Betting on Darwin

Doc Searls interviews Marc Andreessen and Tom Paquin on Netscape's Open Source Strategy.

Evolution is not a force but a process; not a cause but a law. It is not enough to do good; one must do it the right way. No man can climb out beyond the limitations of his own character. —John, Viscount Morley of Blackburn

Only two operating systems are growing in market share today. The one we all know is Microsoft Windows (both NT and 95). The one most of us don't know is Linux. After reading the mainstream press (or listening to politicians like Orrin Hatch or to professional haranguers like Ralph Nader and Gary Reback), it seems clear that Microsoft, that evil monopoly, is making Windows the requisite operating system for every computer-like device in the known universe.

But there is a constituency that won't let this happen. It's the growing community of hard-core technologists, an increasing number of whom now rely on Linux. A case could be made that the only technologists who aren't into Linux remain with another OS for practical or business reasons.

Today, Linux is the technologists' OS. It's a bad idea to bet against them. It's a good idea to bet with them. That's exactly what Netscape did when they released their source code to what everyone suddenly calls the Open Source community, but for years was the Free Software movement. Significantly, Linux is free. But more significantly, Linux's source code is open. Anybody can look at it, tweak it and share their tweaks with friends—which is exactly why Linux is so popular with techies. Quite literally, it's theirs. They made it.

“All the significant trends start with technologists,” Marc Andreessen says.

Lest we forget, the Internet was created by technologists, and its explosive growth is far more an expression of rampant hackery than of commercial activity, personal expression, massive archiving or whatever.

Consider this: over half of all the web pages on the Internet today are produced by free web software—mostly Apache—running on Linux operating systems. Windows NT is gaining, but not necessarily at Linux's expense. In fact, Linux is gaining too, mostly at the expense of commercial systems.

Even many commercial sites use Linux. According to a Commerce Department report released today, the volume of information being processed over the Web is doubling every sixty days. The mind boggles.

If even half of that growth is happening on Linux, the population of supportive techies is at critical mass or better. Netscape knows this, which is why they're betting the farm on Linux, just like they bet the farm on the Net four years earlier, when they changed the whole software business forever by freely releasing their first browser to anyone in the world with an FTP client, and by making development partners out of a ubiquitous user base.

In fact, the farm metaphor is an apt one, because releasing the browser source code on March 31 amounts to a bet that spring is here, the ground is tilled, the seeds will grow, and the crop will be abundant. It's probably not a coincidence that Marc Andreessen grew up on a farm.

He knew in his bones that releasing the Netscape browser source was the “Right Thing” to do, so did many other farmers out there in the hacker community. Chief among these was Eric Raymond, who earned some notoriety when he edited The New Hacker's Dictionary a couple of years ago, but whose fame really caught fire when his essay “The Cathedral and the Bazaar” ( caught the attention of the Netscape folks, and spread from there.

By Eric's metaphor, significant software was traditionally built like a cathedral, “carefully crafted by individual wizards or small bands of magicians working in splendid isolation, with no beta to be released before its time.” But a new model was brought to the world by Linus Torvalds, the creator of Linux: “release early and often, delegate everything you can, be open to the point of promiscuity.” The result was a vast community of techies that “seemed to resemble a great babbling bazaar of differing agendas and approaches (aptly symbolized by the Linux archive sites, which would take submissions from anyone), out of which a coherent and stable system could seemingly emerge only by a succession of miracles.”

The cathedral and bazaar metaphors seemed apt for the worlds that surrounded the code. The corporate world was formal and contained, while the open source world was informal and open.

Not that the bazaar lacks formalities of its own. (A detailed analysis of these formalities can be found in “Homesteading the Noosphere” by Eric Raymond, Every society has its protocols, and the hacker community is no exception. The open source world is a society of peers. “You're not a hacker until somebody else calls you one,” Eric says. What you get out of this society is not just great code written by the best programmers in the world, but massive peer review and an equally abundant source of support.

I found myself pulled into this world when (Linux Journal publisher) Phil Hughes told me I ought to be talking with Eric, since for years I've been a kind of Johnny Talkyseed, spreading the notion that markets are conversations, rather than battlefields. Phil thought the bazaar metaphor was at worst a corollary and at best another way of saying the same thing.

Over the next month, I got a lot of great hang-time with Eric (he was a house guest for a week). I also intercepted Jim Barksdale at an industry forum and turned that conversation into an interview with Netscape co-founder Marc Andreessen and's new captain, Tom Paquin. I attended the Silicon Valley User's Group (SVLUG) meeting where Marc and Tom together announced the open source move and addressed the Linux community on their thinking for the first time.

In brief, here's what Marc and Tom told that community:

  • The timing is right, because both open source and Linux energy have reached critical mass.

  • Releasing the browser source and joining the open source community is a return to roots, both for Netscape and for Marc personally.

  • There's an opportunity to catalyze open source development, not only in the Linux community, but inside two cathedrals: commercial software developers and their corporate customers.

  • If Sun doesn't open Java's source code, it'll happen anyway, outside Sun's cathedral. The open source community should work to bring Sun to a “more enlightened” place.

  • Sun, Oracle and other mainstream developers need to adopt Linux as a Tier One reference platform. Netscape already has.

  • Linux has a couple of problems its development community should address. One is scalability. The other is mission-criticalness. In time, it should be able to run on multi-processing servers and clusters and handle heavy-duty transaction processing, for example.

Marc also made some predictions:

  • Linux will get discovered by the mainstream press within a year.

  • Linux will consolidate the UNIX market and expand beyond it.

  • Netscape will bundle Mozilla with Linux.

  • Mozilla will become the GUI of the future—“the user environment where people can live and work on the Net.” Here it will naturally go “hand in hand, or hand in claw, with Linux.”

  • When mainstream vendors want to start adapting Linux to their particular hardware, they're going to run up against the Linux GPL. Something needs to be done about that, or Linux adoption will be slowed down.

A few days later I was booked for a half-hour interview with Marc and Tom at Netscape. The conversation quickly reached critical mass and ran for an hour and a half. What it comes down to, it's clear, is one line from Tom: “We're betting on Darwin here.” Let's hope he's right.


Doc Searls is the Editor in Chief of Linux Journal