Computerworld, June 21, 1999, p. 96 -- Technology Flashback (1973)
by Mary Brandel
Personal computers hadn't even hit the mass market when researchers started working on what would later prove to be the second phase of the PC revolution: linking these machines to a network. 1977 is widely seen as the year of the PC's big arrival, but Ethernet - the technology that today ties tens of millions of PCs into LANs - was invented four years earlier, in the spring of 1973.
The source of this foresight was Xerox Corp.'s Palo Alto Research Center (Parc). In 1972, Parc researchers were working on both a prototype of the Alto computer [Flashback, June 14] - a personal workstation with a graphical user interface - and a page-per-second laser printer. The plan was for all Parc employees to have computers and to connect all of the computers to the laser printer.
The task of creating the network was assigned to Bob Metcalfe, an MIT graduate who had joined Xerox that year as the self-described "networking guy." As Metcalfe says, the two novel requirements of this network were that it had to be very fast to accommodate the laser printer and it had to connect hundreds of computers.
Metcalfe's previous experience as a student gave him a great head start in networking. "My Harvard Ph.D. dissertation, based on my research at MIT, was about the Arpanet and the Alohanet," which was a packet radio network at the University of Hawaii, he says.
By the end of 1972, Metcalfe and a number of other Parc researchers had completed an experimental 3M bit/sec. PC LAN. The following year, Metcalfe defined the general principles of what he called Ethernet, the technology that made the first PC LAN possible. That same year marked the birth of the first Ethernet board, which you could place in a PC to create a network.
Ethernet defines the wires-and-chips aspects of PC net-working as well as the software aspects of how data is transmitted. A key concept is its system of collusion-detection and recovery, called CSMA/CD, or Carrier Sense Multiple Access with Collision Detect.
With this protocol, devices transmit only after finding the data channel clear. If two devices transmit simultaneously, causing a collision, they delay their retransmissions for a random length of time.
Metcalfe first called the network the "Alto Aloha Network." When he changed the name, he based it on the idea of "lumeniferous ether," the medium that scientists once thought carried electromagnetic waves through space.
Xerox produced hundreds of Ethernet boards that it used internally to access the lab's central minicomputer, access the Arpanet, send and receive e-mail, play games and share files, Metcalfe says.
It wasn't until 1979 that momentum gathered for Ethernet to become a widely agreed upon 10M bit/sec. commercial standard. At that point, Metcalfe had left Xerox and was meeting with Gordon Bell at Digital Equipment Corp. about helping Digital create its own LAN. Instead of coming up with something new, they decided to propose to Xerox that the two companies work together to make Ethernet a standard.
By June, Metcalfe had gotten Intel Corp., Digital and Xerox to agree to work on using Ethernet as the standard way of sending packets in a PC network. "Only then did it dawn on me that Ethernet would be a successful standard worth building a company around," Metcalfe says.
Thus 3Com Corp. was born. Now a $54 billion company, 3Com introduced its first product, EtherLink - the first PC Ethernet network interface card, in 1982. Early 3Com customers included Trans-America Corp. and the White House.
Ethernet gained popularity in 1983 and was soon named an international standard by the Institute of Electrical and Electronics Engineers Inc. But one major computer force did not get on board: IBM, which developed a very different LAN mechanism called Token Ring. Despite IBM's resistance, Ethernet went on to become the most widely installed technology for creating LANs. Today there's Fast Ethernet, which runs at 100M bit/sec., and Gigabit Ethernet, which promises 1G bit/sec. rates.
Looking back, Metcalfe says he would have used different methods to convince IBM of Ethernet's merits.
"I would have recruited major IBM customers to help me demand it from IBM," he says. "And it didn't help that I routinely attacked IBM as being a slow-moving monopoly - much as I now attack telephone companies and Microsoft - so I have not learned my lesson entirely." ------------- Brandel is a frequent contributor to Computerworld. Contact her at <email@example.com>. **************************************************
Jerry P. Becker Dept. of Curriculum & Instruction Southern Illinois University Carbondale, IL 62901-4610 USA Fax: (618) 453-4244 Phone: (618) 453-4241 (office) (618) 457-8903 (home) E-mail: firstname.lastname@example.org