The year 2000 saw the dot-com bubble pop. The myth of a “new economy” predicated on growth rather than profits died as well. Me, I thought we had learned that lesson from Netscape years earlier. Traditional economic influences once again proved their resiliency. Explosive growth of market share does not necessarily translate into profits, and without profits all businesses ultimately fail.
The embedded world, too, had jumped into the growth-over-profits mania. A gaggle of startups promoted Linux as the solution to all of our woes. Most of those companies are still around, somewhat spared the dot-com collapse, though almost all felt its impact and conducted layoffs to appease the banks.
Is the embedded world headed for it's own version of the dot-com crash? How many of these companies have achieved profitability? Most are privately held and don't disclose results, though the spates of layoffs are telling. Perhaps their troubles are the shared woes of the economic downturn. But, then again, maybe not.
Red Hat makes an interesting case study because it's publicly held. The company announced that it has cut its embedded staff, which is charged with pushing Linux and eCos to our world.
Don't get me wrong — I think Linux is a marvelous solution to some problems, and find eCos one of the coolest RTOSes around. It's also a work of art, stunning in it's clean design and almost poetic coding. But cool does not always translate into profits.
Red Hat acquired its embedded operation in 1999 by buying Cygnus for $674 million. Not a bad price for a company that, it's believed, never made a dime. That's not much less than the current market capitalization of the much, much, larger Wind River — a company that is usually quite profitable. Irrational exuberance indeed.
Look at the numbers. Red Hat has reported only a single year of profitability, in 1996 when the company made a half million dollars. Since then its losses have increased every year. In 2001 the company lost $87m on $81m in sales. The 2002 fiscal year was much worse, resulting in a loss of $140m on $79m in sales. The stock reached a high of $150/share in 2000 but lately hovers around $6.
Does this pattern sound familiar? I'd love to know the figures for the other purveyors of open source products to the embedded community. My e-mail in-box gets lots of scary rumors, but nothing substantiated.
The open-source business model is really fascinating. Lest one think it's an idealistic notion consider this quote from Red Hat's web site: “While it is true that Open Source software code is free, taking advantage of its benefits requires a significant investment.” That sounds like a good business proposition, a nice way to make money from a free product.
But the numbers, at least for the biggest of open source companies, suggest otherwise.
Jack G. Ganssle is a lecturer and consultant on embedded development issues. He conducts seminars on embedded systems and helps companies with their embedded challenges. He founded two companies specializing in embedded systems. Contact him at . His website is .
However good Red Hat's open source business model is, it still takes good business management to realize the profits.
In an alternative case of open source, there is Jean LaBrosse's uCOS operating system, which I have used successfully. I presume that LaBrosse is realizing a respectable profit from uCOS, although it comes from his classes, writings, and book sales, rather than from sale of the code itself. Embedded Linux vendors who manage to emulate LaBrosse's operations model, not just the open source business model, may be successful.
Can't wait to see the fanatical responses from the religous zealots grasping their sixties' “free-love” philosophies!!! Peace man! Groovy. Like, why doesn't socialism work man? Can you dig it?