Back in the mid-’90s, my research focused on desktop operating systems. There was a plethora of options for IT organisations, with Mac OS, Windows in the guise of NT and 95, and OS/2 Warp all vying for the attention of IT managers. Even Unix workstation vendors had thoughts of moving beyond scientific and engineering applications to mainstream knowledge worker desktops.

But by the late ’90s, it felt as if I was doing the colour commentary for a horse race whose leader was out in front by 10 furlongs. Still, while it was clear to many that Microsoft was going to dominate the desktop, that didn’t stop some in IT from looking for alternatives.

Then a dark horse emerged. Many people now believe that Linux represents a viable alternative. Today, with mainstream hardware vendors like Dell offering Linux installations and some folks thinking a major shift is about to happen, it’s time to take another look at Linux on the desktop.

Unfortunately, despite major strides in recent years - notably the Ubuntu release - Linux still isn’t viable for most end-users or organisations.

Take a look, for example, at the Dell offering. When it was first announced, I asked company officials whether it was a mainstream product with full support. No, they said. The Linux machines were meant for enthusiasts who wanted a “no Windows” option. Users would still have to pay for the operating system - about $50 less than Windows, which was hardly a major saving - and significant features would be missing because of a lack of driver support.

In short, even though Linux has come a long way in the past few years, it hasn’t come far enough. The latest and greatest hardware still arrives without Linux driver support. Until a vendor is willing to take a gamble and build fully optimised Linux systems, most IT shops simply won’t bother to make the costly transition.

And cost is the hidden factor. While much is made of Linux’s being free, the truth is that software costs account for only about 10 percent of total cost of ownership for PCs.

Finally, there’s the lack of critical application support. Most notable for businesses is the lack of support for Microsoft Office. Yes, there are office suites available for Linux, but the reality is that most organisations are dependent on Microsoft’s applications. Anything with less than 100 percent interoperability and compatibility isn’t going to make it in the business world. And does anyone believe that Microsoft will ship a Linux version of Office anytime soon? Or ever?

And it’s not just business users who are affected. Sorry, consumers, but there’s no version of iTunes for Linux.

So, the search for an alternative to Microsoft on the desktop continues. The fact that a mainstream hardware vendor like Dell is willing to make a Linux effort is laudable, but until such offerings enter the mainstream, we’ll have a catch-22 situation in which vendors wait for users to adopt and users wait for vendors to deliver.

For now and the foreseeable future, it’s going to remain a Microsoft world. Linux still isn’t the answer. And of course, there is always that other Unix-based operating system that has gained popularity over the past few years. It’s called Mac OS X, and it comes from Apple.

Michael Gartenberg is vice president and research director for the personal technology and access and custom research groups at JupiterResearch in New York.