At the recent Embedded Systems Conference in San Francisco, I attended a meeting/press conference held by the Embedded Linux Consortium (ELC). The stated goal of the Consortium is to produce a specification for a standard platform for embedded Linux. That was also the stated goal of the Consortium last April when it met at ESC West. In the past year, very little progress has been made in producing an actual specification. It appears that the ELC is struggling with a number of issues, but the one that has not been addressed is, “Is there really a point to having a standard platform for embedded Linux?”

What the ELC has produced in the past year is an organizational structure, committees, a white paper, a FAQ and a Core Platform Working Group along with a hotly contested intellectual property agreement (IPA), but no spec. Some idea does appear to be gelling as to what this platform spec should look like but from this writer's perspective time and tide seem to be running against the effort.

One factor is time. Today, there is no such thing as a standard embedded Linux platform and we have no way of knowing exactly when there will be. The clock is ticking and the world and the industry are moving forward.

It took a year to put together an IPA that members could live with and now, according to the ELC white paper, the stated intention of the platform is to, “bring together the best of the Linux Standard Base with relevant portions of the POSIX specifications and the Single UNIX specification as a starting point.” There are, however, a number of issues facing the working group including safety, wireless, peripheral interfaces, native real time and Web services to name a few.

Given this laundry list of concerns, it seems that a standard embedded Linux platform will not be able to address deeply embedded applications such as motion control, medical instrumentation, industrial automation/control and real-time control due to the specialized demands of these applications. That is not to say that many of these applications are inappropriate for embedded Linux, but they will not be supported by the platform with its standard APIs and functions. So what application areas does this leave for the standard embedded platform?

Well, the applications that come to mind are things like set-top boxes, PDAs, telematics systems and consumer devices. In fact, these are the kinds of applications mentioned by Consortium members. Indeed, the proposed platform might very well be appropriate and attractive to developers building such devices. There is only one small problem. Who will be the major competitor for design wins in these areas? That's right, class — Microsoft's Windows CE.

Now if you're going to pick a fight, it doesn't seem very wise to start with the 800-pound gorilla. Granted, Microsoft has had a talent for snatching defeat from the jaws of victory in the embedded market, but it also has a habit of trying until it succeeds. That's the advantage of deep pockets. For one thing, building a high-volume device like a set-top box is about a whole lot more than just the embedded operating system, it's mainly about time-to-market, functionality and upgradeability. This means reference designs, design services, tools and support. Except for support, which is the cause celebre of embedded Linux vendors, there is very little that can compete against Microsoft and its CE partners in the other areas. The Wind Rivers, QNXs and OSEs of the world haven't given up on these application areas either. That's the tide that's running against the standard embedded Linux platform.

CE has won in areas where it connects seamlessly with Windows NT-based servers, but along comes something known as carrier grade Linux, which is being supported by another consortium formed around Open Source Development Lab (OSDL). This effort is dedicated to data center and carrier class Linux and one of its big backers is IBM, which has invested huge amounts in Linux. A carrier class Linux with such backing really does have the potential to push Windows NT and XP out of large parts of the market, and this would offer an embedded Linux platform an opportunity similar to that enjoyed by CE — or CE.Net — in the NT world. Ah, but who will survive among the standard platform vendors?

The answer is: one or two companies. Once you standardize on a platform specification, you are selling a commodity and the vendors will go at each other's throats like Dell, Compaq and Gateway in the commodity PC market. IBM belongs to both consortia, as do several other ELC members. If IBM sells both carrier grade and standard platform embedded Linux, how will the smaller companies compete in that market?

The siren song of Linux has lured many brave hearts to sea with a promise of treasure that is truly there. But many brave hearts will sleep in the deep and only a few will bear home the prize.

Tom Williams is Editor-in-Chief of RTC Magazine.

This column, which originally appeared in the April 2002 edition of RTC Magazine, is copyright © 2002, The RTC Group. Reproduced by LinuxDevices.com with permission of The RTC Group.

 
This article was originally published on LinuxDevices.com and has been donated to the open source community by QuinStreet Inc. Please visit LinuxToday.com for up-to-date news and articles about Linux and open source.