Tale of Two Stories

Google has achieved maximum Linux irony by becoming the only commercial enterprise to leverage enormous quantities of free software (10,000+ Linux servers at last count) into de facto web infrastructure: private enterprise as public utility.

Irony or not, Google is a major Linux success story. Even if it doesn't win every search popularity contest (last year, Jupiter MediaMetrix said MSN and Yahoo's search engines still had more "audience reach", although Google led by most other measures), its influence will only grow as the company earns more acclaim and success.

So what's wrong with Google? The press will be glad to tell you. Take two current examples:

"Google vs. Evil: The world's biggest, best-loved search engine owes its success to supreme technology and a simple rule: Don't be evil. Now the geek icon is finding that moral compromise is just the cost of doing big business", by Josh McHugh, in the January issue of Wired.

"A Nation of Voyeurs: How the Internet search engine Google is changing what we can find out about one another - and raising questions about whether we should", by Neil Swidey, in the February 2 edition of the Boston Globe.

One story is about purity, the other about privacy.

Both ask tough questions.

Both come to agreeable answers:

Both tell their stories well, but in the end, when all is said and done, well, a lot more is said than done. Such is the nature of covering an unrelenting success story. You still have to tell a story.

Listen to the managing editors, the assignment editors, the editors-in-chief at every publication covering news, and you'll hear one question repeated like a mantra: "What's the story?"

Swap "problem" for "story" and you get my point. Stories are about problems. Happily Ever After comes after a story is over, if it ever comes at all. Problems are what keep you turning pages. Problems are what make you tune in the latest report from the News Desk, Weather Central, Metro Traffic or the ball game.

Sports, in fact, exists entirely to satisfy the human need for stories. If there's no struggle--if teams and players don't give each other problems--it's not interesting. There's no story. If you're at a game and the home team is up by 30 points with a minute to go, your problem is getting out of the parking lot.

Back in the late 1970s I worked for a parapsychology research organization loosely attached to Duke University. One of my colleagues there was Jerry Solfvin, a deep guy given to asking profound questions. One day, while we were busy sawing through ceiling sheet-rock at the house we shared, Jerry paused to ask, "What is the fundamental unit of consciousness?"

Without even thinking, I replied, "It's the story."

At the time I was reading Writing to Sell, by Scott Meredith, the legendary literary agent. Meredith said every story has a "plot skeleton" with exactly three bones: a character with a problem, moving toward a resolution. What matters most, Meredith said, is keeping the character interesting and the problem complicated. The job of the resolution is to give sufficient cause for hope--and nothing more until it's time to end the book.

I was convinced at the time (and still am) that Meredith's plot skeleton also describes human interests in general. We are naturally interested in problems and complications. We work toward resolutions, sure; but as Steve Jobs famously said, the journey is the reward. Nothing's more dull than a problem-free life of infinite privilege.

Character, problems, complications. That's what keeps us interested in Google--even when the problems aren't that big and the issues aren't that complicated.

Now: What about Linux?

That's what I'm here to talk about.

See, Linux' stories are changing fast. The old World Domination meme is losing its irony. Think for a moment about the oft-quoted Ghandi line, printed onto countless geek t-shirts: First, they ignore you. Then they laugh at you. Then they fight you. Then you win.

Linux is winning. Even Microsoft is now treating Linux' success as a fait accompli. Linux' open-source development methods are cited in Microsoft's latest quarterly SEC filing as a "significant challenge to the company's business model". Microsoft also is being forced by governments around the world to open its source code to inspection, a sign that many highly placed hearts and minds have been swayed by arguments in favor of open source.

This doesn't mean Microsoft is defeated. It does mean, however, that it's not the enemy it was. The stories have changed; the problems are different now--so are the players. Red Hat and SuSE are still around, sure, but now the leading "Linux company" is IBM.

But what does "leading Linux company" mean when the real leaders aren't the usual suspects at all? They're the worker bees at companies of all sizes that are shifting IT dependencies from other operating systems and development methods to Linux and open source. They're IT departments embracing a de facto standard. And more, much more.

Yesterday I had a long talk on the phone with Avery Lyford of Linuxcare, which makes its living in the growing server consolidation business. He said the challenge for Linux right now is to cope with its own appeal, which fans out all over the place. Server consolidation uses virtualization to respond to the problem of server proliferation. "Now you've dissociated the unit of work from the box." There's the issue of "business criticality" and problem-areas such as "failure impact and disaster recovery", where Linux is proving quite helpful. There's procedural stuff like "development staging." On that last issue Avery said, "Every application goes through development, test, staging, and rollout. Linux, for the first time, allows you to switch architectures during that process. So you can develop for Linux on Intel, test on something else and deploy on a third physical architecture. Think about it. For the first time that's true."

And yet the change seems more mundane than profound. It's "best practices" stuff. Whatever business cliché you use, the effect is the same. Linux seems to be a vast magnet, quietly re-aligning ferrous atoms inside IT departments that continue to look the same.

Two years ago it was a big deal that Burlington Coat Factory had massively adopted Linux as its main platform. Now it's common to hear Linux mentioned in the same breath as GE, Goldman-Sachs, Home Depot, Boeing, LSI Logic, FedEx and Reynolds & Reynolds. It's not hard to imagine Linux becoming common IT infrastructure inside every large company whose name isn't Microsoft. And who's to say it won't happen there too?

All the companies listed in the last paragraph have different stories. What they have in common is an adoption pattern in which relatively little money moves to a "Linux vendor". For example, the flow of money to Oracle may stay constant, while the OS platform moves from HP-UX to Linux, while the hardware changes or stays the same. Internal applications that have nothing to do with any brand name in hardware or software also may move from some other OS to Linux without making any noise outside the company--or even inside.

"Use value" (as opposed to "sale value") applications are dark matter to everybody outside the teams doing the work. Changes in those applications don't call for press releases, no matter how profound they may be. Still, the adoption of Linux and open-source development tools and methods causes radical changes in IT infrastructure. There are real stories here.

Over in the desktop space, Linux is still a long way from being a serious threat to Microsoft. But that threat is bound to become real when the Linux desktop feature set and ease of use both match or surpass those of the latest Windows generation. We're still in the "Then they laugh at you" stage of that one, but at some point the laughing stops.

In The Innovator's Dilemma, Clayton Christensen describes how disruptive new technologies utterly transform their industries and cause great companies to fail. Those technologies are always easy to ignore or discredit at their early stages, because they appear too simple, too incapable, too cheap, too downscale, too unlike the stuff they will ultimately replace. But they quietly keep getting better until suddenly the market wakes up and smells the advantages. Then the market's path shifts from the incumbent leader's steady series of innovations to the new technology's course of rapid improvement and adoption. That's the path desktop Linux is on right now.

But this path may not be the one we're accustomed to seeing from inside the traditional Linux market space (where a number of leading vendors yesterday announced the Desktop Linux Consortium). In my recent conversations with executives from Lindows, it became clear to me that we're dealing with a different breed of animal here: a mass market company. While Lindows does some development, it's not driven by engineering. It's driven by a determination to put a new bottom end on the desktop PC operating systems market.

In Open Sources: Voices from the Open Source Revolution (O'Reilly, 1999), Bob Young of Red Hat spoke like a mass marketer when he said, "The Linux OS gives consumers choice over the technology that comes with their computers at the operating system level... Will that user prefer to go back to the old model of being forced to trust his proprietary binary-only OS supplier once he has experienced the choice and freedom of the new model? Not likely." Four years later, that choice is available from Wal-Mart.com in Lindows, Lycoris and Mandrake flavors. No Red Hat. The company driving a wedge through that channel, and the only one selling a $199 PC, is Lindows. I've also heard reports that Lindows out-sells the others at Wal-Mart.com. If so, it's a controlled study in branding.

Of course, Lindows may not be the ultimate desktop Linux winner; but what's interesting right now is that Lindows wants to make a market in ways none of the old "Linux companies" are ready to do: by shifting the conversation from Linux and its development methods to the kind of stuff "consumers" care about, such as brand, features and prices. Yes, Red Hat has ambitions in that space as well, but is Red Hat ready to play a disruptive role in the desktop OS market? Should it? I don't know. I just know there are a bunch of new stories here.

Same with the embedded space, which is at the "Then you win" stage. Last year it was a big deal that Sharp was putting Linux on the Zaurus PDA. Now Sony and Matsushita (which makes Panasonic and other brands) say they're codeveloping their own Linux distro for the whole of consumer electronics. Hitachi, IBM, Samsung, Sharp and Philips are also on board with the effort. This looks like a Game Over scenario, but there are bound to be plenty of problems and complications along the way. (Laptops, for instance. If these guys are so hot on Linux, how about giving us a real Linux laptop?)

All of which means that Linux will have countless real problems to write about. The hard job will be finding problems that don't announce themselves or that old story templates can't possibly describe.

You can bet the mainstream press will be covering Linux' success the way they're covering Google's today, focusing on single "on the other hand" topics. That's fine. But we're not here to do that. We're here to talk about what's actually happening. There's a lot of it going on, and we're counting on you to help us all find out more about it.

Doc Searls is senior editor of Linux Journal. His monthly column in the magazine is Linux For Suits, and his biweekly newsletter is SuitWatch.