COMMENTARY--The end of the world is nigh. Teams of cats are chasing dogs through the streets. Cows are sick of grass and are using cell phones to order pizza. French chefs are dreaming up ways to use bales of hay in gourmet cooking. Scott McNealy is talking about the synergies which can be built between Sun and Microsoft technologies.
|
| ||||
|
| ||||
|
News Focus Microsoft-Sun settle War is over |
||||
|
|
||||
|
|
||||
Sun Microsystems made decisions based on outdated information regarding the competitive landscape. Though Microsoft certainly did, and does, constitute a competitive threat, Sun's real challenge is the growth of a free and open source Unix alternative named Linux. The settlement hints that competitive advantage might be found in a common (albeit loose) front among vendors of proprietary software, each of whom are free to license each other's intellectual property.
The Linux path of destruction
In 2002, Linux's share of servers shipped was 23.1 percent,
up from 22.4 percent a year earlier. In the same, year, Windows' share grew to 55%,
from 50.5 percent in 2002.
My Google net didn't pull in comparable statistics for 2003, but consider the following: though units shipped with Linux grew by 51.5 percent year over year in the third quarter of 2003, sales of Windows servers grew by 23.3 percent. Total server units shipped for the quarter included 210,000 running Linux (an increase of 71,000 units), versus 841,000 boxes running Windows (an increase of 159,000).
Given Windows continued strong growth, it's pretty clear that Linux has more of an effect on Unix sales than Windows sales. The reason isn't hard to fathom.
As I've explained in past articles, there are two dominant programming domains, Windows and Unix. Linux isn't per se Unix, though it is similar enough that technical personnel familiar with Unix have little trouble transitioning to it.
Unix and Windows programming domains do compete, though gains are more incremental, like a tug of war between two rugby teams. The economic "friction" caused by past investment in Unix/Windows software and the existing skill set of your available pool of Unix/Windows technical resources slows transition. Companies moving from Unix to Windows have faced difficulties, as have teams assisting Munich in its transition from Windows desktops to Linux.
Within the Unix programming domain, however, the barriers to transition are much lower. Proprietary Unix, when faced with free (or at least lower cost) alternatives, has a harder time competing. The result is that Unix feels the brunt of free/low cost Linux competition much more intensely than the "parallel" Windows programming universe.
A Seismic Shift: Peace with Microsoft
Sun's decision to change tack can hardly be considered voluntary. In their heyday, Sun was the biggest seller of servers, managing a 48 percent share of Unix server shipments. Today, they have fallen into third place behind HP and IBM. They have lost money for almost three years straight, and their stock has plummeted from the mid-60s in 2000 to 4.90 in April, 2004.
Change is never easy, though, and after years fighting Microsoft, certain habits of thought became as worn as chariot grooves in the streets of Pompeii. What does Sun hope to gain from its efforts?
First, Sun would free itself to boost compatibility with the Windows world. This is important, as Sun's biggest Unix competitors, IBM and HP, are platform agnostic. They happily offer systems running Windows, Linux, or their own proprietary variant of Unix. This allows them to manage higher levels of Windows compatibility than the more ideologically-constrained Sun Microsystems.
Sun might not be on the verge of announcing a new line of Windows servers, but they certainly can do more to ensure seamless operation in hybrid environments.
Ending a culture of Microsoft antagonism, and even more important, licensing the desktop interoperability protocols made available as a result of Microsoft's settlement with the U.S. government, are important steps.
Second, licensing Windows desktop protocols would enable Sun to differentiate Solaris from Linux. GPLed software can't use intellectual property which requires a licensing fee. Solaris, as a proprietary OS, can. This freedom will enable levels of compatibility that Linux would have difficulty matching.
Third, Sun might manage to ride the .NET wave without having to commit to .NET. As I argued in the past, Sun squandered a spectacular opportunity to use Microsoft as a vehicle to popularize Java. Sun can't change its past, but that doesn't mean that Sun can't make it easier for Java software to run within .NET environments. As an example, converting Java bytecode to .NET IL isn't rocket science, and provided the proper Java-like libraries are included, the latest Java code could run "natively" within a .NET environment.
Even so, it's a foregone conclusion that .NET is the future of Windows development, even if you don't think that it will spread to other platforms. Given the obvious popularity of Windows in desktop environments AND server environments, Java's ability to interoperate seamlessly with .NET will be an important selling point.
Conclusion
Assuming Sun manages to end old habits, they might manage to re-invigorate their server business by positioning themselves as a Unix option that slots easily within Windows networks. As long as Microsoft proves popular in enterprise settings, that will matter. Most companies maintain hybrid environments, and a Unix that plays nicely with Windows is a strong competitive advantage.
Furthermore, compatibility with proprietary operating systems is an area where Unix has an advantage over Linux. Unix can compete, provided that it provides a decent reason for people to pay a premium. One important advantage is proprietary Unices freedom to work with fee-based Intellectual Property. It's a smart tactic, and might just be the leverage the proprietary Unices need to keep ahead of open source competition.







