Goodbye Ubuntu phones, and other misguided Linux predictions

Canonical's Ubuntu phone project is dead, representing a failed market prediction for Linux. It's not the first.

Mark Shuttleworth of Canonical recently announced that the company would be abandoning its efforts to market Linux phones running Ubuntu. Also being dropped is the Unity desktop that brought the opposite of its name to Ubuntu's core user base. Perhaps 'Marmite' would have been a more appropriate moniker, so deeply did it divide opinion. Future versions of Ubuntu will return to Gnome in place of the convergence-oriented Unity.

I wrote an article in 2014 about Ubuntu's plans for a Linux tablet, mentioning the supposedly-impending Ubuntu phone launch. I wasn't alone in my enthusiasm for the idea, but in hindsight it was always going to be a small market. Apple and Google have such strong footholds that no other phone OS is likely to dislodge them in the near future. Bigger companies than Canonical have tried and failed to make much of a dent.

So, regardless of my personal disappointment, Canonical's change of direction makes sense. It also represents another in a long line of inaccurate predictions about Linux's future. It's worth revisiting some of the others.


Linux on the desktop

In the late 1990s and early 2000s, Linux was packaged in retail boxes with manuals and installation CDs, not unlike Windows. Industry pundits were confidently predicting that Linux would take chunks of business away from Microsoft (this list from 2001 is typical of many). I still have a boxed copy of Mandrake Linux from that era, and there were many others vying for the public's attention.

It didn't catch on. Whatever the arguments in favour, two major factors prevented its take-up. One, Linux couldn't run all the familiar Windows programs, and two, Windows was 'free' anyway when buying a new PC. Today the percentage of Linux desktop users hovers around 1-2% as measured by web client usage. That figure seems unlikely to be going anywhere soon.


Linux phones

Dead, as noted above, only that's not the whole story. It's true that there are no widely-available phones running a pure version of Linux, yet Android is based on a Google-modified Linux kernel.

There's all sorts of stuff on top of it, including Java, but drop to a terminal and you're in a Linux shell, with all the usual commands available. By this measure, Linux is one of the biggest mobile operating systems in the world. It just doesn't look like Linux.


Linux tablets

I have one of these, one of the first true internet tablets: a Nokia N810 from 2008. Nerds loved it, along with its successor, the N900. Surely Linux tablets would become an unstoppable trend?

Unfortunately, nobody else much cared, and that line faded out in step with Nokia's declining market share. Various attempts have been made to launch pure Linux tablets since then, so far without major success. But see the Linux phones section above: any tablet running Android has more than a little Linux in it.


Linux servers

Red Hat stamped its mark on the enterprise server market in the 1990s and is still doing so. With a few other Linux companies it took the view that providing support rather than just software was a sound business model. That turned out to be one of the better predictions in this list.

Linux also took on web hosting with the LAMP stack (Linux, Apache, MySQL and PHP). This free combination dominated web hosting for many years, though now the gap between it and Microsoft's server solutions is – depending on how you measure it – quite narrow.

'Server' now means so much more than a web server or corporate file/mail box in the backroom, of course. Cloud services are often hosted on Linux servers, with a wide range of [Something]-as-a-Service offerings delivered via an underlying Linux OS. Even Microsoft uses Linux for some of its Azure cloud services. In this respect the Linux server market has exceeded its proponents' wildest predictions.


Linux IoT

This is one direction in which Canonical's repositioning is designed to take Ubuntu, and it's easy to see the reasons why. Nobody mass producing millions of internet-connected 'things' is going to want to pay an OS licence fee for each one. With Linux they don't have to. See also: Raspberry Pi and similar tiny computing devices. So we can confidently predict that the future internet of things is going to be running Linux... possibly.

All of this serves to show that when it comes to free, open-source software written by a disparate bunch of talented people without a central board or committee, there's no predicting what will happen next. I suppose that's half the fun, but it's disappointing at times. As one of the small group of desktop Linux users, I was looking forward to owning a real Linux phone.


Also read:
Will Ubuntu’s Linux tablet shake up the market?