Open source has grabbed a big part of the server and app dev market. Apple has redefined the mobile device market and rendered Windows Mobile devices beyond passe. Firefox has blunted Internet Explorer's dominance, reversing the ActiveX hegemony for interactive Web apps, and removing one more barrier for widespread Macintosh usage. Vista is a dud -- "a piece of junk," one Gartner analyst called it yesterday in an interview.

This week, it got worse. Google is now challenging the Office cash cow and Microsoft's cautious steps into on-demand services through its new offering to let software vendors build Google-based apps. Salesforce.com has been remaking itself into a business app development and provisioning platform, and is likely to link up with Google next week on a joint offering.

Almost everywhere you turn, Microsoft is challenged. (Exceptions: Its SharePoint collaboration platform is well liked and the fastest-growing product in Microsoft's history, noted Gartner analyst Neil MacDonald, which could help preserve Office's fortunes against Google's incursions. Its Windows Server 2008 is a good OS that will serve businesses well. And its app dev tools dominate in the Microsoft-centric world of .Net.)

An enterprise strategy may have blinded Microsoft

Microsoft spent a small fortune to beef up its enterprise creds in its investments in business apps like Great Plains Software and the Dynamics CRM tools, in an attempt to get a way from its PC origins and the lack of serious that 1990s businesses treated Microsoft with as a result of those origins, notes Gartner's MacDonald. It succeeded, creating a US$1 billion business.

But despite that effort, Microsoft is still No. 4 in that space, even though it gained the enterprise creds it wanted. But Microsoft may have fought the wrong battle, argues Gartner analyst Darryl Plummer. The world has moved beyond enterprise computing to global computing, where the needs of one user are as critical as the needs of millions.

That requires a strong customer focus -- ironically what Microsoft had in its PC-oriented days -- that "enterprise class" doesn't deliver. Enterprise apps are optimized for large groups, not also for individuals, Plummer says. That's why it's a new breed of provider -- Salesforce.com, Google, Amazon.com and Facebook -- that are emerging as the new opportunities for both markets and technology.

A lock-in strategy doomed to fail

The "enterprise" apps are in what fellow analyst Yvonne Genovese calls a "terminal state of decline," consolidating into four vendors: SAP, Oracle, IBM and Microsoft. These vendors are trying to own their customers through a strategy of owning the entire stack that enterprises use, such as by adding their databases and middleware under their apps. The goal, she says, is to make an enterprise dependent on that integrated stack, creating a moat around their customers (and locking their customers in it). The four companies have different tactical approaches but the same goal.

SOA's abstraction and standards-based integration approach are meant to counter that lock-in strategy, but the app vendors already have figured a way to make SOA useless for that purpose, Genovese notes: They don't use standardized business processes, even when they use standard SOA mechanisms, so their apps won't in fact interoperate with independent services. Unless of course they are designed to be part of the app vendor's "ecosystem," which exists to lock customers in and competitors out.

These vendors may get a five-year lock-in from their impose-our-stack strategy, argues analyst Richard Hunter, but customers will eventually escape to something else -- just as happened when AOL played the same game in the 1990s. That "something else" is likely to the likes of Google and Salesforce.com, he says. He and his Gartner colleagues all cite cloud computing as the vague but increasingly clear space that this new world will exist in.