Off the Beat: Bruce Byfield's Blog

Unity, Ubuntu's default desktop environment, has always been ironically named. Begun after conflicts between Ubuntu and GNOME, Canonical Software and Ubuntu developed it as a solo effort after other projects ignored Mark Shuttleworth's famous challenge to develop an interface to rival Apple's. Moreover, with less than a six percent market share in the 2016 LinuxQuestions survey, it remains among the least popular of Linux desktops.

Under these circumstances, the only surprising aspect of Shuttleworth's announcement that Ubuntu was dropping Unity in favor of GNOME is that the decision took so long. Still, it's a gutsy decision that allows Canonical to concentrate on servers and OpenStack, the only divisions of the company that currently seem capable of turning a profit.

When Shuttleworth issued his challenge in 2008, a focus on interfaces seemed a sensible next step for the Linux desktop. At that point, the Linux desktop had matched proprietary desktops like Window for functionality, thanks in no small part to the early Ubuntu releases. However, by the time Shuttleworth made his announcement, both GNOME and KDE, the main Linux desktops, had already started their own work on interfaces. As a result, Unity was never the unique improvement that it would otherwise have been.

Even more seriously, despite an ingenious widget or two, Unity never did manage to live up to the original hopes for it. Contrary to original intentions, Unity was never easier for new users than any of the other desktop alternatives. To the contrary, when more than two windows are required, it is the most cumbersome interface since Windows 3.1 -- especially on a small screen.

Part of the problem were some arbitrary decisions, such as moving the title bar control buttons to the left side and displaying menus in the top panel. Although there was some vague talk of freeing the title bar for other purposes, these purposes never materialized. The changes only made sense if you were familiar with such arrangements in OS X.

A similarly arbitrary decision was to lock the launcher on the left side of the screen. The positioning was sensible enough, given the wide screens that were becoming popular at the time, but not the insistence on locking the launcher in one position. Unsurprisingly, utilities to change the launcher position soon followed.

Other choices proved equally controversial. A Contributor's Agreement that allows Canonical to re-license the code guaranteed that Unity would remain a largely in-house project. The heads-up display, intended to replace menus, was only partly implemented, leaving major gaps in its functionality uncorrected. Even worse, the decision to add Amazon search results by default to search results was widely condemned as an invasion of users' privacy.

All these issues created divisions between paid Canonical staff and Ubuntu volunteers -- especially when Shuttleworth used his veto power on behalf of the staff, demonstrating how little control volunteers actually exercised. At times, user revolts seemed close, although they were apparently prevented by belated diplomacy.

Unity and Ubuntu Touch
In the last five years, Unity development was focused on convergence, the development of a common interface for everything from phones and tablets to laptops and workstations. From a technical perspective, this goal seemed promising. On a small touch screen, Unity seems far more efficient than on a twenty-three inch widescreen monitor.

Unfortunately, Ubuntu shipped phones and tablets with convergence incomplete. In Ubuntu Touch, the version of Ubuntu developed for such devices, plugging in a keyboard, mouse or monitor gave almost the convenience of a laptop, but the ability to install desktop packages was never made simple enough for average users to take advantage of the feature.

In addition, like many first time manufacturers, Canonical and Ubuntu had trouble breaking in to markets that were already heavily saturated. They chose to partner, but none of their partners seemed particularly interested in promoting Ubuntu Touch devices. In the case of bq's Aquaris M10 Ubuntu Edition tablet, it didn't help, either, that the hardware was mediocre.  Although the Aquaris M10 tablet eventually sold out, a second production run was never made. And, like the Ubuntu Touch phones, the Ubuntu Edition tablet was never distributed in North America.

With such problems, convergence never did manage to catch buyers' or manufactures' imaginations. Had Ubuntu Touch been available five years earlier, perhaps convergence might have become an asset, but, by the time Ubuntu phones and tablets were available, the market had become accustomed to having a different interface on each device. Even an application like KDE Connect, which helps Linux and Android devices to interact, has not received nearly the attention that might be expected.

Under these circumstances, the last reason to continue Unity development has failed. Sensibly,Ubuntu is switching to GNOME. Considering Ubuntu's previous discontent with the GNOME project, and the fact that the switch will not happen for a year, I suspect that Ubuntu will carry a forked version of GNOME, although all that remains to be seen.

Already, a fork of Unity has been announced. Like Trinity, the fork of the KDE 3 release series, it may continue to be used by a few percent of users. But otherwise, after a decade marked by dis-unity, Unity's day appears to be done.