Ether where art thou

Came across a retrospective view on the development of Ethernet by Bob Metcalfe and associates at Xerox Parc in the 70s. For me, Ethernet was where I cut my teeth on technology. As a student in the Science Communications Masters program at Boston University in about 1982 I had to write a thesis in the form of a technology assessment and I chose, for some mere but serendipitous reasons [my friend and schoolmate Rich Mack suggested it], Ethernet. So this Economist story from last fall is of special interest. Yes, the Ether still seems to be a place that gets my attention.

My study led me to conclude that local-area networks would be a billion-dollar-a-year market by the end of the 80s. What do you know: They were! Looking back on my study recently I see that I overestimated IBM and its token ring baseband, and underestimated Ethernet [and standards-based networks] baseband; I did correctly foresee the problems broadband alternatives would have in this market, but I remembered that; it’s the errors I tend to forget. Hard to recall but in fact the oddman out in this survey was the PBX, which many thought would be the center point for future computer networking.

Only pulled it out to see if I had talked with Metcalfe during the course of my research. In fact it was another fellow [Steve Palmeroy] from his company [3Com] that I talked too. When I met Metcalf [working as a reporter] at a meeting of the Electronic Freedom Frontier Foundation in the mid-90s, when he was still W.C. Fields-like prognosticating the collapse of the Internet [someday he will be right], I dutifully thanked him for talking to me, though it wasn’t him, years ago. Penguins were chirping as we talked, and later as Esther Dyson made a speech, for the event was held at Boston’ Aquarium.

Anyway, during the Ethernet hunt were first-principle effects. For example: This was before Moore’s Law was so entrenched, and in conversation with a Mostek engineer I was walked through the types of volumes Digital Equipment Crop. was expecting for integrated circuits that embedded the better-part of the Ethernet protocol work. People familiar with these things know that an “X” pattern is formed as chip prices go down and unit volume goes up.

A magic moment happens and the technology takes off, and by that fact, takes off even more. I also learned from this project - well really from my professor, Kirt Olsen, who was teaching me how to do this technology analysis – that you can count yourself lucky knowing on or two things when you are filling in tables representing, for example, vendor market shares. Knowing one thing can help you deduce another, if you put your mind to it, and dont resist a little naive enthusiasm.

Last update: 4/4/2004; 8:46:52 PM.