The market has rejected Linux desktops--get over it

I've been running Linux on PCs since 1998, when Red Hat still cared about the desktop and Mandrake was supposed to be the distribution that was going to bring Linux to the masses. That was also about the time that the mainstream media got infatuated with the story of the free operating system from the Finnish hacker that was going to bring down Microsoft Windows.

Spoiler alert: I'm going to give away the ending now. It never happened.

In the decade since it was first proclaimed as the "Windows killer", Linux on the desktop has made virtually no progress in real adoption numbers. According to market share trackers (based on real PC activity and not just sales) such Net Applications, StatCounter, W3Counter, and others, the market share of Linux has been hovering around just 1-2 percent of total PC operating system installations for a decade.

Even in the past two years since the Netbook phenomenon began with Linux as its primary OS, Linux market share has failed to make a major jump. The chart below, based on Internet visitors tracked by Net Applications, shows the trajectory of Linux desktop market share over the past 24 months.

Notice that Linux market share got a little bit of a bounce (mostly from netbook sales) in the first half of 2009 but has been dipping since then. Even so, the top line here is the 1 percent market share threshold, so the peaks and valleys are pretty insignificant when viewed from the perspective of the larger desktop OS market.

Despite this consistent evidence that Linux desktops were going nowhere, pundits, analysts and Linux enthusiasts have been repeatedly predicting that Linux was on the verge of a breakthrough on the desktop. At the end of nearly every year, some writer or publication has prognosticated that the following year would be "The year of the Linux desktop".

Here's a quick selection of these Linux prophecies:

• 2009: Year Of The Linux Revolution (FastCompany)
• 2009: Can Ubuntu Linux become a Windows killer? (TechTarget)
• 2008: Linux's year on the desktop (ZDNet)
• Is 2008 the Year of the Linux Desktop? (Linux Magazine)
• 2007: Forrester Calls Desktop Linux a Credible Threat to Windows (eWEEK)
• 2006: The year of desktop Linux? (GCN)
• 2005 will be the year of the Linux desktop (Express Computer)
• 2004: The Year Linux Grows Up or Blows Up (eWEEK)
• Linus Torvalds Says 2004 is the Year for Desktop Linux (Slashdot)

Desktop Linux: What happened?
Why hasn't Linux succeeded on the desktop? There are several simple reasons:

1. It's still too much of a pain--While Ubuntu has made Linux much, much easier, it's still not quite as easy to hook up a new printer, connect a digital camera, or manage your work calendar, for example, as it is on Windows. Plus, on the other end of the spectrum, Mac is even easier than Windows for most tasks and it has the same Unix underpinnings as Linux.

   As a result, in the last few years a lot of the hard-core techies who are the primary candidates for Linux have instead jumped to Mac OS X as a Windows alternative.

2. The divide and fail strategy--The energy behind the Linux desktop movement has been divided up between a lot of different players, from Red Hat to SUSE to Ubtuntu to Debian at the software level to hardware players such as IBM and Dell at the PC level. A decade ago, the thought was that the force of Linux attacks from multiple angles would ultimately outflank Microsoft Windows. Instead, it has diffused the force behind Linux and dulled its attack from a marketing perspective.
3. Not enough innovation--The primary value proposition for Linux is that it's just as good as Windows--or at least "good enough"--and costs a lot less. Occasionally, you'll hear that Linux is more secure or more stable than Windows--which can be true, but that's mostly based on its Unix foundation. But, what innovative features has Linux brought to the world of desktop operating systems? The only one I can think of is the desktop manager / virtual desktop (which Mac OS X eventually adopted as its "Spaces" feature).

   The technology industry (and the consumers and businesses that support it) are still driven primarily by innovation, and the Linux development community has spent too much time trying to copy Windows and not enough time innovating on its own OS.

4. Businesses want someone to blame--As my colleague Bill Detwiler says, IT professionals prefer to have someone to point the finger at when critical systems blow up and it leads to lost revenue or productivity. If you have Windows desktops (or even Mac), you've got a big target to point your finger at if you're having a PC software problems, and someone predictable to call to help figure it out.

   On the other hand, if your IT department went with Linux desktops then you'd be going out on a limb. If something goes wrong--like users losing productivity from incompatible software--the finger could get pointed back at the IT leader who made the decision to take a non-standard Linux approach, since there's no big software vendor to blame it on. In other words, Linux can expose IT leaders to more risk.

What about Google Chrome OS?
Now we've got Google Chrome OS being hailed as the latest savior of the Linux desktop. Google is taking a very different approach than Ubuntu or SUSE. The search giant is taking its Chrome Web browser and turning it into Web-only OS that will boot instantly, rely solely on Web apps, and drastically minimize local storage.

The Chrome OS will technically have Linux as its foundational software but it will not allow users to install Linux applications or even get to the Linux command line. It will be a non-standard, custom Linux kernel that serves only to boot the Google Chrome Web browser as quickly as possible.

Chrome OS is an intriguing concept and it will be one of the first big tests of the extent to which consumers and businesses are ready to accept the paradigm shift to cloud computing. However, it's a concept that's probably still several years ahead of its time and unlikely to make a major impact on the PC market in 2010. It's also a little spurious to call Chrome OS part of the Linux desktop movement since the only thing it really does for Linux is to strip it down and get it out of the way.

Bottom line
It's time to stop all of the misguided predictions about Linux becoming a force on the desktop. That ship has sailed. The masses don't want it. Businesses don't want it. Even Google can't change that.

Linux is still building major momentum in servers and mobile devices. In the data center, Linux is replacing lots of Unix servers and is more than holding its own head-to-head against Windows servers. In mobile, Linux quietly serves as the underpinning for both Google Android and Palm webOS, the two platforms that pose the biggest challenge to the incumbents in the smartphone market. However, on the desktop, it's time to just admit that the market has rejected Linux.