I'm guessing it was twenty-five years ago, maybe more, that I witnessed the first of countless operanting system arguments. This one was, "Which is better:" Unix or CPM?

History, one might think, has proven the argrument ridiculous. Yet today we live in a world where the majority of servers are derivatives of Unix and the majority desktop is a derivative of CPM. Perhaps in time those will derive from Unix as well (that's still my long-term bet); but alas we still live in the meantime.

Still, taking the Long View (for which, at fifty-seven, I'm more qualified than most), there are a number of verities that I believe are proven out, or well on their way. So, this Thanksgiving Day, I thought it would be fun to share a few of them with you. Some of them I've all but beaten to death. Others are new. But since some build on others, I'll risk repetition. Here goes....

1) Linux is what happened when Demand supplied itself. So is every other free software open source project, (plus many a standard) that now serves as commodity infrastructure.

This is something you cannot see if you're busy, like most editors and the analysts they quote, looking at markets in from the supply side, "driving" demand.

Demand Supplying Itself is the most important economic fact in computing and communications. Alan Kay said "The best way to predict the future is to invent it." Kay, of course, was a technologist. Likewise, Marc Andreessen said "Technology trends start with technologists." The same holds for economic and social trends that are enabled by technology -- especially the networked technology built by the demand side for its own purposes, now leveraged in countless directions for other purposes.

Consider this. Most of the industry in music -- the effort that goes into making, performing and listening to it -- isn't by (what we still call) the music indusrtry. It's by individuals, working alone and together. Networked, they will completely undermine, reform and replace the music industry we inherited from the late Age of Industry. Broadcasting too. Count on it. (But feel free to help history along. I want to see this one happen in my lifetime.)

2) The computer industry is turning into a construction industry.

It's not just coincidence that we talk about building, architecture, design, platforms, structures, construction, sites, locations and addresses. Construction is a good working metaphor for what we do with software -- and hardware too -- or we wouldn't borrow so much of our vocabulary from it.

3) Commidities are Good Things. So is Commoditization. Thank your free software and open source creators for both.

Two questions.

First, Why do commercial software makers whine about the "threat" of "commoditization" (while analysts and press, mostly paid for by vendors, whine along in harmony)?

Second, why has nobody ever whined about the "threat" of commodities, such as those that give us food and shelter, that we can't live without?

The answer to the first is that the software industry -- what we've known of it, so far -- is still in its infancy, and hasn't learned to get along in a world where people learn to make good money in businesses commodities make possible. Some of that money is in selling commodities. A lot more of it is in selling and servicing the things commodities make possible.

This knowledge is a ten thousand year old hat in construction. It's brand new in software. Be gentle with it.

4) The world is filling up with open source and free software commodities. Without those commodities we wouldn't be able to create a thousand new industries and categories.

The list goes far beyond LAMP. In fact, it's moving toward endless. This equips us to make all kinds of stuff all kinds of ways, and to develop all kinds of expertise, at building, running and servicing all kinds of stuff that goes far beyond the technologies alone.

Think, for example, what will happen when the explosive new phenomenon we call "podcasting" replace one of the primary roles broadcasting has played as a companion while we drive around in cars, fly in airplanes, ride in trains and make our coffee in the morning. That role is Telling Us Stuff We Want to Know.

My pal Buzz Bruggeman called me this morning from somewhere in Arkansas. He's driving across the country, from Orlando to Portland, listening to podcasts of The Gillmor Gang (of which I'm a member). It occurred to him, as he drove along, that he was also interested in podcasts that tell him more about each of the places he passes, right down to the Funkyville Foursquare Gospel Church and the Red State University. He suggests those should be available for the downloading, possibly as a service yet to be offered (are you listening, Sirius and XM?) on satellite radio. Buzz isn't a technologist. He's a lawyer who knows what technologists can do. Which is why he called me, and why I'm passing his ideas along to the rest of you.

5) Openness matters. Modularity matters more.

I'll thank Esther Dyson for this one. I was writing about DIY-IT (Do It Yourself IT) for her newsletter, Release 1.0 (the May 2004 issue), going on about the good model provided by the construction industry, when she observed that the Larger Virtue of openness, its practical context, is modularity.

Modularity is what makes stuff work together. It's also what contains, even quarantines, intellectual property claims, keeping them from infecting whatever it is we're building.

Consider the case of a door. In most houses, in most countries, doors come in standard sizes. They also open and close in standard ways. Doorknobs and latches go in standard places along one edge of the door.

The basic door is a commodity, along with the wood (or glass, or plastic, or metal) used to make it. So let's take your basic wood door, like you'd buy down at Home Depot. You can put whatever doorknob and latch assembly you want in that door. Those items may in fact be full of intellectual property. The brand name may be trademarked. The mechanics may be patented. But there's nothing about any of that intellectual property that tells you what kind of door you use, or how ou build the rest of your house.

Likewise, Georgia-Pacific, Weherhaeuser, Dow-Corning and Johns-Manville construction materials are not "platforms" on which you are forced into dependency, and required purchases that work only with products made by those companies.

The construction industry doesn't suffer the form of vanity (common still in computing and other electronics industries) of what we'll call "platformitis". That's one reason why...

The deepest virtue of computing, as with economics, governance and many other instutions (even marriage) is independence.

Nobody likes other people, or other companies, or other industries, telling them what they can and cannot use. Liberty and independence are graces of enviornments where an abundance of commodities deliver us to what John Updike calls "the age of full convenience".

Infinite liberty and independence is an oxymoron, of course. There are natural limits. Civilization is a web of dependencies and interdependencies.

But basically, we all want opt-in, not opt-out. And least of all, the absence of choice. We want our choices to be voluntary. And we'll go out of our way, some of us, to raise the level and number and range of voluntary choices.

The best businesses are independent ones.

For years Dave Winer has had an argument with open source, even as he's helped fill the world with open source software (most recently, Frontier), open standards (RSS, SOAP, XML-RPC) and open-ended categories (outlining, blogging, podcasting). What he;s been arguing for is independent software developers. Lots of those developers have come up with commercial software without which the world would be a worse place. Such was the case with outlining, blogging, presentation software (which, for my money, began with MORE) and lots of other categories that have since been commoditized by open source alternatives.

Business categories are ecosystems. They are full of dependencies and interdependencies, as well as independencies. Now that we are entering an age of commodity surfeit in software, we can begin to discover how independent businesses can make money in independent ways with independent customers. And where most or all of the dependencies and interdependencies we choose are voluntary.

I'm no economist. But I'm old enough, and I've seen enough, to know that we're finally seeing a lot of industries, businesses and categories finally grow up. And thanks to that fact, we'll see countless new ones grow out of the rich soils fertilized by the old ones. We need to learn what that ecosystem is about, even as we continue to create it.

And while we're at it, thank the technologists for giving it to us.