Windows Insider

Upgrading Windows XP-Based PCs Is Like Replacing an Aging Bridge

The widespread popularity and pervasiveness of Windows XP for means IT shops looking to upgrade before the end of support has a daunting task in front of them.

• By Greg Shields
• 02/24/2014

I think of bridges, the kind you drive over, when I consider the industry's "Should we retire Windows XP?" debate.

Bridges and the roads they link are critical to transporta-tion infrastructure. They connect locations that are otherwise disjoined and accomplish their job with remarkable success.

So much success, in fact, that they're an entirely forgettable part of the everyday commute. Once in place, you rarely think about them. As long as they work, they're the last thing anyone thinks needs changing.

Fixing a bridge or replacing it entirely is an inconvenient activity. Doing so takes time. The process often involves schedule setbacks, cost overruns and incomprehensible activities that are tough to appreciate when you're idling in construction traffic.

With mere weeks remaining in its operational lifecycle, Windows XP can feel a lot like a bridge. It's been around so long you've gotten used to its presence. So comfortable, in fact, that anything new just seems like an inconvenience.

Like the 7,980 U.S. bridges experts have designated "structurally deficient and fracture critical," according to SaveOurBridges.com, the aged-but-still-functioning Windows XP is ready for a preventative overhaul. That process is fraught with peril, not because upgrading an OS is particularly difficult, but rather from all the externalities -- the "everything else" -- that must be addressed in completing the project.

Hardest of all is managing users' perceptions. Same as with bridges, users just want to get to work. A new OS feels more like a traffic jam than an exciting technological advancement.

With all this conspiring against, how do you survive the looming death of Windows XP? To me, the answer is simple: Act like a bridge builder.

Inform the Complaints
There's a curious trend I've noticed in IT lately where technologies are taking the place of traditional, everyday communication. In most situations, this makes a lot of sense. A technology for automating an always-on VPN connection is far superior to training users which buttons to click.

Yet there are situations where the old-fashioned training class still holds merit. OS migrations are one such example.

The bridge builders do it. Government entities hold town hall meetings to inform and educate their constituency on upcoming projects. Most are held with the knowledge that few will attend. It's the holding of the meeting that's arguably more important than people attending. It offers an opportunity to be heard.

The same holds true in OS migrations. An old-fashioned training class can do wonders for educating the interested while buffering the complaints.

Invest in Heavy Equipment
I'm a big fan of Microsoft's free OS deployment tools. I've also been using them since the 20th century. I know my way around Windows PowerShell, and can compatibility test, package and deploy just about any software package in existence.

But not everyone's investment in heavy equipment has paid off. For everyone else, an array of for-cost solutions exists both from Microsoft and third parties for managing application compatibility, gathering inventory, deploying OSes and applications, and preserving user state information.

Buying the right set of tools, however, is fundamentally important. Bridge builders invest in heavy equipment because they can continue to use it for decades. The right tool for your use is one that'll continue to manage desktops throughout the rest of their operational lifecycle.

Use Detours Strategically
OS migrations are all too often held back thanks to a tiny subset of problem applications and devices: A line-of-business app runs only on Internet Explorer 6 (God help you), a vendor won't upgrade their support statement, another vendor no longer exists and so its app will never support a new OS.

These are surmountable problems because "the looming death of Windows XP" needn't necessarily mean its wholesale eradication. Solving this problem is an excellent use case for virtual desktop infrastructure and desktop virtualization -- for just those problem apps. A no-longer-supported Windows XP might never touch a production network again without significant protection, but we're IT people and a little virtualization can go a long way.

Nearly 8,000 bridges are structurally unstable and long past their recommended lifespan. Like Windows XP they still work, but for how long?