After several “stays of execution” and a brief resurgence due to the Netbook trend of a few years ago, it seems that Windows XP is finally on its last legs, with less than one year of official support remaining from Microsoft. This is certainly not a surprise to IT pros and, despite four years to adopt Windows 7, TechRepublic’s survey and my own informal polling (glancing at laptop screens on airplanes) indicate XP is still going strong, and many organizations have no plans to replace the timeworn OS. What’s going on?

The strange Windows dynamic

Early releases of Windows moved at a fairly rapid clip. During the early days of my IT career a release would occur every couple of years, and with the growing importance of Internet-based protocols, enterprises rapidly adopted each release. XP seemed to be a plateau of sorts for Microsoft, as the company provided service packs that fixed bugs and added compelling new enterprise features. Microsoft gave users six years of an ever-improving OS, and then botched the initial Vista rollout so badly that consumers and enterprises simply opted to forego it.

Windows 7 fixed most of the troubles of Vista and provided a modern UI while increasing performance, but also taught IT Pros an important lesson: you can skip Microsoft’s upgrade cycle without penalty.

Now, Microsoft’s latest OS, Windows 8, delivers a bold departure from the past and, for many, a bridge too far from XP, which continues to power the majority of the world’s PCs. Combine a major deployment effort and extensive end-user training with a user community that’s successfully panned two major Windows releases, and you have a recipe for ennui.

So what if we stay with XP?

There’s a seemingly valid contention that sticking with XP after Microsoft’s end of life for the OS isn’t a big deal. The daunting cost of Windows 8 and rumors of a modified release of Windows 8 that returns some familiar features furthers this sentiment. Most large companies have applications and custom code from the COBOL days, and no doomsday scenarios are playing out, so why should XP be any different?

While this logic is sound, XP remains one of the major targets for exploits, primarily due to market share. Just as infamous bank robber Willy Sutton quipped that he robbed banks “because that’s where the money is,” malware and targeted exploits will go after systems with a large installed base. Surely new exploits will be found after Microsoft ends support, and users will have to rely on Microsoft extending support yet again, or a third party layering protections atop Microsoft’s code.

But does the OS even matter?

With applications increasingly moving toward web- and cloud-based platforms, some are going so far as to question the relevance of the desktop OS beyond a modern “dumb terminal” that powers a web browser and little else. While we seem to be moving in that direction, recall that predictions of the “end of the desktop” have been with us since the late 1990s, and a malware-infested OS that runs little beyond a web browser presents just as grave a threat when an end-user plugs it into the LAN behind the firewall. While many IT Pros view an option to stay with XP as “free” when couched against the cost of a major OS upgrade, self-supporting the platform will incur a definite and rapidly increasing cost once Microsoft support ends.

So what’s an IT Pro to do?

As of this writing, Windows 7 seems the most compelling migration path. The OS is mature and stable, hardware compatibility is mature, and thousands of organizations have already blazed the trail to Windows 7, identifying most of the caveats. Unlike Windows 8, the after end-user will require little familiarization training with the new UI and will likely have seen Windows 7 or Vista on another machine. Windows 7 buys you extended support until 2020, giving Microsoft several years to perfect their vision for the future of computing.

If you’re considering migrating directly to Windows 8, wait for the rumored “Windows Blue” update, which may bring back familiar tools like the Start button and refinements to the user interface. While XP may be “good enough” today, the cost and hassle of self-support is an unappealing and expensive proposition.