Linus Torvalds wants the Linux desktop. Too bad no one else does. This is never going to get any better, so let's give it up.
Linus Torvalds may still want a Linux desktop, but no one else does. And even if they did, by the time the requisite ecosystem could be developed, the need for a desktop — Linux or otherwise — will largely be gone.
So, why this persistent fetish for a Linux desktop?
Linus wants the Linux desktop
Linus Torvalds, founder of Linux, has a knack for provocation. The man who can blister bad ideas with furious broadsides on Linux kernel mailing lists also has a penchant for seeing past immediate technology roadblocks. But where the proverbial "Linux desktop" is concerned, Torvalds may have a blindspot.
At LinuxCon in Chicago this past week, Torvalds reinflated dashed desktop hopes with this comment:
"I still want the desktop. The challenge on the desktop is not a kernel problem. It's a whole infrastructure problem. I think we'll get there one day."
I don't. The briefest glance at market share data suggests that I'm not alone, either. While hundreds of millions of people want Linux powering their smartphones, and millions of businesses are content to let Linux run their servers, virtually no one wants Linux running their laptops and desktops.
What they do continue to want, however, is a better personal computer experience.
Still haven't found what we're looking for
Greylock venture capitalist (and former CEO of Mozilla) John Lilly calls this out in a recent blog post, arguing that "our software and systems were mostly not delivering on the promise" of personal computing. For effect (and because it's sadly true), Lilly quotes tech visionary Mitch Kapor, writing in 1991 about the sorry state of personal computing:
"Despite the enormous outward success of personal computers, the daily experience of using computers far too often is still fraught with difficulty, pain, and barriers for most people, which means that the revolution, measured by its original goals, has not as yet succeeded."
As Lilly notes, Kapor's lament is as true in 2014 as it was when he penned it in 1991. Or close.
While we in tech like to assume that technology, including personal computing, is easy, the opposite is actually true. It doesn't take long to figure this out, either. All you have to do is talk to people who aren't actively involved with technology each day. They're the ones that call you to ask how to print out a PDF, or install a router, or whatever allegedly menial technology task they're trying to perform to get through the day.
As Wes Miller, research VP at Directions on Microsoft, humorously tweeted recently:
Apple has been successful, in part, because it has reduced the complexity of personal computing. But even Apple has largely failed to overcome the PC's clunky interaction with human experience.
Could Linux do any better? Probably not. Linux, developed historically by and for geeks, may be the least likely candidate to improve the consumer experience. While companies like my former employer, Canonical, are focused on streamlining the user experience, a pretty UI is just one of many, many things that needs to change.
Torvalds hints at this when he says the Linux desktop problem is one concerning the "whole infrastructure" around Linux. In other words, the Linux desktop needs to remove all complexity associated with third-party software, drivers, etc. Jack Wallen covers three key things that business users would need from a Linux desktop, but once you factor in the more consumer-y things, like support for preferred music and app stores, Linux is a bridge that's much too far.
I ran a Linux desktop for years, including Fedora, Ubuntu, and SUSE distributions. Each had its share of glitches. None of these was something that I couldn't figure out with a little time searching Google, but most users don't have that kind of patience or even that level of expertise. It needs to "just work." The Linux desktop, quite simply, does not.
So, let's move on, just like the rest of the world has. No one outside geeky events like LinuxCon pines for the Linux desktop anymore. We should be content that they're pining for something even better: the Android smartphone, which makes the Linux "desktop" relevant for the next 20 years, even if it wasn't relevant for the last 20.