Deep perspective

Got an email this morning asking if my interest in geology had ever drawn me to Frenchman Mountain, over which the Sun rises, east of Las Vegas. Truth is, I hadn't noticed much more than the fact that Frenchman Mountain is there.

What's unusual about Frenchman Mountain, it turns out, is its exposure of The Great Unconformity, which is the boundary between the often pretty layered sandstones that make up so much of the West's dramatic landscapes — especially its brightly colored canyons and rock formations — and the much older and mostly dark basement rocks on which they began layering about 550 million years ago. The basement rocks are upwards of 1.7 billion years old. Even though in most expsures they were beveled almost flat by erosion long before the first extant sedimentary layers were deposited, these basement rocks are what remains of old mountain ranges about which relatively little is known. These are the same basement rocks exposed as Vishnu schist at the bottom of the Grand Canyon, and elevated as many of the mountain ranges and peaks of Colorado.

An unconformity is a zero where history used to be. It's many volumes of history that have long since been lost forever.

There used t be a program on TV called Mutual of Omaha's Wild Kingdom (woops, turns out there still is). In it the host, Marlon Perkins, would always compare something in wild nature to the benefits of Mutual of Omaha insurance. Just as the nutcrested bloodwrench protects her chicks from predators, Mutual of Omaha protects its clients from unfortunate but unpredicable events in everyday life... So, in the manner of Marlon, I'm about to use geology as a metaphor for computing history.

Because, while all of computing history is known to us, we sometimes act like it isn't. By which, I mean, we forget.

So I'll bring back to our collective attention two foundational developments in our industry that have largely been forgotten.

The first is when Ethernet father Bob Metcalfe convinced Xerox (which owned the Ethernet patent), Digital Equipment Corp. and Intel to release Ethernet to the world as an essentially free standard. (It was more complicated than that, I know, but let's not digress more than I am already.) This allowed Bob to make his business bones with 3Com, and for Ethernet to serve as the most useful and ubiquitous base networking protocol. What mattered here was the freeing of Ethernet, which allowed countless companies to make money selling stuff that used Ethernet, rather than by selling Ethernet itself. The unconformity here continues to reveal itself every time somebody asks how it's possible to make money selling something universally useful (the Internet, Linux, Python, USB, whatever) that's free.

The second is when Craig Burton and others at Novell saw the potential for selling networks as services, rather than as "pipes and protocols." That was the genius of Netware, which began by providing file and print services over anybody's hardware and software, finishing a job that Metcalfe started, which was discrediting the belief that the networked world needed to be a skyline of self-contained silos built on foundations of proprietary networks and operating systems that, like everything above that depended on them, barely interoperated with anything outside themselves.

I believe right now we're on the threshold of finally coming to a full understanding of A) why free stuff creates and sustains whole markets; and B) that what we call the Internet is actually a growing collection of ubiquitous available services, each based on an idea, a standard or a technology that is essentially free.

In a less wordy way, I'll use my guest time here at the Standard to talk about that, and point to other people and places that say all this stuff better than I.

Posted by dsearls, June 14, 2004 01:07 PM | | TrackBack