If you Google the phrase "Microsoft's worst nightmare," the range of hits you get is entertaining for its breadth. Various pundits have proclaimed that the software giant's very worst nightmare is everything from Linux, Google and Firefox to software as a service, Cisco's digital home business and the Sony PlayStation.

They can't all be Microsoft's worst nightmare, so which one earns that distinction? The correct answer is none of the above.

The best way to gain insight into what Microsoft truly fears is to research the musings of Bill Gates. I've had enough encounters with Gates over the years to have recognized that the more he dismisses a competitor, the more he fears it. In 1999, when I asked him how concerned he was about Linux, he blew it off as a "darling of the press" that was nothing more than the equivalent of the first Windows NT kernel.

In a 1995 interview, when I asked Gates for his views on the emergence of Netscape, he derided media interest in the company as an "overreaction" because "an Internet browser is a trivial piece of software." Enlightening as that comment was, there was something else Gates said in that interview that revealed his worst nightmare.

I asked Gates what trend or development had occurred in the technology sector in the past 20 years that really caught him by surprise. His deadpan response: "Kaleida and Taligent had less impact than we expected."

Gates was referring to two software joint ventures formed in the early '90s by Apple and IBM that were already fading into oblivion. There was something different in his tone -- a biting sarcasm -- that reflected a degree of scorn that he seemed to reserve for the Apple/IBM combo. And it was telling.

Microsoft's worst nightmare is a conjoined Apple and IBM. No other single change in the dynamics of the IT industry could possibly do as much to emasculate Windows.

Buzz about an Apple/IBM merger has arisen cyclically ever since the 1991 formation of the PowerPC microprocessor partnership of IBM, Apple and Motorola. It was especially loud in late 2004, when IBM sold its PC business to Lenovo, seemingly making way for Apple to fill the void. Timothy Prickett Morgan accurately wrote at the time in The Linux Beacon , "IBM has the reach to make Apple pervasive; Apple has the means to make IBM cool."

It really doesn't matter how cool IBM is. But the pervasiveness of Apple is a different story, because a pervasive Apple would be a boon to corporate IT.

As it is, IT shops that are looking at Apple as an alternative to Microsoft are being thwarted by Apple's ditziness in the enterprise. Robert L. Mitchell wrote about why the insurrection could happen in the enterprise -- and why it probably won't.

Mitchell spoke with Dale Frantz, CIO at Auto Warehousing, who since last year has been working to migrate his core IT operations from Microsoft to Apple. "The biggest weakness at this point, I'd say, is the lack of a cohesive enterprise strategy on the part of Apple," Frantz said. Mitchell couldn't get anyone at Apple to even talk about its enterprise strategy.

The only surefire way for all of that to change is for an IT powerhouse with a dominant enterprise presence to acquire Apple. I wrote in my blog last week about how close Oracle CEO Larry Ellison came to partnering with Steve Jobs to buy Apple in the late '90s, and how that might have transformed Apple into more of an enterprise player.

It's unclear how close IBM has come to acquiring Apple in the past, or how interested it might be now. What is clear is that such a move would be warmly welcomed by a lot of IT shops that hate being dependent on Microsoft. And there's not a more nightmarish scenario hovering over Redmond.