Slashdot Mirror
New Numbers on Linux Market Share Soon
prostoalex writes "New numbers on Linux market share are due this week. As far as global PC market is concerned, Gartner claims 5% of all PCs shipped this year ran Linux OS, although by the time the PCs were actually on the user's desk, only 2% of them run Linux. In the server world IDC estimates that Linux-powered servers comprise 28.3% of all server sales in 2004."
Wish I could edit my own posts - just as after I sent this I found an interesting article by Tim O'Reilly. He suggests using book sales to measure market trends.
I don't know if this will turn out to be accurate, but it's at least somewhat objective. A neat idea.
It is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail. - Abraham Maslow
"So while there are a small number of users purchasing their computers with Windows pre-installed and migrating to Linux, that number is effectively buried by businesses doing just the opposite."
MOST users running linux don't purchase a computer with it preinstalled. Since linux requires far less in terms of hardware to get the same or better performance usually they put a little memory in the computer they have and install linux on it. That includes businesses.
But what dwarfs the linux factors one way or another is the windows site licenses. For every desktop that an alternative OS was purchased on there are hundreds in which it was not, where they simply paid the MS tax despite their site license (like they've always done before).
Of course the difference in the numbers should be obvious, this could well constitute a pretty big chunk of the market, significantly reducing what is believed to the size of the market. This means x number of sales is really a larger percentage of the market than it is portrayed to be.
When it comes down to it, compared to legitimate copies, there really aren't that many pirated copies of windows... there just aren't. While they are common among techs, giving us the impression they are rampant, in reality I'd be surprised if techs and their families amount to even 1% of the market.
The kid/teenager of the house isn't an OS installer anymore than his parents or grandparents. He is wise in that he can successfully work the mouse and install most programs... a far cry from a pirated OS installation.
With linux on the other hand, there is a strong prevalance of technically literate users (the reasons for this are debatable and not the issue here). Almost every linux user can install the OS. Couple this with the fact that companies normally act as if linux is a "cheap and inferior" solution. Normally the pc's that come with linux preinstalled are in the $200-300 range and worth more like $150, they are usually crap a literate user wouldn't touch.
Aside from the price on the pc's, I fully admit I'm educated guessing the numbers. But from what I've seen... well I've never actually seen a system with linux preinstalled on it. I've seen lots of linux systems mind, many I've setup and have lots of friends using linux. Most of their computers are homebuilt (but not all). All in all, among desktop users I'd guesstimate about 200 linux pc's. Not a single one of them would be counted in these numbers.
In the businessworld it's much the same. Support contracts are an issue for obscure software only in small businessland. Corporations want accountability, small business wants it to work and wants someone to call to fix it when it's broke, they don't care about fingerpointing.
On the business side I've setup countless workstations and several hundred linux servers. Out of all of them only one was even a purchased license, all the rest were download editions of the software. A support contract would be pointless, if they have a support contract it's still us they call if they have a problem, we are local and can fix the problem before they finish holding.
You also don't need to buy a boxed version for updates. Really using the vendor update mechnism is probably the last thing I'd recommend to a customer. With redhat distros in particular, redhat drops support too fast and is slow on the updates in comparison with well known and trusted 3rd parties (*cough*freshrpms*cough*) who still provide updates for redhat version 6.2 last I checked.
5% of the desktop market, I doubt it's that low. 5% of oem preinstalls, perhaps. As for whether it had that OS on it when it hit the desktop, if you consider that, you have to consider all the rest I've mentioned above and more and the result is the desktop market, not the OEM preinstall numbers gartner is claiming.
I still sometimes get a little pissy about articles/reports/surveys like this. Then after I spout my peace, I can't help but laugh. Linux has an advantage that most commercial desktops can't even approach. See, if I sell Windows, Solaris, Mac, or any of the other commercial operating systems, I know exactly how I'm doing.
We all "know" MS has 90-95% of the market. The numbers shipped, the dollar amounts, all point to this "fact". Same with the others.
Linux doesn't come close. From a dollar perspective, most distros pale compared to the others. From a "shipped" point of view, well, who counts little Linux shops in their numbers? This is about Dell and the big folks. And there is the dualboot/wipe issue.
So, why do I laugh? Because, using these statistics, nobody will realize how many people actually use Linux until it's right in their faces. In other words, theoretically, MS could still ship 90-95% of the market, only to turn around one day, and find out that only 10-20% of users actually use Windows (with a few more using it occasionally).
So, realisically, the better way of measuring this would be to measure the "other" sales related to Windows. Antivirus software wouldn't count, neither would Office software, or games. (These are necessities for dual booters, or things that might only be available for one OS.)
My pick would be the "cheapy" software that people tend to buy for their computers. The productivity stuff, or "make your computer easier to use" kind of stuff. Better yet, if you want a long term guage, try the "educational" aisle.
In other words, to guage an OS's success, compare it's market. Find something unusual about that OS, something that no other can share, and use that as a guage. Exact numbers don't matter, but trends can point out a lot.
If money/users seem to be disappearing from these markets, yet the hardware folks are actually doing pretty well, you might want to bump up your Linux/OSS numbers a little.
Microsoft has just released their much anticipated hands-free cordless mouse. Warning, it may hurt a little at first.