November 29, 2011, 12:47 PM — Repeat after me:
Don't use DistroWatch as a measuring stick in any way for the popularity of a Linux distribution.
Seriously, stop it.
In fact, why are we even asking the question at all?
"Popularity" is a term that smacks of our days in high school, when we thought we should care about social standing and where we fit in that ranking. Now apparently, we seem to be locked into this notion of figuring out which distro is most popular, too.
This is a silly question, for multiple reasons.
First, the real data is hard to get. There is no common download tracker for distros. If you think DistroWatch is it, think again. DistroWatch doesn't count downloads or boxes: it counts page hits for each distro's page on that site. "Only one hit per IP address per day is counted," the site explains.
Which means that anyone coming to DistroWatch to find out about what distro works best for them will see the rankings list and, thanks to positive feedback, most likely click on the distros that are near the top of the list--thus keeping those distros on top of the list.
DistroWatch's ranking system has never been indicative of true installation numbers. Here's what Ladislav Bodnar, the owner of DistroWatch, said in 2006:
"There is absolutely nothing flawed about DistroWatch page hit ranking. It represents page hits on the individual distribution pages, period. Now, if you'd said that the DistroWatch page hit ranking is a flawed way of measuring the popularity of Linux distributions, then yes, I would have agreed..."
Would I, as a journalist, like to find out which distros were indeed the most used? Sure, I'm a sucker for numbers just like any other geek. But the second reason the notion of distro "popularity" is so completely silly in Linux these days is because I can run virtually any application on any distro. In the old days, when differences between distributions were larger, that might not always be the case. Today, even though the Linux Standard Base seems to be given little more than lip service by distribution developers, most applications can run on Linux with little to no porting.
And those applications that can't can use virtualization to run on a client or server box anyway.
Interfaces are not an issue, because I can run any interface I want on any distribution. Yes, it's nice when a stock distro runs a great interface I like right out of the proverbial box. But if it doesn't, who cares? I'll just tweak the interface until it does, making three reasons why distro popularity is moot.
The final reason? HTML5 and other web-based applications. While I believe there will always be a need for native applications, we cannot ignore the migration of software tools from the desktop to the web. The more applications that come to us as Software as a Service, the less need there will be to differentiate between distributions.
Are all distribution comparisons worthless? From a technical standpoint, no--like any software, some distributions will always be put together better than others. And some distributions, by their nature, may be better for some audiences than others.
Distributions, I think, will end up going one of two ways. They will either be "legacy" distros that will have the package management and configuration tools to enable users to trick out their platforms as needed with greater and greater ease. Or, they will be niche distros, designed to perform certain functions or satisfy a certain customer.
If this segmentation does come to pass, then it will be pointless to compare what will essentially be platforms consisting of modular, swappable apps and interfaces… or distros will be too niche-driven to make fair comparisons.
We can highlight innovation and progress, but the nature and maturity of open source really means that the distro wars are over.
Read more of Brian Proffitt's Open for Discussion blog and follow the latest IT news at ITworld. Drop Brian a line or follow Brian on Twitter at @TheTechScribe. For the latest IT news, analysis and how-tos, follow ITworld on Twitter and Facebook.