WIND is an enabler for a growing trend toward concentrating on "core" and outsourcing "context" (phraseology from Geoffery Moore's "Living on the Fault Line")
The casualties among those companies will likely be tho ones who sacrifice time-to-market and precious R&D resources integrating, qualifying, and tweaking "open source" software in a misguided attempt to prevent "lock in" and save royalty payments while their more nimble (and wise) competitors just leverage WIND's products and pay WIND the toll.
Am I on the right track here?
You got it!
WIND is the integrator, the packager, the enabler of connectivity in the new era of computing and communications. But there is an important aspect to WIND’s central role that needs to be understood.
As everyone knows, I have been a long-time fan of QCOM. I believe QCOM pulled off one of the greatest miracles of the 20th Century when it manipulated CDMA into becoming a significant wireless technology. Once it was clear to me that QCOM had succeeded in embedding CDMA in the wireless world, BUT NOT BEFORE, I really liked the company. Why? I liked QCOM because it owns a successful proprietary standard in an explosive market.
WIND is the opposite of QCOM. WIND is not trying to force the world into its proprietary mold. WIND provides everything needed to connect all the acronyms on any proprietary hardware platform. This means that all vendors with proprietary technologies must connect to WIND software to play in the mainstream. There are no exceptions. For example, ALL network processor semiconductor companies must provide reference platforms based on WIND’s software solutions or else there customer cannot expect to get to market – and the semiconductor companies believe this.
A connected world requires WIND. If WIND went away tomorrow, the very next day, the semiconductor companies would begin anointing a replacement. On the other hand, the connected world only wants one WIND; albeit, it will always want one or two others to keep WIND honest and to keep everyone from feeling trapped.
To some extent, Linux is now playing the support role to WIND. High profile Linux design wins by high profile WIND customers sends a useful message to everyone.
But Khan makes an important point about the availability of Linux plug-ins and the undeniable popularity of Linux, not as open-source, but as a standard, free OS. Of course, you would argue that, no matter what one thinks of Linux, you can’t align your fortunes with any fly-by-night Linux vendor. And, I argued that you shouldn’t roll-your-own Linux platform solution. As pointed out by James, plug-ins will be increasingly available for VxWorks as well, but still Khans point stands.
This means that Linux is somewhat stranded and probably doomed forever to the support role I mentioned above. However, there is one other possibility. If the market is serious about Linux, then WIND will offer Linux just like it offers Java. It could be the Linux API; it could Linux as a task under VxWorks AE; or it could be pure Linux in addition to VxWorks. If someone in the world is going to package Linux with all the acronyms needed in this connected world, that someone must be WIND.
Note that a WIND Linux package eliminates your concern about vendor choice, and it eliminates my concern about companies mucking with open source code, and it guarantees platform flexibility and access to all the open acronyms. And that’s why this solution would dominate the embedded Linux world.
Oh, and BTW, the package will not be free and no one cares.
Allen |
