Canonical founder Mark Shuttleworth is re-taking leadership of the company as Jane Silber steps down from the role of CEO and joins the board of directors.
As he returns as CEO, Shuttleworth has unique challenges ahead, as well as the opportunity to give Canonical a sharp focus and a lean structure that may attract investors (a move he is reportedly exploring).
Here is what Shuttleworth should do to as he takes Canonical into a new direction.
Divert resources from desktop & mobile: The year of ‘desktop’ Linux or mobile Linux will never come; Linux is already there with Android and Chrome OS. As Henry Ford said, “If I had asked people what they wanted, they would have said faster horses.” Canonical should stop wasting resources waiting for the ‘faster horse’ (i.e., Linux on desktop and mobile) and divert resources towards enterprise.
Create Ubuntu Enterprise: Canonical should follow the lead of SUSE and Red Hat and create a subscription-based enterprise version of Ubuntu. Ubuntu Enterprise may have a lifespan of ten years and come with commercial support plus a private repository of enterprise applications.
Create Ubuntu Community Edition: Creating a subscription-based enterprise edition of Ubuntu doesn’t mean Canonical should ditch its massive user base. Instead, the company should create a free of cost version of Ubuntu that has all the features of enterprise, except for a shorter life-cycle, no access to commercial support or private repositories.
The Community Edition of Ubuntu will be sponsored by Canonical sponsored, but a project managed by the community, which will also serve as the upstream for Ubuntu Enterprise, just like Fedora and openSUSE are upstream for Red Hat Enterprise Linux and SUSE Linux Enterprise.
The Community Edition of Ubuntu can be managed by a foundation, which is a mix of community members and Canonical employees. Once again, Shuttleworth doesn’t have to go any further than looking at the openSUSE Board for a model.
It will be a win-win situation for both Canonical and the larger community: Canonical will benefit from the project as upstream for Ubuntu Enterprise and the community will have more say and control over the project that they love.
Having a Community Edition may also solve the problem of having too many forks and unofficial derivatives of Ubuntu. These derivatives are wasting developers limited resources by essentially doing the same thing (i.e., making it easier for users to use Ubuntu). Canonical-controlled Ubuntu makes it hard for people to get the features they want in Ubuntu; a community-controlled Ubuntu will solve that issue.
There is a reason why you don’t see any successful forks of Fedora, openSUSE or Debian (except for Ubuntu), because these community-based projects empower the community to serve everyone.
Be a consumer: Jim Zemlin, the executive director of the Linux Foundation once told me that one of the many reasons organizations use open source projects is that they have “discovered that they want to shed what is essentially commodity R&D and software development that isn’t core to their customers.”
Canonical had been doing the opposite. Instead of using existing projects, even if that meant some compromises, they tended to create their own in-house projects. Not only does this approach go against the idea of open source, it wastes resources. My advice to Canonical is to start consuming as much as you can, and when you do make changes, do it upstream.
Snip Snap: In last week’s announcement, Canonical gave hints that it would go back to Gnome and Wayland. But it still seemed attached to Snap. Canonical should ditch Snap and adopt Flatpak for desktop and Docker containers for IoT and cloud. I understand Canonical might have its own reasons behind developing Snap packages, just the way it had reasons behind Unity, Mir and Upstart. Don’t make the same mistake again, Canonical. Drop Snap, even if it means some compromises.
That leads us to the last point:
Give respect to earn it: You can’t expect a community to include features that you need if you are not seen as a good citizen. Canonical and Shuttleworth have a long history of upsetting members of the open source community: Unity was created because Canonical couldn’t work with Gnome; Banshee was kicked out of Ubuntu because Canonical wanted to take a cut of sales; there was a dispute between Canonical and Linux Mint over licence fees; the founder of Kubuntu was kicked out of his own project; there was a nasty battle within the Debian community over the use of systemd vs Upstart; Mir was announced with heavy criticism from the Wayland and Xorg communities. You can see there is a very long list of confrontations between Canonical and other open source communities.
You won’t find any such conflicts in the Fedora or openSUSE communities. Tribalism and cutting others short to make yourself look big will never work in the open source world. You can’t insult people who work in their own free time (and often company time) to create and improve upon the project that you use in your products.
I think it’s about time for Shuttleworth to start the healing process. The new message from Canonical should be community first. In my meetings with the CEOs of Red Hat and SUSE, they both made it clear that for them community comes first. As Red Hat CEO Jim Whitehurst said, “If it’s important to the Linux community, it’s important to Red Hat”. That should be the new message from Shuttleworth to the community: “If it’s important for the open source community, it’s important for Canonical.’
The 5 Cs
All of this boils down to 5 Cs: communicate, compromise, consume, collaborate, contribute. Looking at last week’s announcement, Canonical seems to be heading in the right direction. They are tightening their focus on enterprise, cloud and IoT because that’s where the future is. It’s too early to say how successful Canonical will be as it transforms into an enterprise company. That also depends on whether it truly wants to transform itself into an enterprise company or continue to chase the desktop mirage.
Take the red pill, Mark!