Since the turn of the decade or so, it's been regularly reported that every year will be the year during which Linux finally breaks out of its box and gains a major foothold at the desktop level.
It hasn't happened yet. I don't think it ever will, either.
According to March 2009 information compiled by Wikipedia from a variety of sources, here's how the desktop operating system market looks:
- Windows: 90.99%
- Mac OS X: 4.81%
- Linux: 1.05%
Here are prior year values:
- Mac OS X: 4.95%
- Linux - Q1 2008: 1.01%
- Windows: 94.48%
- Mac OS X: 3.98%
- Linux: 0.68%
The values don't add up to 100% due to the way that the data was analyzed, but the information is accurate enough to be significant. Presently, Mac OS X has a more than 300% advantage over Linux while Windows enjoys a lead not that far shy of 9,000%. Although the Windows lead has closed over the past couple of years from what was a 13,500% lead over Linux in 2007, Linux is far, far from close to overtaking Windows.
Even though Microsoft has made friendly overtures to the open source community, the fact remains that Windows continues to be the major force in the desktop computing environment. Quite frankly, I don't see that changing and, if it does change, I seriously doubt that Linux will be the contender that unseats the king. From the consumer perspective, Apple has the mindshare that could further erode the Windows lead. Even in the netbook market, which, at first, looked poised to be a niche in which Linux could gain a foothold, Microsoft has managed to turn the tide in its favor.
There are a lot of Linux distributions out there. Some are friendlier than others when it comes to the user experience. However, even with open source efforts such as WINE, Linux desktops remain incapable of running the vast majority of programs that people are used to running. There are open source alternatives, such as OpenOffice, that can mostly replace some of these commercial titles, but the open source programs are not widely used and, as such, aren't known to the general public. There is also the issue of enterprise software that runs only under Windows. Cloud applications such as salesforce.com, are growing in popularity and may help Linux take hold in some enterprises, but I don't see this trend killing Windows anytime soon.
I also believe that Windows 7 will have a net positive impact on the Windows market. Vista was, rightly or wrongly, not exactly well-received by pretty much anyone. Even though the product has improved, that taint remains. At this point, all indications are that Microsoft will hit one out of the park with Windows 7. Even in the netbook market, which was originally thought to be the perfect target for Linux, Microsoft has made some big splashes. Windows XP has proven to be a great fit for the space and Microsoft has plans for a application count-limited edition of Windows 7 aimed at this space. Time will tell whether or not the Windows 7 Starter Edition's three application limit will be a hindrance to users of the devices.
When it comes to simple mind share, Linux doesn't stand a chance against Apple and Microsoft. If anything starts to eat significantly into Windows' market share, it will be Mac OS X. Between constant TV ads from both companies touting their relative strengths and the fact that these two platforms are the only ones seen in places like Best Buy (you know, where regular citizens buy their computers), it's hard for many people to think of a desktop computer as running anything but one of these two operating systems.
In the enterprise, Windows remains firmly entrenched for a variety of reasons. First, most enterprises have relatively attractive licensing arrangements with Microsoft. Sure, Windows and Office are still expensive, but using these de facto standards means that document sharing and other common needs simply work. Second, people are used to Windows and Office. Even with the major changes in Office 2007, I personally still find it less foreign than OpenOffice, and I'm an IT guy with fifteen years of experience in Windows, Linux, NetWare, VMware, Office, OpenOffice, WordPerfect... you name it. The critical mass of people seem to feel that, although the ribbon in Office 2007 seemed like a horrid idea at first, once people started to use it, they found that they actually like it. Again, one could say the same thing about OpenOffice, but so far, migrating from Office to OpenOffice has not been that common, with a few high-profile exceptions. And finally, as I mentioned before, the companies that produce enterprise software are also helping to keep Windows entrenched in the enterprise by creating client software that runs only on this platform. I'll be the first to admit that Windows has its flaws, but consider the perspective of the software developers. Would they rather support a single client application on one OS or is there truly a need to develop a client for an OS base that has barely reach 1% of the total? From a financial perspective, the answer to that question is a no brainer.
Obviously, the server space is a whole different animal and, in this arena, Linux plays very well. Bear in mind that this space is handled by people with very high levels of technical knowledge that have a willingness to explore and work around issues that might arise with their desktop of choice. Most people, however, and particularly those in the enterprise space, use their computers as tools by which business is done. Time to tinker is a luxury and doesn't add to the bottom line.
Sure, Linux is "free" in that a distribution can be had for no up front cost. For some of the reasons I've mentioned above, though, this freedom only goes so far. At some point, probably sooner rather than later, someone will hit a wall and a workaround will need to be developed. This goes back to my point about "tinkering". Yes, Windows requires tinkering sometimes, but most things just work and people can focus on the output rather than the tool.
None of what I've written should be interpreted that I hate Linux and open source. Nothing could be further from the truth. That said, there is a time and a place for a particular tool and, for the foreseeable future, I don't see much of a place for Linux at the desktop level. Between mindshare, what seems to be developer difficulties to focus on end user needs, lack of software applications and other choices (namely, Windows and Mac OS X), there is simply little reason beyond "free" to choose Linux for general desktop use.