OSI: The Internet That Wasn’t

How Quickly We Forget

History is written by the winners, as they say. And in the fast-moving world of technology, history can mean things that happened just 15 or 20 years ago. In “The Internet That Wasn’t,” in this issue, Andrew L. Russell, an assistant professor of history and director of the Program in Science & Technology Studies at Stevens Institute of Technology, in Hoboken, N.J., explores just such a case: an alternative scheme for computer networking that, despite years of effort by thousands of engineers, ultimately lost out to the Internet’s Transmission Control Protocol/Internet Protocol (TCP/IP) and is now all but forgotten.

Russell first wrote about the competition between that scheme, called Open Systems Interconnection (OSI), and the Internet in 2006, for the IEEE Annals of the History of Computing. During his research on the Internet and its precursor, the ARPANET, “OSI would creep up as a foil, something they didn’t want the Internet to turn into,” he says. “So that’s the way I presented it.”

After the article was published, he says, veterans of OSI “came out of the woodwork to tell their stories.” One of the e-mails was from a computer networking pioneer named John Day, who had worked on both TCP/IP and OSI. Day told Russell that his article hadn’t captured the full scope of the story.

“Nobody likes to hear that they got it wrong,” Russell recalls. “It took me a while to cool down.” Eventually, he talked to Day, who put him in touch with other OSI participants in the United States and France. Through those interviews and archival research at the Charles Babbage Institute, in Minnesota, a more balanced, complex history of networking emerged, which he describes in his upcoming book Open Standards and the Digital Age: History, Ideology, and Networks (Cambridge University Press).

“It’s almost alarming that something that recent can be so easily forgotten,” Russell says. On the other hand, it’s what makes being a historian of technology so rewarding.