Allen, would you be so kind as to assess for us the negative or positive impact of real-time Linux to WIND's future . . .? I am STILL unconvinced that RT Linux will not have an impact upon our great company . . .
Khan also asked me about the Linux threat. I didn't answer before because many other issues are of greater importance. Since it looks like Linux concerns will hang around for awhile, I'll tell you what I think.
I suspect the availability of a quality OS with source code will be attractive to many embedded systems shops that stubbornly insist on using their own home-grown OS. Exactly how Linux might become a serious mainstream alternative escapes me, but I don't deny that it could happen. The main point is that it simply doesn't matter, one way or the other.
Java is almost a perfect metaphor for Linux. Java landed in Internet land with great fanfare, driving up the stock price of any company including it in their name. You may not recall, but embedded Java threatened all RTOS vendors. Sun was making special Java silicon, as well as a JavaOS, with the announced expectation that the write-once, run-anywhere feature of Java would propel SUNW to total dominance in the embedded world.
Now, four years later, where is Java? Actually, it is about where reasonable people thought it might be by now. There are a few hundred embedded Java designs, many of which may prove successful. Most of the anticipated problems with Java for embedded use have been ironed out. Of course, the total number of embedded designs during this period measure over ten thousand, so Java has failed to live up to its hype, and probably never will.
Oh, there is one other thing about embedded Java. SUNW didn't end up with most of the embedded designs. WIND did. In fact, with the merger, WIND must account for 75% or 80% of all embedded Java designs.
Linux is Java by another name. It probably will be important in the embedded world, but much more slowly than advocates expect today. And when and if it does succeed, it will be subsumed by the dominant embedded player, just like Java. This is what WIND says officially; this is what happened in a similar situation (Java); and this is the only thing that makes any sense.
As for the free part, it doesn't threaten royalties as many fear. I can't present all the reasoning here, but suffice it to say that the value-add that WIND brings to the table generally extends well beyond the OS. From an economic point of view, OS royalty is just one mechanism for converting value-add to revenues. At this point in the maturity of embedded systems, the operating system per se is NOT the major contributor to value-add. This is why WIND is happy to subsume Linux or anything else demanded by the market place-with no expected negative impact on its business model. (If you find this terribly confusing, I will try to explain it better.)
As I said before, the only thing I found scary about Linux was the market caps of Linux IPOs. With the hot air out of the Linux balloon, the threat has subsided.
Allen |
