How many Linux users are there really?
It’s a darn good question, and there isn’t a darn good answer.
In one way, we’re all Linux users now.
As Jim Zemlin, the executive director of The Linux Foundation, points out, “I am not joking or trying to be trite, but the answer to this question is: every single person in the modern world every day. Everyone who searches Google, picks up a phone and uses telecommunication infrastructure, watches a new televisions, use a new camera, makes a call on many modern cell phones, trades a stock on a major exchange, watches a weather forecast generated on a supercomputer, logs into Facebook, navigates via air traffic control systems, buys a netbook computer, checks out at a cash register, withdraws cash at an ATM machine, fires up a quick-boot desktop (even those with Windows), or uses one of many medical devices; the list goes on and on.”
“It is hard to think of someone in the developed world who doesn’t touch Linux every single day. The better question here is who isn’t a Linux user,” Zemlin concluded.
He’s got a point there. If you buy something from Craigslist or keep up with friends on Facebook, you’re using Linux. To be exact, you’re using Big-IP 9.4.6, which is an embedded high-speed networking system that incorporates Linux. Do you watch videos on YouTube? Linux again. Google? Yes, they run Linux too.
Gartner and IDC Don’t Count
By this yard-stick, everyone on the planet who uses the Internet is a Linux user. While certainly true, it is only one way of measuring Linux use. The major analysis firms that track operating system use, Gartner and IDC, focus on the server market. Their numbers ‘indicate’ that Linux is a powerful player in the server market, but their research doesn’t translate well into giving us an actual number of Linux servers.
Gartner, of late, tends to focus on server hardware sales. So, while we know that IBM and HP both sell Linux-powered servers, Gartner’s data doesn’t easily tell us how many of those systems are shipping running Linux. IDC, on the other hand, tends to look at the overall cash value of server sales. That’s valuable information, but, since Linux tends to be the cheapest server operating system offering, this methodology also under-represents Linux’s impact on IT.
In both cases, the big two research firms are only looking at server sales. They’re not looking at what IT departments are actually running in their server rooms. In my own consulting experience, most of the SMB (small-to-medium sized businesses) use re-purposed x86 desktops and servers to run Linux.
They can do this because Linux is cheap. With Linux, any PC from the last eight-years makes a perfectly acceptable small business file and print server; Web server, VPN (virtual private network) server, etc. etc.
How do you measure Linux’s low-tech server impact? It’s not easy. For every corporate customer running the latest RHEL (Red Hat Enterprise Linux), they are probably a dozen businesses running Fedora or CentOS.
And, then there’s the desktop. Here, the numbers get truly fuzzy. We know, without a doubt, that there’s are tens of millions of Linux servers in use. But when it comes to the desktop, we just don’t know.
It would appear that a good rule of thumb would be the number of user desktops that report they’re running Linux to Web server logs. For example, we often cite the percentage of desktop Linux users as being under 1% because Net Applications Web site survey shows, in its January 2009 survey, that Linux has only 0.83% of the market.
Can You Trust Net Applications’ Numbers?
The question that isn’t often asked though is: “Can you trust Net Applications’ numbers?” According to Roy S. Schestowitz, editor of Boycott Novell, the answer is: “No.” According to a recent Boycott Novell blog, “Microsoft and Apple put money on Net Applications’ table, so rather unsurprisingly, the results satisfy both companies. GNU/Linux, on the other hand, is not able to pay Net Applications for favorable bias.”
And, in addition, to other points Schestowitz writes, “Net Applications admits its statistics are flawed (skewed)” and “Net Applications keeps its methods secret and the dataset likewise.”
Whether Net Applications is unduly influenced by Apple and Microsoft is outside the scope of this story. What can be said though is that Net Applications’ desktop numbers certainly can’t be taken as Gospel.
In my own research of the Web traffic history of computer, but not Linux-oriented, sites, with millions of hits per month, I found that Linux desktops averaged at 6.8%. That number is biased towards technically adept users, and not the general population. That said, I think it’s a more accurate number for the computer-literate population than the Net Application numbers are for the general population.
That said, with the rise of netbook sales, Linux is getting into many more hands. For example, ASUS CEO Jerry Shen said in October 2008 that ASUS would hit its expected number of 5-million ASUS Eee netbooks sold. If three out of ten of those netbooks are running Linux, as the company has said, then that alone has brought 2-million more desktop Linux users on board in 2008 alone.
Combine that with Dell and other major OEM (original equipment manufacturer) pre-installed desktop Linux sales and the Linux hardcore who install their own distributions, and I think that 10-million desktop Linux users is a reasonable estimate.
Of course, compared to Windows users, that’s not a big number. Linux’s millions though is a significant number, and not just to Linux users. In its latest quarterly announcement, Microsoft admitted that, “Client revenue declined 8% as a result of PC market weakness and a continued shift to lower priced netbooks.”
If the Linux desktop is big enough to be hitting Microsoft’s bottom line; it’s certainly big enough to matter.
But, we still don’t know exactly how many Linux desktop users there are… yet. Cole Crawford, a Dell Linux engineer, is working on the problem. He has created statix to anonymously track the number of Linux desktops worldwide.
Statix is a client/server program. It uses both a Python client and a hosted Python CGI (Common Gateway Interface) back end to track Linux desktops is running. Eventually, it will also be able to track the kernel and distribution of statix-using Linux desktops. The client will not send any personal information about the user back to a centralized database. It’s simply a mechanism that can be used by any Linux distribution to say “Linux desktop user here.”
This is a purely open-source project and needs your support. With just a bit of work, statix will finally be able to give us a hard answer to that annoying difficult to grapple with question: “How many Linux users are there anyway!?” I, for one, would really like to know the answer.
Related Posts: On this day...
- Undercover TSA Agent smuggles gun through scanners multiple times - 2011
- Epic Beard Man interview - Thomas Bruso - 2010
- URL redirection service makes innocent URLs look sinister - 2010
- Buying movies vs pirating movies - 2010
- Google Maps face-blurring technology in action - 2009
- OpenX: More Than Just an Ad Server - 2009
- How fanboys see their respective worlds - 2009
- Time Magazine from 1999 - 2009
- Body scanners replace metal detectors in tryout - 2009
- HOWTO: Making a USB Key Bootable in Linux - 2008