Can Linux's market share be adversely affected by the competition?

Rodney Gedda

Rodney Gedda is the former deputy editor of CIO and former editor of Techworld.

In an article about the Linux kernel I wrote last week I threw in a few lines about Linux's market share and how it may have pressure put on it by Windows 7. It would be good to know if that's fact or fiction. Some commentators instantly labeled the article as being biased. Of course, that's just not true, but such claims prompt me to speculate how much merit market share data really has.

As always, I welcome people's feedback.

There's no doubt market share statistics have some merit. If you gather the operating system and Web browser details from visitors to your Web site then you have some idea of the composition of the operating systems people use.

It's not perfect, of course. A known problem would be the type of Web site being visited. If the Web site is Apple-related then you can bet a good proportion of the readers will be surfing with Mac OS X, and vice-versa.

If you distribute the statistics across a multitude of sites and exclude the technology specific ones, then you come closer to an “average” value of operating system use among the Internet-using population.

Before someone highlights the other problem, let me explain it first.

Most market share data for operating systems thrives on Internet connectivity. There may be many thousands of Linux installations on computers without regular Internet connectivity. However, this is becoming less common as inexpensive wireless networks pop up all over the world.

Back to Linux's market share and the advent of Windows 7.

When I wrote: “The advent of Windows 7 in October may drive Linux's desktop market share down even further” I didn't mean Windows 7 will take over Linux installations (it may, but I suspect only in very small numbers), rather it will put downward pressure on the existing Linux install base as a percentage of the total.

If the pie is larger, and more of the pie is being taken up by Windows and Mac OS X, then where does that leave Linux?

Linux gets “marginalised” even though the total number of Linux desktop installations actually increases.

Here's a basic example. Say I have 100 berry plants. Of those, 20 are blueberry plants and 80 are raspberry plants so 20 per cent of my berry plants are of the blueberry kind.

One year I plant 10 raspberry plants ( a new variety) and two of my raspberry plants die and, instead of replacing them, I plant blueberry plants in their place.

So the actual “market share” of blueberry plants has decreased even though blueberry plants replaced some raspberry plants.

The fact that a new release of raspberry plant (or operating system in the case of Windows 7) came along, and because it already had a dominant position in the market, ended up expanding the market size doesn't mean blueberry plants won't replace raspberry plants and still decrease in market share.

I do believe Linux will replace a lot of Windows installations after Windows 7 finally appears. That's not what Linux is up against right now.

A new Microsoft operating system comes along, many more computers are sold with it pre-installed, and the overall “computer pie” grows with Windows increasing the size of its piece. That's what Linux is up against.

Suddenly my “downward pressure” comment wasn't so biased after all.

But, as I've said before, there are a few key things coming together that will make Linux on the desktop more attractive going into 2010. Everything from device drivers to Twitter tools.

In any event, we can expect Linux, Mac OS X and Windows to do some tough competing this 2009-2010 season.