Fat or thin: It's not a weighty issue

Many CIOs, I'm sure, are taking up time at company meetings to discuss the issue of thin clients with the CEO, CFO and IT staff. Each discussion, about whether the company should introduce thin clients into the enter-prise, is accompanied by detailed reports that outline the cost associated with changing the existing client/server setup to accommodate a new model of computing. And then the numbers are counterbalanced by the huge savings on support and maintenance associated with moving toward thin technology.

Perhaps the executive meetings spark some heated arguments on whether thin clients belong. But it doesn't have to be that way, because as John Duffy, senior vice president of communications services at Cigna Corp., in Voorhees, N.J., pointed out to me, it's not about re-engineering the infrastructure with new pieces of hardware and software. It's about designing applications to work in either fat-client or thin-client mode.

In other words, you can keep that hefty memory hog of a PC on your desk but enable it to work as a fat client or a thin client where it makes sense.

"If you look at it as a virtual thin client, it makes it a lot easier for people to make the decision to get into thin-client computing, because it isn't an either/or decision," Duffy said.

Making the call
Here's how Duffy deduces what applications operate in fat or thin mode: If it is used heavily by many people, it makes sense to put it on the desktop in fat-client mode. If it is used occasionally, it might make sense to put it on a centrally located server to download on demand.

Now, simply re-engineering applications to work in an either/or mode isn't necessarily going to ease support issues--which thin clients are supposed to do--because users will still be grappling with hard-drive crashes and device driver problems. But simply making the transition to thin clients has some support issues of its own. The transition, I would assume, would be rolled out in stages, which means fat and thin clients would coexist. And support staff would still be splitting time be-tween desktop issues and users trying to get comfortable with their new computing devices.

Face it, fat clients are here to stay.

Whenever I hear a big thin-client vendor like Oracle talk about how these devices are going to take over the desktop world, my ears start to ring. Remember that old marketing ruse, something about mainframes being dead? I haven't seen a big-iron graveyard. They didn't die, they evolved.

Desktops are going to do the same thing, with thin clients a part of the equation, be it an actual device or a virtual thin client, like the direction Duffy is moving in.

The biggest infrastructure issue doesn't have anything to do with the desktop device, but rather with network bandwidth. As long as a company has a backbone to support the variety of systems that pepper the enterprise, then the actual end-user device is a moot point.

Keep your mainframes, keep your client/servers, add some thin clients if you want--or just design your applications to work for the environment. Because it's not about risking it all with thin clients. It's about retrofitting what we already have to support a cost-and-support model that works for each individual department or user within your very large organization.

Do I make it sound too easy? What are your thoughts on fat and thin clients in the enterprise? Let me know in the talkback below.