Good Luck, Ubuntu

I have been a huge fan, supporter, and promoter of Ubuntu since about 2006. All the things that attracted me initially were: gratis, free/libre, community, just one CD to install. As I used it more I liked the technical merits as well, things such as timely releases, hardware support, server version, application availability, and LTS. I used it at home and work whenever I could. But things were not always perfect. Things popped up here and there that made Canonical’s direction for Ubuntu more important than the community’s. Things such as CLA, proprietary Landscape, Launchpad, bzr, Unity (initial releases; I love it now), Upstart, Amazon-in-Dash, etc. There was a visible Not Invented Here (NIH) syndrome in Ubuntu, which was good to some extent but then became overwhelmingly powerful. In an effort to control all the things they cared about, Canonical started deviating from the wider Linux ecosystem.

Upstart was a great idea and executed well enough that it was included in other distributions. But then Red Hat’s NIH kicked in and they created an alternative, systemd, and convinced other distributions to adopt it. Debian and Ubuntu are the standouts in not having embraced systemd yet. I have used systemd and as a user I loved it. It makes it easier to manage my services without getting in my way. I don’t know how I would like it in a sysadmin role but I don’t think it’ll be awful. So now if the wider Linux community accepts systemd then why ignore it?

Traditions are important but innovation and change are even more important. Canonical is driving that change in Ubuntu, especially with Mir. But it’s such as ill-advised change. When Wayland has been the community darling for a while, and Ubuntu kept saying they would replace X with it, why go the Mir route? Control; that’s what Canonical wants. It’ll allow them to compete with Google and Apple and make free software available to everyone in every form factor. Canonical is essentially changing its target audience from everyone (users, software developers, sysadmins, enterprises, etc.) to some (users, carriers, hardware manufacturers, enterprises, application developers, etc.). They don’t want Ubuntu developers anymore, they want developers building “apps” for Ubuntu. Which is fine if that’s their goal but they should realize that a lot of their original community members will leave as they don’t fit in anymore. But with Mir the real fear is that a lot of good work done recently with open source graphics drivers will be outdone by manufacturers targeting Mir to the detriment of the rest of the Linux world.

If all the secrecy around Ubuntu phone, Mir, QT-based Unity, etc. was because Canonical feared competition would take their ideas and build products faster than they could, they have missed the bazaar of free software where ideas are presented, discussed, and developed in collaboration for the benefit of all. So what if someone else takes your idea and builds a proprietary product? Our (your) goal is to create a free software product.

Unless you have the financial resources of Apple, Google, and Microsoft you can’t build an OS end-to-end on your own. You need a volunteer community to do that. Red Hat and SUSE understand that and try to build better community relations. Yes, they are not perfect and can be heavy handed when their goals are at stake but they haven’t discarded the community like Canonical has. If what Elizabeth Krumbach says has any weight in it (and I would agree it does), then the Ubuntu community will be less involved (if any) with code development as time goes by, and will be relegated to QA, documentation, support, etc. So the community will support whatever the Canonical cathedral code-dumps on the world without any input on the direction of the code.

Of course, Canonical and Shuttleworth have spent a lot of time, energy, and money to make Ubuntu what it is today. They do need to make money to continue Ubuntu. But at what cost? Is world dominance (with free software) a worthier goal than free software? Is free software meant to be dumped on users in “tada” moments? Free software performs two functions: gives rights to users as well as promotes collaboration. If Canonical doesn’t want collaboration on Ubuntu from a wider community then it’s their eventual loss. They’ll be hard pressed to continue their growth without a significant contribution from a diverse community. Red Hat is now a billion-dollar company but it also can’t create a complete OS and all its other offerings on its own. You need to be the size of Google and Apple to pull that off. Given this situation, can Canonical afford to disenfranchise its most ardent supporters?

Many like me will move away to other distributions. Debian is the obvious choice. But I would like people to give Fedora a chance as well. It’s not as easy as Ubuntu or stable as Debian but it strikes a good balance between innovation and breaking things. Once you get used to Fedora’s quirks and cadence you’ll find a truly free OS that makes community an essential part of its existence. Meanwhile, I wish Ubuntu good luck in Canonical’s endeavors. I hope by Ubuntu 14.04 we’ll see Canonical reach its goals and prove us all wrong.


Do I care about the end product or the process of creating the product? If you’re a free software enthusiast you’ll most likely want to be involved in the process, from beginning to end. Yes, the end product is important but so is the process. And you can’t build a product if you don’t have enthusiasts willing to step in and do the hard work. Well, you can, but you’ll need huge financial resources to do so. Does Canonical have the money and willingness to take Ubuntu forward on its own?

One Response to Good Luck, Ubuntu

  1. Thanks for the interesting post and the kind words about Fedora.

%d bloggers like this: