I normally stay out of the Linux conversations because it's like placing oneself between two packs of rabid, fanboy wolves. Not that being enthusiastic about your operating system of choice is a bad thing. It's just a lot of flame to handle for one meager columnist.
That said, I couldn't help but notice a number of articles passing around the Web this week, praising Linux for pushing past the one-percent adoption rate for desktop operating systems. Huh? One percent? That's like throwing a ticker-tape parade for a one-year-old. I mean, kudos to Linux for making it this far and all, but I think that people are selectively focusing on the "concept" of the number a bit too much. Because when you dig a little bit deeper into the statistics, you'll find that Linux's big "Achievement Unlocked" isn't really that big of a deal at all.
When we talk about Linux, it's often just that. Linux. Be it Linux for your desktop or Linux for your server farm, Linux is Linux. Would you run Windows XP on an enterprise Web server? No. You would use an operating system featuring a different name, one that's sure to distinguish it from Windows' consumer-focused versions. But what about Linux? You'd run some variant of Linux, definitely, but you'd be correct in just calling it Linux.
Why is this important? Because when reports come out that suggest Linux has a one percent adoption rate, you might be tempted to write that off as Linux in its entirety. In actuality, the figures that Net Applications are quoting only consider the operating systems of typical desktop machines. While a number of reports on the data adequately disclose this fact, it's still an easy mistake to make if you just happened to scan a headline and see, "Linux at One Percent!"
And of these figures, you have no idea of the sample size, nor the profiles of the individuals being sampled. I'm not suggesting at Net Applications is wrong, and that the figure for Linux desktop adoption is in the twenties or something. Nevertheless, other sites that survey batches of users peg the Linux adoption rate a little bit higher than one percent. Are we surveying your average Best Buy shoppers? Are we surveying IT geeks? Who's right? I don't know. And another question to throw into the mix: How might this number change if we toss laptops and netbooks into the picture? And how does the survey qualify dual-boot users?
Going back to my original point for a moment, the one-percent figure somehow suggests that Linux is barely scraping by. While this might be true on consumer machines, I think it's erroneous to ignore the impact of Linux in the enterprise market. IDC estimates that Linux-based enterprise software sales will reach $35.5 billion by 2013. While that's still quite a ways away from the projected Microsoft-based software sales--206 billion--the growth rate for Linux software is expected to outpace the general market growth by five times. Microsoft software sales will barely keep pace.
It's difficult to break this number down into the raw, mano-a-mano competition of total Linux server environments versus total Microsoft-based server environment in the enterprise market. But I can guarantee that the number of Linux servers in existence is greater than one percent. And were we to take enterprise hardware into account, I believe that Linux would appear to have a far greater showing than the paltry one percent (or thereabouts) figure quoted for the desktop market. I realize that wasn't the point of the original survey. The original survey cared about desktop machines and desktop machines only. But why are we limiting to this isolated definition of Linux OS installs? It would sure be nice to have a little bit more to celebrate, that's all. I mean, when's the last time you saw a pre-built desktop machine being offered with a Linux OS instead of a Windows-based one anyway?
More interesting to me than Linux's alleged one percent adoption rate are the growth curves presented by Net Applications' data. It took Linux nearly a year to carve out an additional .2 percent of the market. At this rate, we'll all be dust in the wind by the time Linux takes one-fourth of the desktop market share, assuming the figures keep tracking along the same growth pattern. While the adoption rate of Microsoft Windows fell nearly three percent over the past ten months, Apple became the leech for 2.8 percent of those users. Were we to lump all of the Windows refugees into a pile, Apple would have pulled 86 percent to Linux's 14 percent. That's practically a six-to-one ratio.
The most compelling story isn't Linux's big jump into single-digits (on the left side of the decimal, that is). I'd keep my eye out on Apple if you want to see evidence of solid, market-affecting growth. Outside of the enterprise market, Linux just isn't growing in the mindset of your average consumer--provided the numbers and survey samples match up, that is.
So what's the takeaway? Don't believe everything you read (Desktop Linux adoption rate is at one percent), don't ignore a sizeable market to assume the sky is falling (Enterprise Linux is pretty big), and even if the numbers all pan out in the end, don't believe that Linux's one percent accomplishment is anything significant (Time to buy Apple stock?).