It might be time to give this idea some serious consideration.

MSDOS begat Windows 1.0, which begat Windows 2.0, which begat Windows 3.0, which begat Windows 3.1, which begat Windows 95, which begat Windows 98, which begat Windows ME, which begat Windows NT, which begat Windows 2000, which begat Windows XP, which begat Windows Vista, and the list goes on.

If you've been in the IT business for as long as I have, you no doubt have more than a passing familiarity with most, if not all, of the products in the Windows OS family tree. Now that Windows 7 is all the rage, you're probably planning for (and possibly dreading) your next big upgrade. It's a given that you're going to do it, right? Not so fast! Perhaps it's time to consider replacing Windows with Linux, especially on the desktop. In this article, I'll examine why I believe the desktop OS isn't as important as it once was. I'll also look at the issues that may affect your decision whether to renew your reservation at the Microsoft camp or whether to look for new territory.

I'm old enough to remember batch processing and interactive time-sharing systems. The computers were expensive, and users interacted with them via dumb terminals. Back then, the users made do with the tools they were given. Change requests were put into committee and, if approved, would make their way into the development queue. The advent of the desktop computer changed all that, empowering users to explore new packages, such as spreadsheet programs, that allowed users to sidestep the application backlogs so common in corporate data-processing departments. Centralized computing begat distributed computing. And distributed computing begat a huge desktop software industry. We all know the story about how a certain Harvard dropout managed, through luck, timing, superb marketing, acquisitions, and odious anti-competitive behavior, to create a company that dominates the desktop software market to this day.

The funny thing about human history is that it tends to repeat itself. What's old is what's new. It's been happening in fashion (I've recently seen kids wearing bell-bottom pants and hip-hugger jeans), and it's happening in the computer industry as well. Dumb terminals and time-sharing systems are coming back in style in the guise of Internet-connected systems. The dumb terminals are represented by our desktop systems running Web browsers, and the time-sharing systems are represented by the servers they connect to. Web services, Software as a Service, and cloud computing, aided by broadband Internet connectivity, threaten the current desktop-centric paradigm. Ironically, this is occurring at a time when desktop horsepower is becoming cheap as dirt. I contend that desktop as we know it is becoming an anachronism. Probably sooner than later, the desktop computer will be little more than the equivalent of a terminal. I see that occurring now both in the applications that fuel our business and in the way we interface to our business partners. Our main line of business applications (written in Java) run as well on my Linux desktop as they do on my colleagues' Windows desktops. And the banks we deal with are now doing everything over the Internet. Firefox works fine for that, thank you. I find that I am as likely to reach for my Motorola Droid to log on to a remote system or to check my email or schedule as I am to go to my desktop computer. Perhaps you're seeing the same thing in your own shop. The ramifications of this are clear: the choice of desktop operating system is becoming irrelevant. And if that is true, then you really owe it to yourself to consider your other options before upgrading to the next iteration of Windows.

Once you've decided to evaluate Linux as your desktop computer, you'll need to choose a distribution. A distribution is a collection of a Linux kernel and various utilities and productivity applications. There are literally hundreds of distributions, each of which has been designed for a specific purpose. Some are targeted for server use. Others work well as a desktop. Still others have been created to load on "small" hardware. Some are even there because of someone's vanity. Surfing to DistroWatch.com gives you access to an incredible list from which to choose. Just beware: if you were confused by Microsoft's Vista flavors, you can really get a headache trying to pick a specific Linux distribution. My personal favorite for desktop deployment is Ubuntu from Canonical. Ubuntu has been designed to make it an easy transition for the Windows user. Canonical has taken care of signing license agreements where necessary so that an Ubuntu distribution is pretty much ready to go out of the box; many other distributions require the user to hunt down and install packages, such as MP3 decoders, by themselves. Commercial support is available from Canonical, so you have someone to call for help when you need it. That'll make it an easier sell in the corporate environment. The biggest recommendation I can make for Ubuntu as a desktop OS is that it passes the "mom" test. My mother is in her 70s, and she has no trouble using her machine for Web surfing, email, and word processing. It only took about a half-hour of training to move her from her Windows box to the new Ubuntu box. Though I recommend Ubuntu, there are many others from which to choose. Since they're all available as ISO downloads, there is little cost involved in trying different distributions—just bandwidth and blank CDs or DVDs.

While I'd love to see everyone deploy Linux on their desktops, I'm cognizant of the limitations many IT shops find themselves in. If your company has taken the Microsoft "blue pill," then extricating yourself can be quite expensive, if even possible. It all depends on how far the MS tentacles extend into your enterprise. Linux can play quite well in a Microsoft network. Authentication can be done against your Active Directory server, and, of course, Samba has been around for many years, so file-sharing is not an issue. Where you may find yourself running into a wall is with the applications you use in your business. If they've been written with Windows as a target, then you're out of luck for easily running them on the Linux desktop. There are some workarounds—I use virtualized Windows XP instances and VNC to do this—but the cost of the complexity introduced can exceed the savings that you'd realize switching. Even if you can't justify replacing your Windows network (I'll avoid any jokes about "if it ain't broke" and Windows), then perhaps you can find a place where it would make sense to deploy Linux. As an example, we have an Ubuntu instance used for accessing Facebook. It sits outside the normal network but is accessible remotely from inside. Thus, those who require Facebook access can have it without endangering their business desktop with the malware that somehow finds itself on the social networking sites.

Is it time to replace Windows with Linux? I'd love to be able to readily say "yes," but unfortunately most companies have become so invested in that platform that a complete switch would be too disruptive.  Unless and until such a company decides that an upgrade from one Windows version to the next is as costly in time and money as a total conversion to another platform, such as Linux or McIntosh, then the best it can do is realize some savings using open-source software. Much of that is available for Windows; the most striking example I can offer is the OpenOffice productivity suite. We use that here for virtually all of our desktops computers.

I do believe that it is possible to replace Windows with Linux in new enterprises, though. In a new business, it may very well be possible to do everything, from the server to the desktop, without using Windows. As I said earlier, the desktop OS choice is slowly but surely becoming irrelevant. A company with a Linux network may be ahead of the curve and be better-positioned to save money and do business in the future. At the very least, it's a possibility that deserves a thorough investigation.

Barry L. Kline is a consultant and has been developing software on various DEC and IBM midrange platforms since the early 1980s. Barry discovered Linux back in the days when it was necessary to download diskette images and source code from the Internet. Since then, he has installed Linux on hundreds of machines, where it functions as servers and workstations in iSeries and Windows networks. He co-authored the book Understanding Web Hosting on Linux with Don Denoncourt. Barry can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it.

Also Read

Practical OSS: MkDocs for the IBM i, Part 2

Practical OSS: MkDocs for the IBM i, Part 1

Practical OSS: MMAIL from EASY400