I took an inventory of the computers and operating systems in my household, and found that my wife and I have a total of nine copies of Windows, one copy of MAC OS 9.1, and one purchased Linux distribution. But this doesn't tell the truth about our computer use patterns any more than the recent Gartner Group survey -- the one that shows Linux servers with only 8.6% of the market -- tells us how many servers really run Linux.My household currently owns a total of 10 computers, but six of them are obsolete units sitting out in our storage shed. All of the old boxes came with Windows pre-installed. We have a total of four computers actually hooked up and operating; three laptops and one desktop. One of the laptops -- my wife's -- is an iBook running Mac OS 9.1. The one desktop came with Windows but now runs nothing but Linux. One of the two Intel laptops was purchased with no operating system, and has never run anything but Linux. The newest laptop -- the HP Pavilion 5340 I am using to write this article -- came with Windows ME installed. I immediately installed Linux on it but have not yet deleted the Windows partition. I have a retail copy of Windows 98 (SE) I purchased back in 1999 while researching a story, and it is on my shelf, not currently installed on any computer. And I own only one purchased copy of Linux, Mandrake 8.0.
Go ahead. Do the math. We own a total of 10 computers -- and nine copies of Windows.
But we have three copies of Linux installed and working, despite have purchased only one (a trick that would violate federal law if we tried it with a proprietary operating system) as opposed to only one installed copy of Windows -- that I have only booted up a few times, ever.
Suddenly the numbers look different. Now, counting only installed operating systems, we show 60% Linux, 20% Mac, and 20% Windows. We could even take things a step further and count only "operating systems in daily use," which would show our house running 75% Linux, 25% Mac, and 0% Windows.
Here's the key phrase in the Gartner Group survey:
"The study results indicated that in the traditional server market in the United States during the third
quarter of 2000, 8.6 percent of server shipments were Linux-based systems."
As my colleague Jamie McCarthy points out, you could use Gartner Group's logic to say, "100% of babies born in this country had unpierced ears, therefore, pierced ears are a rarity."
I suspect that an awful lot of servers -- and home computers -- get counted as "Windows" machines because that is how they were sold, even if they now run Linux; or *BSD; or BeOS; or (yes, it's still around) OS/2.
Then comes the question, "Why would anyone buy a Windows-loaded computer instead of one with Linux on it?" Answer: the HP Pavilion has a case and keyboard I find exceptionally comfortable to use, a bright screen and a 20 GB hard drive, and was an excellent value (on sale) for $1,745. The fact that it came with Windows instead of Linux (and has some features that apparently are designed to work only with Windows) is annoying, but it was still the best unit I could find, per dollar, for my particular needs. I'm sure many commercial buyers buy Windows-loaded servers because they get the best price by purchasing low-end, stamped-out units preloaded with [whatever] than they would if they held out for servers built to their specs -- including their preferred operating system.
It is not hard to install Linux. And it's not hard to install RAM, either; if I can buy additional RAM for half the price the original machine manufacturer wants for it, I'll get it and put it in myself, thank you. Most of the high-end commercial computer users I know (including people who run ISPs and hosting services) are willing to install RAM and operating systems themselves if this will save them $200 or $300 per unit.
I understand why Microsoft is scared of Linux. Windows workalikes for the Linux software I use daily, all of which came with Mandrake 8.0 in the "Powerpack" box that cost $65 (minus a $20 mail-in rebate), would run well over $1,000.
And I understand why Microsoft would help sponsor a Gartner Group survey that shows Linux server usage much lower than other server operating system surveys have shown it to be, really I do.
But I also understand how statistics can be manipulated to give false impressions. And because I am perfectly happy with Linux, I don't think I'll switch to Windows anytime soon, no matter how many times I hear that Linux is a "hobby" operating system; or that not as many people use it as we think; or that the Linux desktop (which seems perfectly lively to me) is dead. Or whatever bad thing about Linux the people in Redmond are saying today -- or say next week or next month.