CONNECTICUT LIBRARIES
June 1997, Vol. 39, No. 6
CONNECTICUT LIBRARIES
June 1997, Vol. 39, No. 6
Where Wizards Stay Up Late: The Origins of the Internet
"Is this going to slow me down when I'm searching for biology statistics?" asked my college senior son. It was December 1995, and I had just tried to impress him with the World Wide Web and how libraries were going to use it.
Fancy typefaces and pictures on the Internet caught my son's attention. This was not the plain vanilla Internet he knew. He was relieved that his father finally had something more to say about online services besides his usual complaints about e-mail. But my son suspected that access to scientific resources might slow down when people like his dad started spending large amounts of time on the Net.
I doubt that my son ever experienced any really serious delays. But then, he was an undergrad, not a professional researcher. What I did hear on the radio and see in the newspapers were articles about Internet II and vBNS (very high speed Backbone Network System). The Internet had become so clogged with e-mail, business, and entertainment traffic that scientists were organizing their own private, "super-duper" network.
Where Wizards Stay Up Late (Simon & Schuster, 1996) makes it clear that the scientists and researchers only have themselves to blame for the delays. They created the Internet, used it, misused it, and evolved it into such a great thing that they couldn't keep you and me off it. This is a fun book, full of people with real personalities, people with incredible technical knowledge, people with dreams.
ARPA (the Department of Defense's Advanced Research Projects Agency) appreciated that computers greatly enhanced the research capability of scientists on university campuses across the nation, but it was frustrated with communications. Back in the 1960's, one of ARPA's offices was a room crammed with several types of terminals because no single terminal was compatible with the wide variety of computers used by ARPA scientists. And the number of computers was increasing since every researcher wanted his own! Remember, we are talking about the computers of three-plus decades ago: large main frames, very costly, proprietary in design, difficult to program, and expensive to maintain.
ARPA sought to control costs, end duplication, and encourage cooperation among its scientists by interconnecting incompatible computers over a reliable network. None existed, and the telecommunications authorities of the time were of no help. "Packet switching" is a key telecommunications process today. Thirty years ago the nation's telephone monopoly, AT&T, openly scoffed when ARPA proposed it. At a seminar for ARPA scientists, ninety-four AT&T experts patiently explained why packet switching would never work.
We do not always appreciate how appropriate the "highway" metaphor is for the Internet. Sometimes, we employ the "pipeline" metaphor; i.e., we envision sending a message roaring down a tube toward its destination. The pipeline metaphor breaks down because it assumes we are sending all of the information contained in the message down the same tube at the same time. Instead, what we actually do is more like shipping cargo long distance on trucks traveling the interstate highway system. We break down the message into "packets" of information and, figuratively, load them on trucks. We enclose destination and re-assembly instructions, and we dispatch the trucks on various routes in order to avoid detours and delays. Close down one highway and the dispatcher re-routes the "truck." As the trucks reach the destination, the packets are unloaded and the message put back together again.
The figurative interstate highway system of our metaphor is the distributed network of the Internet. Creating the system's "dispatcher" was a real challenge for ARPA. Research computers (hosts) at each campus were partnered with an Interface Message Processor. The IMP was a minicomputer wired and programmed to route packets of information over the network. A host sent a message to its nearby IMP, which then communicated (over the distributed network, using packet switching) with its fellow IMPs. They processed the message and forwarded it to their hosts. Once it was demonstrated that this system would work reliably, ARPA put its efforts into message congestion control, Telnet, and protocols for file transfer.
When the Internet was first introduced in the early 1970's, it was not an immediate hit. Many researchers saw it as a drain on their time and resources. Others believed that they had nothing to learn from their colleagues. Some were simply over-protective of their hard work. Interpersonal communication, not scientific cross-pollination, transformed many reluctant researchers into Internet enthusiasts. It was e-mail that won them over!
So, send your e-mail, join your listservs, and chat electronically. It is in the finest traditions of the Internet, and it almost certainly will continue with "Internet II: Return of the Nerds."
Return to Looking at Books Online