Due Diligence: The Big Bad WolfBy Eric Nee | Posted 07-15-2003
As if cast in the image of its creator Linus Torvalds, Linux has become the fair-haired boy of the computer industry. Over the past dozen years, Linux has grown in popularity and matured in capability at a fast if not spectacular rate. Today it stands on the cusp of becoming one of just two contenders for the mantle of server operating system of choice. But Linux's once-rosy future is now in doubt. That's because the sight of the fair-haired boy happily skipping down the path of success has finally attracted the wolves.
One small but cagey wolf is SCO Group, which claims Linux infringes on SCO's Unix source code. Money may be the main motivation for both the company's $1 billion suit against IBM Corp. for allegedly violating its Unix license, and the threatening letters it has sent to corporations that are using Linux. But the result of SCO's threats is to create uncertainty, which over time could cause corporations to back off their support of Linux.
In the long run the small wolf isn't the problem. The real threat comes from the much larger, more dangerous and more cunning wolf that's intent on killing Linux. That wolf is Microsoft Corp. The parable has been told many times before in the IT industry, often with an unhappy ending: A young if not necessarily naive upstart gains a foothold in the market with a new technology, rivals jump in, and a bloodbath follows, most often with Microsoft the ultimate victor.
What is noteworthy this time is that the outcome of the story could very well determine the future of corporate computing. Microsoft wants more than to have Windows running on every computer, from the largest data center down to the smallest cell phone, as ambitious as that plan may be. The company's ultimate goal is to leverage its domination of the operating system into every nook and cranny of the corporate application environment. Microsoft has already succeeded on the desktop, and it is well on its way to doing the same on the server.
That's why the current brouhaha over Linux is so important. Linux is the industry's last hope for an alternative to Windows über alles. Just consider the alternatives: Solaris? HP-UX? z/OS? NetWare? It's a lineup of has-beens that is getting more pathetic as time goes on.
Linux is the only server operating system besides Windows that continues to gather momentum. In the first quarter of this year, sales of Linux servers (measured as a system with hardware and software together) were up 35 percent over the same period last year, says IDC analyst Jean Bozman. By contrast, the overall server market was down 3.6 percent. Microsoft's Windows dominates the server operating system market, accounting for 49 percent of the new units shipped in 2001, according to IDC. Linux was next with 26 percent, followed by NetWare and Unix at 12 percent each and all the rest at 2 percent.
As fast as Linux is growing, it is still a nascent and vulnerable operating system. It is largely confined to Internet-related tasks like mail, print and Web servers. Despite support from SAP AG and PeopleSoft Inc., few corporations have moved enterprise applications to Linux. Joe Farrelly, CIO of Aventis Pharmaceuticals Inc., is typical. He was quoted in a recent Merrill Lynch report as saying Linux ran on just 3 percent of the company's servers.
Still, there's lots of heft behind Linux. IBM Corp. has thrown its full support behind it. The company began offering Linux on its mainframe computers in early 2000, and now "we have 600 customers using Linux on mainframes," says William Zeitler, senior vice president and group executive of IBM's Systems Group. "It's accelerating as more and more people see the value in it."
IBM's success selling Linux servers is one of the reasons SCO is pursuing its lawsuit against Big Blue. The Linux community is mostly sanguine about SCO's suit, saying it has no merit. While that may be morally true, it might be legally incorrect. It's not possible to know until experts have a chance to review the contracts, examine the actual source code, and look back at the history of how the code was written, says Stacey Quandt, a Linux analyst with Forrester Research.
In Quandt's view, corporate customers have not yet been deterred from buying Linux servers. But if the lawsuit continues to drag on, risk-averse CEOs and CFOs may avoid potential problems and postpone Linux deployments. That would be music to Microsoft's ears. Such customers would probably turn to Windows instead of Linux. And Linux's legal problems would lend credence to what Microsoft has been saying all along: You can't trust an operating system that doesn't have a true owner, a company that will make sure not only that the right features are added but that it is all done legally.
After first ignoring Linux, and then dismissing it, Microsoft has in the past year gone after the operating system aggressively. In his annual letter to employees sent out in June, Microsoft CEO Steve Ballmer targeted Linux, writing that it would receive "our concentrated focus and attention." The war is being waged on two fronts. The first is a test of technical features and the first shot is Windows Server 2003. This is a fight Linux can handle, and one that will benefit customers by providing them with better products.
The second front that Microsoft has opened against Linux involves the hijinks Microsoft is known for. In May, Microsoft signed a licensing agreement with SCO, helping to lend credence to SCO's claims. It was also widely reported that Microsoft had recently told its sales force that under no circumstances should it lose a government sale to Linux, even if that means heavily discounting the price. In addition, Microsoft has announced it will donate as much as $1 billion worth of software to nonprofits over the next four years, a move some see as an attempt to pre-empt Linux making inroads into this cost-sensitive market.
If Microsoft succeeds in squashing Linux, it will be bad news for CIOs. First, they could expect even more billions to flow from their budgets into Microsoft's pockets. Much of that money will come from operating system sales, but an equal amount will come from applications. Look at what Microsoft has accomplished on the desktop. In the quarter ending March 31, sales of desktop system software totaled $2.5 billion, and desktop applications totaled $2.3 billion. Server system software brought in $1.8 billion, while server-based enterprise applications totaled just $150 million.
Microsoft clearly has catching up to do on the server applications side, but it is moving fast. In the past 21/2 years the company spent $1.1 billion to buy Great Plains Software Inc. and $1.45 billion to buy Navision, both makers of business applications for midsize companies. Oracle Corp.'s recent hostile offer for PeopleSoft will no doubt trigger other buyouts in the enterprise software industry, whether or not it succeeds, and don't be surprised if Microsoft joins in.
Alternatively, Microsoft could choose not to make any more big acquisitions, and instead enhance the software it has. The company believes that a new set of interactive business applications will be developed that make use of XML, its own .NET, and other Internet-based tools. Microsoft also believes that by controlling the operating systems on all types of computers, it can create better applications. Whether they really are better remains to be seen. But you can be sure Microsoft will use its control of the operating systems to try to tightly integrate applications across the different platforms, and drive out competitors. The result would be fewer startups because venture capitalists won't fund them. Innovation would slow and competition would evaporate. It's not a happy ending.
Will the story end that way? That depends on a whole host of unknowns. But it doesn't have to. "Little Red Riding Hood" ends with the wolf in Grandma's clothing being killed by the hunter.
Eric Nee, a longtime observer of Silicon Valley, has served in a variety of editorial positions at Forbes, Fortune and Upside magazines. His next column will appear in October. Please send comments and questions on this column to email@example.com.