet grease monkey,” he played an essential role in developing the gateways and protocols — the rules of the road of data transmission, if you will — that provide the backbone of today’s information superhighway. His specialty? Routing protocols.

“I wanted to send electronic mail, but in order to do that I had to write the software protocols. We wanted to teleconference rather than travel to meetings; we wanted to connect networks together, so we had to write the software,” he explained.

One well-known example is the Fuzzball router, an inexpensive mini-computer loaded with software written by Mills. Each Fuzzball was connected to SATNet, providing a network link that Mills and his colleagues used to research, experiment and test the protocols and technology they were developing.

“There were nests of Fuzzballs across the globe in the U.S., England, Germany and Italy,” Mills said. Later, they also would be used in early National Science Foundation (NSF) internetworks connecting university and research campuses to the Defense Department’s ARPANet.

Mills also actively participated in the evolution of the Internet protocol (IP), transmission control protocol (TCP), file transfer protocol (FTP), simple mail transfer protocol (SMTP), Telnet and other standard methods of data transmission upon which modern researchers and the world’s Internet users rely. On his watch were gateway architecture, network and internetwork routing algorithms, subdividing IP networks and “growing pains” — which often meant working all hours of the day and night, and traveling across the country and Europe to collaborate with others involved.

“If it didn’t work, the fickle finger would point to me,” he smiled, recalling many a night sprawled on his office floor, waiting for phone calls alerting him to the latest problem.

In particular, his Network Time Protocol (NTP) was essential to the early development of the ARPANet, precursor to the modern Internet. The protocol enables precise time synchronization, which is critical to the accuracy of electronic feats such as air traffic control coordination, online experiments, stock-market buy and sell orders and streaming video on the web.

UD a “virtual post office”

Meanwhile, Dave Farber arrived on campus in 1977 with an ARPANet Interface Message Processor (IMP), the predecessor of the modern-day network router, and an ARPANet connection. UD was among only a small number of universities that had an ARPANet connection.

In the early 1980s, researchers at academic institutions across the nation and overseas wanted to talk to one another. In a small computer science laboratory in DuPont Hall, Farber, now Distinguished Policy Fellow of Electrical and Computer Engineering, and a small cohort of students and faculty operated the first global “post office” for academic email.

“Delaware was the central hub in the early days of academic networking. You didn’t talk to anyone in the academic world without going through Delaware,” Farber explained.

UD set up mail servers using telephone lines and the ARPANet so that academic institutions across the nation could dial in to drop off or receive messages as often as they were willing to pay for phone calls. The result was CSNet — a computer science network designed to tie universities together.

Bob Caviness, then chair of computer and information systems, and Peter Warter, then chair of electrical and computer engineering, supported and encouraged the project, according to Farber.

CSNet worked on a relay system. A school would dial into the UD computer, establish a telephone connection and download all email desired to UD. Then UD faculty and students collected and delivered the mail from multiple institutions.

“My student, Dave Crocker, sometimes stayed up 24 hours a day managing the relay,” Farber remembered. When word got out, Farber and his colleagues were delivering email to 500 sites daily, including foreign countries. They quickly realized the critical value of electronic correspondence.

“Each time the system went down for repair, within minutes our phones were ringing off the hook with people from all over the world saying, ‘we sent you something and you haven’t replied, what’s wrong?’ ” he continued.

Farber’s team was also among the first to recognize privacy issues associated with email and develop tools to debug email without viewing the contents.

As students graduated and began working, industrial research laboratories also began clamoring for email access, Farber noted. Tech giants like Bell Laboratories, IBM and Hewlett Packard, among others, were willing to pay for connectivity, as were a number of academic institutions. Farber responded by extending the mission of CSNet to include industry, lowering the cost to academia.

The project involved 380 schools in the United States and 30 countries around the world. Early users overseas included Japan, China and Russia. South Africa, which recently celebrated two decades of having the Internet and acknowledged UD, was among the last to join as the Internet was coming into popular use in the late 1990s.

“Our role was to generate ideas on how email could work, set the direction and formulate the needed infrastructure; then train students who would carry on afterward,” said Farber.

As organizations began demanding quicker service, permanent links and phone lines, the Internet became commercialized and CSNet faded into the background, replaced by NSFNet, a collection of regional networks.

UD was part of SURANet which was built under the auspices of SURA, the Southeastern Universities Research Association. Networks in other regions, including NYSETNet in New York State, NEARNet in New England, JVNCNet in New Jersey, among others, were interconnected to form the national NSFNet.

At UD, Daniel Grim was an early proponent of Internet and Ethernet technology, promoting their use and adoption on campus.

“One requirement of participating in the national network was to deploy networking technology on the campus,” he said. Up until then, the only networking at UD connected the computing center in DuPont Hall to the electrical engineering and computer science departments.

Grim, who today serves as the University’ chief technology officer, chaired UD’s Network Planning Task Force charged with developing a plan to construct a campus-wide network. The task force delivered its report in 1989 and implemented a network that reached key locations around campus in the early 1990s. Today the network extends to every building on every campus across the state.

A shift in thinking

While Mills and his colleagues suspected early on that the Internet would be huge, he admits to periodically worrying that the Internet was growing too fast; that computers were not big enough and that lines were not fast enough to handle the volume.

“In retrospect, I needn’t have worried,” Mills chuckled. “Moore’s law states that computing power and memory double every 18 months. Astonishingly, this observation has remained true from the ’80s until now.”

In the mid-1990s, thinking surrounding the Internet began to shift. Networking people began to consider more than simply connecting the networks together; they pondered how to enrich the infrastructure and transfer the technology to public use.

The main challenge, Mills continued, was could they “reach the time when the sun never sets on the Internet protocol.” On February 8, 1994, they succeeded, when the South Pole received its first Internet connection.