It's been more than a decade since the Linux desktop became usable. It's older than Mac OS X, and technically speaking, X11-based GUI computing pre-dates Windows. Yet here in 2009, Linux desktops comprise a very small segment of the computing world. This isn't due to lack of options, performance, or stability; it has more to do with Linux being viewed as a geek option only. And quite frankly, it is.
I don't think this is a bad thing. After all, a GUI desktop is a tool, like any other computer program, and those who know how to use it have a leg up on everyone else. Simple things like middle-button paste are terribly useful but are completely lost on the majority of the computer users the world over. Maybe that's the way it should be.
One of the main roadblocks for ubiquitous Linux desktops is hardware driver support. Desktop and laptop manufacturers use a wide variety of chipsets for various components like Wi-Fi, video, and so forth. Most of the time, Linux drivers can be had for these chips, but not always. Hardware manufacturers are generally lax about providing Linux drivers for their hardware, even today, and the drivers that may exist might not be feature-complete, turning the process of getting a Wi-Fi card to work into a three-hour task. That's far more time than most people would ever want to spend on such a mundane process. Honestly, it's more than I want to spend too. However, the benefits of running Linux on my workstations and laptops far outweighs these issues.
That's the key. I see the time required to wrestle Windows into submission as taking far more time over the life of the system than the potential pitfalls of installing Linux on the system to begin with. The upfront investment is outweighed by the relatively constant care required to keep a Windows installation sane and virus-free. A stitch in time saves nine.
So maybe this is how it should be, at least for home computing. Those who know and understand computing at a greater level than the rest of the world choose to use tools that allow them more time to further their work and their productivity, while those who can't be bothered run whatever is on the system when they buy it. The haves and the have-nots, perhaps.
Then again, there are all those Mac users out there who have the best of both worlds -- an attractive, fully functional desktop system with no worries. The rising number of Mac users speaks to the success of Mac OS X, perhaps aided by the problems facing Microsoft Vista.
Frankly, it's still too early in the life of the personal computer for the Linux desktop to see widespread adoption. Microsoft found fortune by providing a relatively easy-to-use GUI for business computing that ran on standard hardware. While the Mac OS was better than, say, Windows 95, it ran only on Macs, which cost more. That made the decision for most companies and thus the status quo.
There are other major factors preventing the rise of the Linux desktop. First and foremost is business computing. While it's certainly possible to set up and configure Linux desktops with centralized LDAP, it's not as easy as Microsoft's Active Directory. In fact, if there's one thing that Microsoft has done extremely well, it's AD. That, coupled with the Office suite and Exchange, makes the case for Windows in the workplace. There will be Macs and perhaps even Linux systems on the same corporate network, but they'll be one-offs and most likely will be configured to play nice with Active Directory. Thus, Microsoft owns the corporate network.
To change this, a few things need to happen. Obviously, a realistic substitute for Active Directory and Exchange has to be developed and be significantly cheaper than Microsoft's offering. There are plenty of ways to handle e-mail, and OpenLDAP and others can certainly perform directory duties, but there's no clear-cut management interface that ties them together in the same way that AD and Exchange are managed. That's absolutely key.
Imagine that for a few hundred dollars, a medium-size business could implement a fully AD-compliant directory with e-mail support identical to Exchange, with support for Windows, Mac, and Linux clients. Imagine that there was no licensing to worry about, and a client desktop can run on any hardware and run Windows applications via virtualization or WINE. My guess is that it would be a big hit. Apple could do something like this if it put its focus on business computing, and any number of Linux vendors and ISVs have tried to do this in the past with mixed results.
However, times are different now. Microsoft's waning along with the economy, and businesses are more cost-conscious than ever. Vista was released too late, requires significant hardware upgrades, costs too much, and is generally viewed as a failure. The time is right for a real competitor of Microsoft to appear in the business network.