On Mayday (May 1, 2000), a point when financial markets seemed to have lost faith in Linux as a Big Trend (many Linux stocks lost most of their value in the first third of the year), a big chunk of change — $37 million, to be exact — was invested in Lineo, Inc., which is emerging as the leading embedded Linux software company.
In an interesting twist, the list of sources for that money includes only three venture capital firms. Another fourteen investors are actual or potential Lineo customers, including familiar names like Motorola, Samsung, Misubishi, Compaq, Citrix and Acer — plus a raft of motherboard, laptop and component manufacturers in Taiwan, Japan and Korea. These include DaiShin Information and Communications, First International Computer, Global Alliance, Hikari Tsushin, Arima and Mitac International. The VCs are Egan Managed Capital, J&W Seligman and Astoria Capital Partners.
These manufacturers are Players. "There is a substantial interest on the part of major manufacturers in embedded Linux. And we include in that category a variety of software, hardware, components and solutions for embedded systems. These are smart companies that got to where they are by knowing how both to predict market trends and sense what's happening right now," says Lyle Ball, Vice President of Communications and Co-Founder of Lineo.
The most interesting aspect of this news is that it appears to be something unusual: a very traditional Old Economy play. These manufacturers want to put Linux in their products; not just to score a big run-up off a Lineo IPO (which, of course, they certainly wouldn't mind).
To get the significance of this, consider the little-discussed fact that every company has two markets: one for its goods and services and another for itself. Before the New Economy showed up, the latter market was extremely secondary, even for publicly traded companies. Value was All. Growth mattered, but there was no prevailing imperative to take a new company public or to run its value up to the sky overnight. But the get-big-quick imperative of the New Economy has biassed business conversations over the last several years, so that talk about investment has drowned out talk about the fundamentals of business. And this kind of talk has been endemic to the commercial Linux market ever since Red Hat went public last August and branded Linux instantly as the hot Growth Topic of 1999.
But this investment appears to be operating on the Old Economy imperative, which is to support product and service innovation. At least this is what a long and manufacturer-heavy list of investors suggests.
What this also suggests is that the Linux will probably expand from servers to appliances and other embedded devices more quickly than it will spread to clients — although there is no shortage of desktop and laptop manufacturers on this list of investors.
As Lineo sees it, the embedded Linux market shapes up this way:
Internet Infrastructure
Retail Business Systems
Transportation Systems
Home automation
Consumer devices
Industrial controls
"These are categories where we already have customers or expect to have them as demand spreads," Ball says. He expects demand to spread quickly because Linux's advantages lower the threshold of adoption well below competing operating systems, including familiar embedded OSes such as Wind River, ISI and QNX. These virtues include: open OS code source, an in-place worldwide support infrastructure, a huge developer community and base of knowledge, and relatively easy code adaptation across different processor types. "Linux is the Switzerland of internet connectivity and infrastructure," Ball adds. "You're not making NetWare Internet-savvy here. You're taking something that's extremely native to the Net and adapting it to a whole class of new Net-native devices. And you're doing it with the support of many thousands of Linux developers, all over the world, versus relatively few working for proprietary embedded OS companies. That's why our community is bypassing the development schedules that have determined embedded growth in the past, and accellerating a whole new schedule. We're putting the embedded Linux growth schedule on Internet Time."
Linus Torvalds seems to agree. He has been talking up embedded Linux for a while now, and his employer, Transmeta, has more or less the same intentions as Lineo.
"Now you can imagine a Coke machine as a Net-native connected device," Ball says. This might have been thinkable in the old embedded processing world, but it wasn't do-able, because there wasn't a Net-native embedded OS that was familiar to the very people who are doing the most to deploy the Net itself. Now that it's do-able, it will be very interesting to watch progress across Lineo's list of market categories.
— Doc Searls