Last week, founder Mark Shuttleworth opened the Ubuntu Online Summit with a challenge to Linux desktop developers.
“I’m issuing a call to people who participate in every desktop environment,” he said, “to set aside our differences, to recognize that the opportunity now is bigger than those differences, to create experiences that spans phones and tablets, and PCs, to bring all of our applications, none of which are on one desktop environment or another.”
His words were rhetorically stirring — and provoked no major response whatsoever. Although some news sites reported his words without comment, probably most companies and projects have heard too many similar calls to action for this one to be effective.
Falling on Deaf Ears
The problem is not just the fact that Canonical, Ubuntu’s commercial arm, has gone ten years without turning a profit, making it a less than desirable business partner. Nor is it simply the fact that Canonical has had several failures in product development, such as Ubuntu TV or crowdfunding the boutique phone Ubuntu Edge. Increasingly, such concerns may be affecting reactions to Canonical and Ubuntu, but from the whispers I sometimes hear, I suspect that Shuttleworth’s peers have simply grown suspicious of his motives.
Over the years, Shuttleworth has made several calls for co-operation. In 2008, he called for free software projects to synchronize their release cycles. A few months later, his challenge was to develop an interface that rivaled Apple’s. However, as Aaron Seigo points out, while such goals sound sensible, they sound like code for everyone else in the free software industry to accept Canonical’s priorities and leadership.
The same suspicions might be applied to what Shuttlworth calls convergence — the development of common interfaces across everything from workstations and laptops to tablets and phones. Many, such as the KDE project, regard convergence as a low priority.
So far as anyone besides Shuttleworth is interested in the goal, they are already working towards it using their own interfaces. Some common code for hardware compatibility might be shared, but what could co-operation mean except working on a common interface, which presumably means Ubuntu’s?
In fact, given Canonical’s past record in innovation, the goal of convergence might be dubious. Ubuntu began strongly by tidying up the Linux desktop and making features the norm that should have been implemented long before. However, its more recent record is full of innovations that seem low-priority — even unnecessary — if the goal is improving the Linux desktop. For example, Upstart to replace init, Unity to replace GNOME, the Heads Up Display to replace menus, Mir to replace Wayland as an X Window replacement, and, more recently, Snappy packages to replace .deb packages.
With this record, why would the rest of the industry accept Canonical’s judgment about where their efforts should be spent? Too often, Canonical’s priorities have seemed questionable at best.
Moreover, these efforts to innovate have frequently taken place in-house — and may of them have ended with Canonical making similar calls to help. Canonical seems to have a history of taking on more than it can manage, then calling for the community it avoids to help from its difficulties.
In addition, these projects can be seen as attempts by Canonical to control key projects. Unity, in particular was born out of Canonical’s impatience with the give and take of working with the GNOME project, and Shuttleworth’s inability to convince GNOME developers to share his vision. As a kindergarten teacher might say, Canonical seems to have trouble playing with others and sharing.
As if that was not enough, another of Canonical’s failures was Project Harmony, an attempt to develop principles for contributors’ license agreements for free software projects. Canonical’s own contributor’s license agreement was widely denounced when first released as contrary to the spirit of free software, because it gave Canonical complete control over contributed code, even allowing the possibility of re-licensing it.
When all these things are considered, the silence from the rest of the community seems understandable. Responding to Shuttleworth’s call to action sounds very much like a call to accept his priorities, to develop his way, and under his leadership. None of these sound compelling, nor even in the interests of any organization other than Canonical. Considering that Canonical has even managed to rouse its Ubuntu volunteers into revolt, the case for accepting its leadership is questionable at best.
Popularity vs. Community
Canonical does have a solid claim for being the most popular Linux distributions. Admittedly, Linux Mint may appeal more to experienced users, but Canonical’s popularity remains strong.
Given this popularity, Shuttleworth often talks as though the interests of Canonical and Ubuntu are identical with the interests of Linux and free software in general. Like his previous calls for action, this latest one seems more of the same. However, this identification has never been one that a majority of the free software community has accepted.
Rationally, cooperation on a single desktop and goal might seem the way to go. In practice, though, many free software developers have other goals than commercial success, such as offering users diversity. Many have spent years developing their chosen software. Most seem unlikely to abandon their efforts easily — and definitely not for goals that they care little about or to follow a leadership whose record is so spotty.