August 23, 2010
by sjvn01
0 comments
As everyone who ever tried it knows, trying to count how many people use a particular Linux distribution is almost impossible. Now, Rick Lehrbaum, founder of LinuxDevices and a friend and former editor of mine, has tried a new and interesting way to count Linux users on his new site, LinuxTrends: look at Google search results for the various Linux distributions.
Some of the results aren’t surprising. Ubuntu has become far more popular than the other mainstream distributions of 2004/2005: SUSE Linux, Fedora, Debian and Mandrake/Mandriva.
Still, these distributions’ decline to the Ubuntu juggernaut is nothing like as bad as the fall in popularity seen by the second-tier distributions of 2004/2005. Of Slackware, Gentoo, Arch and CentOS, only CentOS, a RHEL (Red Hat Enterprise Linux) clone, has retained much of its popularity. Newer second-tiers Linux distributions that are well-regarded such as PCLinuxOS, Puppy and Sabayon also aren’t doing well. The one exception is Linux Mint, which is an Ubuntu-based distribution.
This reflects what Sean Michael Kerner, one of my fellow Linux journalists, and I also have seen. When we recently got together at LinuxCon, we agreed that there’s a lot less interest these day in the minor-league Linux distributions than there was in the mid-2000s.
Like any attempt to count Linux users, Lehrbaum’s study doesn’t tell the whole story. For example, do you count Google users? If you do, and that means almost everyone on the planet with a net connection, then Goobuntu, Google’s server brand of Ubuntu Linux, takes the prize as both the world’s most popular distributions and one of the most obscure.
As I hashed out with my fellow writers at the LinuxCon Media Panel, trying to determine how many Linux users are really out there is almost impossible. I’ve also wondered just how important it is that we have Linux numbers.
I know, I know. As a baseball fan, I get that we all love numbers. Let me put it this way: I think we can all agree that Linux has become more important than ever. If you don’t buy that, ask yourself when the last time was you used Google, Facebook or any of the major social networks. If you use any of them, congratulations: you’re a Linux user, albeit once removed. By that standard, I’d say Linux was batting somewhere about .998.
If you’re looking at servers, which is where most of the analyst firms get it wrong, you have to keep in mind that these groups are only counting new servers with their installed operating systems. These organizations don’t even make an effort to count on those older servers, or PCs converted into servers, that are running Debian, openSUSE, or CentOS.
On the desktop, Linux has never done as well as its fans would have liked. At the same time, more people are using it on the desktop than ever before. Keep in mind that even though the overall percentage of Linux desktop users doesn’t appear to be going up, there’s ever more people using desktops of all sorts … for now, anyway. I still think that the rise of Linux-powered smartphones, TVs and tablets may yet make the whole issue of the desktop moot for casual computer users.
In the meantime, I can agree with Lehrbaum’s finding that, in so much as the Linux desktop is concerned, Ubuntu is number one. On servers, I’m sure it’s RHEL. Will the leader board still look that way in another five-years? Good question, and I don’t have a good answer. It will be interesting to find out.
A version of this story was first published in ComputerWorld.
