Linux for Suits

The best LinuxWorld Expos were the early ones. There were two in 1999 alone, both in the San Jose Convention Center. The second one, in August, had an official attendance of 14,278. It felt like ten times that many. My favorite memory of that show was sitting among thousands of geeks packed into a vast space where they could hear (though barely see) Linus Torvalds speak, hanging on every word as if Linus were Billy Graham calling the Faithful to a crusade. Never mind that Linus' whole schtick was the antithesis of box office, and that most of what he wanted to talk about—as always—was incremental progress on the Linux kernel.

Linux energy back then was like the electric charge that swells in hills below a gathering thundercloud. The high-tension wires that crossed the computing world sparked and glowed with vast anticipation of a world where the advantages of open over closed, free over captive, common over exclusive, were all as plain as day to the Faithful—but to few others.

Now the storm has passed, lightning has flown, and the world we expected is largely here. In GhandiCon (www.faqs.org/docs/jargon/G/GandhiCon.html) terms (first they ignore you, then they laugh at you, then they fight you, then you win), we've pretty much arrived at GhandiCon Four. Of course, the future is not evenly distributed. Desktop Linux, for example, has been arriving asymptotically for years. All we need to close that gap is one smart hardware OEM move, which has to happen eventually. (See UpFront for the Dell IdeaStorm story, which is very encouraging.)

Meanwhile, the device drivers keep piling up. If we spelled out the whole LAMP stack, it would be more than 145,000 letters long. Today, it's kinda hard to build “solutions” to anything requiring computing and Net connections and not to take advantage of so many free and open building materials. There is also a huge demand market for smart techies who not only know how to build with those materials, but how to improve them as well.

Yet the number of Linux queries (www.google.com/trends?q=linux) on Google has trended downward during the past two years. Although the news volume has held steady, the query volume today is about half what it was at the end of 2003. (That's as far back as Google goes, and it doesn't give precise numbers.)

That's why fighting for Linux today is like fighting for geology, botany or the periodic table. There may be some holdouts around less sensible paradigms, but what's the point? The Linux Revolution has become the Linux Establishment. We've won. Now what?

Good question. (That's what you say when you don't know the answer. Good question.) Here at Linux Journal, we like a good cause as much as the next magazine. And, we'd like to celebrate Linux's victory in exactly the way you'd expect any born fighter to behave: by looking for new fights.

Fights are naturally interesting. That's what story theory says. For a story you need only three elements: 1) a protagonist—somebody or something you care about and can identify with; 2) a problem against which the protagonist struggles; and 3) movement toward a resolution. You don't have a story if your protagonist isn't interesting, the problem is pointless, or if there's no movement toward an end state. That's why sports and war stories are so compelling.

So, what will our story, or stories, be? I'll suggest four and leave the rest up to you.

Linux and the Net have grown together ever since Apache became the standard Web server in the mid-1990s. Yet while Linux rocks on, the Net is becoming trapped in carrier silos. Net users today are no less trapped by their phone or cable companies than personal computer users in 1999 were trapped by Microsoft Windows.

The difference is that every carrier is its own Microsoft, every Net service is as crippled as Windows, and customer choice (in the US, at least) is between Tweedle-telco and Tweedle-cableco—or just one of those. These carriers still look relatively good to customers because the connection speeds they offer (labeled “broadband” or “high speed”) are many times higher than dial-up. It's too easy to forget that dial-up was what broke Net access wide open, making it available to nearly everybody—and did it in spite of the phone companies, rather than because of them. If it hadn't been for the original dial-up ISPs—The Little Garden, Panix, Batnet, Earthlink and even AOL—the Net still would belong only to universities, government and big business.

Now customers think the Net is gravy on top of their phone or cable TV services. They don't realize that the Net is the real base utility, and that it can carry any kind of gravy you like, including telephony and television. Almost nobody talks about all the businesses a wide-open Internet makes possible, mostly because the cablecos and telcos support consumption and discourage production. The “Net Neutrality” fight is a red herring. Most “high-speed Internet” customers have never experienced truly neutral service. Instead, they've enjoyed asymmetrical bandwidth and port blockages, without ever tasting what they've been missing.

Things are much better in some other parts of the world. Japan and Korea have notoriously high bandwidth at low prices, for example. The country with the highest broadband penetration is Denmark, with Estonia not far behind. However, all is not rosy there either. Networks may be fast in Korea, but Microsoft's market share in many categories, including desktops, verges on 100%. Broadband growth in Europe has recently slowed in regions (including Denmark) where incumbent carriers are making comebacks.

The big fight here is between independence and dependence, between citizens and monopolies (or duopolies), between local initiatives—backed in many cases by local governments—and some of the nastiest state and federal politics you're ever going to find. In every case, the protagonists are individuals, local groups, local companies, local governments and local utilities.

The problem they face is a combination of duopoly entrenchment and well-lobbied protection at the federal and state levels. The telcos alone are the biggest-spending lobbying group in US history—even bigger than the pharmaceuticals. And, they don't just work Congress. Some carriers are working at the state level to make it against the law for anybody to carry the Internet other than themselves.

Linux folks can help enormously here, because Linux techies—our readers—know how to build good, strong, reliable, easily fixed and easily improved solutions. And, they know how to do it on the cheap. We've been watering grass roots for up to two decades or more. Stallman taught us what freedom means, and Torvalds taught us how to have fun putting it to use. We have a lot of leverage.

Sometime this year there will be more than three billion mobile phones in the world. To put that in perspective (relying on last month's LJ Index), compare that to 1.4 billion credit cards, 1.3 billion land lines, 1.1 billion Net connections, 800 million cars, 200 million computer games, 100 million PVRs and 85 million iPods.

Cell phones are networked computing devices. A growing percentage of them run on Linux. Yet the OpenMoko (openmoko.com) and Trolltech's Qtopia Greenphone (www.trolltech.com/products/qtopia/greenphone), both wide-open working prototypes, are rarities. They face an enormous uphill battle against silo'd alliances between cell service carriers and equipment makers, such as Nokia and Motorola.

Yet the world needs open phones. In fact, I'd hazard a prophesy that open phones are inevitable, because there will be far more money to be made because of open phones than will ever be made with closed ones (and closed services offered only by carriers). We're starting to see vertical cracks in the closed wall of mobile telephony in settings such as universities, where rogue companies like Rave Wireless (disclosure: I consult them) provide students with custom (based on open) phones that run on familiar networks (such as Cingular and T-Mobile), but that do far more than the closed phones sold at stores by those same networks. Users are even free to do their own programming, create and add their own features and services. With each crack of this kind in a vertical market, the chance improves that open phones will become the norm rather than the exception.

The protagonists here are Linux techies, but working a much larger world of possibilities. The problem, as ever, is less a matter of closed systems than of the mentality behind it.

Once markets start to open up, it will be easy to fill whole magazine issues with stories of clever hacks and deployment successes.

Desktop Linux has been approaching without ever arriving since the mid-1990s. Most Linux Journal readers are there already, of course. But they're wizards. The muggles are still on Windows and Mac boxes. What we've needed for the duration is one or more of the major hardware OEMs to wake up and smell the volume. Last year, Lenovo began selling Linux-loaded ThinkPads in a committed way, but not aggressively. Lenovo didn't push it. This year, Dell set out bait in the form of IdeaStorm, a site that had all the look of a “conversational” marketing ploy, but instead served as a hole in the Windows-only dike that has been holding the Linux desktop river outside of Dell's headquarters for the duration. That hole quickly widened to a river of its own, flooding through Dell's product development system. (See the IdeaStorm story in this issue's UpFront.)

I'd love to be a fly on the wall when Michael Dell tells Steve Ballmer that Dell will be selling Linux-branded laptops and desktops in a much more public way, because the company has no choice: the market demands it.

HP, Sony and the rest won't be far behind. As the volume grows, so will the portfolio of applications and the sum of expertise about Linux desktops and laptops.

This is a category that will explode very quickly. I'm willing to bet right now that in June 2008, Linux Journal will have an unavoidably personal focus to every issue—for the simple reason that there will be too much going on with desktops and laptops.

Or maybe not. We don't know yet. Lenovo, HP and Dell may continue quietly to fill orders for desktop Linux without ever marketing it aggressively. This won't go on forever, but the asymptote may still stay flat for another year, two or three.

Meanwhile, desktop and laptop Linux are still worth fighting for, just like we've been doing for the last decade or more.

I have a confession to make. Or a Make to confess. I love Make magazine. I wish we'd done something like that first. Kudos to Dale Daugherty and the O'Reilly folks for pulling that one off and doing a great job with it—also for not running too much Linux-type stuff in there.

When I started with Linux Journal in the late 1990s, we were basically a how-to magazine. To a large degree, we still are. Most of our readers are hands-on types in any case. Problem solvers. Most of our writers (myself excluded) are too.

So I'm wondering...now that Linux is (or can be) in nearly everything, what can we make or fix that's one or more layers up? What can we do with MythTV that's beyond a set-top box? Pluto is a cool (and Linux-based) whole-home automation, security, entertainment and telecom system. But, it's still a system. A deep and under-appreciated (and under-deployed) virtue of openness is modularity. You want to be able to mix and match different stuff from different makers, including (especially) yourself. We should be making Legos with Linux, not just embedding it in finished closed products that work only with themselves.

Here the fight is for the right and ability to build what you want, any way you want to build it. Although Make is oriented toward doing fun hacks on already-made stuff (turning a mouse into a robot or adding temperature control to a coffeemaker), we'd angle more toward making the modules, and the things-with-modules that allow anybody to build anything. Our protagonists would be the same DIY-ers we've had all along, but the problem would be Building Anything. Fun problem.

Years ago, I talked about how the software industry was turning into a construction industry—when architects, designers, builders and their specialties would all be independent of any one company's platform or development environment. Now we're almost there, but not quite. The fight here is to make Linux and its endless variety of “stacks” into the base materials with which people can put together using their own virtual Home Depots.

Although it's easy to point to the exemplary successes of Linux-built giants such as Google and Amazon, it's just as easy to overlook the degree to which the practical value system behind Linux development has become the default approach to networked progress.

Yet even as Linux and the LAMP+ stack have become standard building materials, there's nothing to stop them from being used in service of a proprietary mentality that seeks to lock in customers, lock out competition and lock down markets. As Steven Hodson puts it (www.winextra.com/?p=354):

It's still customary for VCs to ask their potential portfolio companies, “What's your lock-in?” This is an Industrial Age mentality that needs to be exposed as a value-subtracting anachronism in a world where creation and choice yield abundances that can be put to countless productive uses. You should want to build goods and provide services that customers choose freely. You should keep customers because they want to stay, not because you've trapped them in a silo.

Even Steve Jobs this year came out and said the record industry would be better off without DRM. That's because he's no less trapped than any of his customers.

The protagonist here is nothing less than the cause of freedom, which will never be old Gnus. (Pun intended.) The problem here—the enemy—is a mentality that's as old as the Industrial Age.

The battle for freedom, of course, is one we've been fighting all along. The difference now is that the logic of lockup is more and more exposed, and its flaws are more and more evident—though not yet widely obvious.

The fight, then, will shift from ideals to practical matters. How do you make money by building with free stuff and putting it to use, rather than just by selling it? How is software more useful and important as it becomes less and less of an industry? How do you get more work done, and become more valuable as a contributor because you're working with free and open goods?

These are still new questions, even though Linux Journal has been a living answer to all of them since 1994.

So now the question goes to the floor. What are the Good Fights you want to read about in Linux Journal? You tell us. Write to info@linuxjournal.com.