SuitWatchFriday, December 14, 2001
A Linux model for consumer electronics
At its rollout event in January 2000, Transmeta showed off a bunch of prototypes expressing the embedded possibilities of Crusoe microprocessors. The most interesting one to me was a flat panel display in a frame that featured latching male and female USB connectors on all four sides, suggesting a build-it-yourself device that might be a DVD player, a camera, a camcorder, a computer, or all the above and then some.
One year later, at Macworld, Steve Jobs said there were three eras in the history of personal computers: the standalone PC, the networked PC, and what he called "the hub of your digital lifestyle."
And just this week a Slashdot story pointed to a DigitalTechnica piece that talks about "the expanding role of the computer as an all-encompassing home entertainment system."
There's a thread here, and the harbinger is in the middle item. After all, harbinging is what Steve Jobs and Apple have been doing for many years. Credit where due: Steve and/or Apple were largely responsible for driving the GUI, 3.5" floppies, laser printing, digital multimedia, SCSI, USB, FireWire and DVD-R, among other things. They were late to adopt Unix in a useful way, but they trumped Microsoft big time by putting Darwin under OS X, creating the first GUI desktop platform with an open source foundation and command line to boot.
Apple bears watching, but less for what they're up to than for where they're pointing -- which is toward a revolution in consumer electronics. This is a business with a long history of highly controlled, patented and branded innovations that flow largely out of giant Japanese corporations linked together not only through their own kieretsus and cartels, but also highly bound up with a distribution system assures maximum flow of boxes from a few manufacturers to billions of consumers.
We became somewhat acquainted with the nature of this business when DVD cartel failed to respect the existence of Linux, requiring linux geeks to independently provide their own DVD decryption with DeCSS.
Breaking the consumer electronics business free of its giants and cartels isn't a job for Apple, which at best will only enlarge the envelope of the existing system. And it's certainly not a job for Microsoft, although the Beast from Redmond is certainly a threat to Japan, Inc.
This job requires independent developers of all kinds, working outside the system.
I've got some history here. I've worked in consumer electronics retailing at various points in my life, and I've always been obsessed with electronic toys -- in spite of the annoying way those toys' makers make damn sure I become dependent on them exclusively for service, upgrades and peripherals (such as the annoying Memory Stick in my Sony digital camcorder). I've watched how these companies treat their markets more like corrals into which they can herd large populations of customers.
In the early 90s I also did a lot of work with the U.S. arm of Hitachi's semiconductor group. There, among other things, I helped market the SH microprocessor, now better known as the SuperH (on which there are GNU/Linux developments). The SH was originally developed to serve as the brains of Sega game machines, although it was also highly suitable for other purposes, such as controlling the brake systems in automobiles. (By one naming story, SH stood for Sonic the Hedgehog.) It was at Hitachi that I got my first up-close view of what consumer electronics marketing is really about.
For context, let's look at how marketing really works.
In school they say marketing is about making and selling stuff that customers need. Theodore Levitt, the father of the marketing concept, said the only purpose of marketing was "to satisfy the customer, no matter what." In other words, necessity should mother your inventions.
But in fact marketing also works the other way around, especially with technology. Invention is the mother of necessity. The best new inventions -- whether they're practical, fun, or both -- cause us to need them. We got along fine without copiers, cell phones, computers and digital cameras, but now we need them all.
Consumer electronics is largely about inventing necessities, and at Hitachi this begins at the HCRL: the Hitachi Central Research Laboratory. Much of the cream of Japan's technical Ph.D. crop goes to work at HCRL and its equivalents at NEC, Sony and Fujitsu, where they do amazing scientific work, much of which becomes patented, and some of which turns into products of almost baffling variety -- yet with a black-box sameness from one company to the next, and from one product to the next. (How many remote controls aren't black with gray lettering, even though they're almost impossible to read in the dark rooms where they're used?) Because these guys are also very good at following market trends by following each other. The final customer has approximately nothing to say about it. The feedback loop is barely there, even today with the Net saturating everything.
It is significant that the PC didn't come from any industrial giant's central patent pump. Not even IBM's. The Internet didn't come from that source, either. Nor did FireWire. Nor USB. Nor, most emphatically, Linux.
From a consumer electronics perspective, the PC has a problem: it's fundamentally generic. The Net has the same problem (or AOL and Microsoft wouldn't continue to make their instant messaging non-interoperable, in spite of the fact that they run on the Net). And so does Linux.
But that problem is a huge opportunity, especially for enterprising inventors who understand how big a market can be when you're not trying to own some of it by herding it into a corral. How can we put applied generica to work here? Is there an invention we can make that mothers some necessity?
Let's look at the drift toward this "lifestyle hub" business the PC finds itself in. Sure enough, the Net, USB, FireWire and simpler electronic connections are giving PCs enormous gravitational pull on consumer electronics peripherals.
Now, if the PC is truly is turning into a "hub," what's the next step? Apple and Microsoft both see themselves in the "digital lifestyle" hub business, and it's not hard to see where they're going with their strategies. My question is, are there any other strategies that open up when you start with Linux?
On the PC, Linux is a hermit crab OS. It likes to run on any box at all, no matter brand it is, or what other OS might have shipped on the box in the first place. But in embedded systems, Linux gets much more purposeful. It becomes a first-place strategy.
What would Linux' first-place strategy be in this "digital hub" business? Is there an opportunity here for somebody other than a Usual Suspect? Will somebody come through with a device that's all inputs, outputs, processing and convenience? That you can plug anything into and that plugs into anything? That mediates not just between recording and playback, but between one device and another, or between any hubbed peripheral and anything else out on the Net?
I think there is. But it won't happen if we wait for the Usual Suspects to make the next move.