A re-evaluation of Canonical
Off the Beat: Bruce Byfield's Blog
Like many people, I've always thought of Canonical as a FOSS company. Recently, though, I realize that I've been guilty of a category error. Unlike Red Hat, Canonical does not have an open source business model. Rather, like Novell or Google, it's a company with a mixed business model in which free software and proprietary practices mix as convenient.
Nothing's wrong with a mixed business model, no matter how disappointed it makes me. As much as I'm tempted, I can't even bring myself to call it immoral, although I would say it's not part of best practices.
But I do wonder about the mistake in my thinking. I spend hours every day tracking what's happening in the free and open source software community. So how could I miss something so simple and obvious?
In self-defense, I can only say that I'm not alone in my assumption. In fact, thousands make the same mistake. But why?
A failure of expectations
Probably the first reason for the mistake is that Canonical contributes what must be a good proportion of its resources to developing Ubuntu, which genuinely is a FOSS product. Many Ubuntu leaders are Canonical employees, and Mark Shuttleworth, the founder of both Canonical and Ubuntu, has the last word about Ubuntu's direction as much as he does about Canonical's.
Under these circumstances, it's not always clear in what capacity Shuttleworth (or any other Ubuntu contributors at Canonical) are speaking and acting. Are they speaking in as Ubuntu contributors, or Canonical employees?
In a cynical mood, I might wonder if the distinction isn't sometimes deliberately blurred. There's no denying that Canonical benefits hugely from being perceived as a FOSS-based company -- much more so than it suffers when members of the greater community complain about its apparently inconsistent behavior. The obvious comparison is between Canonical and Ubuntu and Google and Android, although Ubuntu remains far freer than Android.
Certainly, there's no denying that Canonical has often gone to some lengths to cultivate the FOSS community. For instance, if you look at the early entries in Mark Shuttleworth's blog, you'll find Shuttleworth talking as though he were a developer instead of a company CEO -- just another member of the community You'll find him expressing his opinion on one FOSS issue after another, as though he nothing else to think about.
I don't question for a moment that his interests are genuine, but his emphasis on them, rather than any aspect of business do suggest an effort to demonstrate that Canonical is good citizen of the community.
Nor, given Canonical's contributions to the desktop, was there ever much questioning of the image being projected. Dissenting voices exist, but, for the most part the community was content to take Canonical at its own self-proclaimed worth.
The community wanted to believe in the eccentric multi-millionaire who was apparently altruistically accelerating FOSS development. Most of its members ignored any early signs to the contrary, such as the slowness with which the Launchpad code was released.
So far as I know, Canonical never declared itself a FOSS company in so many words, but it was more than willing to be perceived as one. Quite possibly, in the early years of the company, the intent really was to be one.
However, in the last couple of years, the signs that proprietary practices were also part of Canonical have become impossible to ignore. Its websites -- at least in the top-level pages -- began to talk about Ubuntu as an operating system in its own right, and to eliminated any mention of GNU/Linux. Proprietary services, such as Ubuntu One, and a Software Center with non-free items, started to become standards parts of Ubuntu.
For complex reasons, Canonical began to prefer in-house projects it could dominate (like Unity), rather than working with other existing FOSS projects (like GNOME). Canonical began to be interested in contributors' agreements that allowed it to relicense code contributions however it wanted. Instead of trying to improve relations with other projects, it began to look for contributors for its projects.
Such changes may reflect the growing urgency to make Canonical profitable. But at each one, the FOSS community -- and I am no exception -- has screamed betrayal. However, perhaps what we should have been doing is re-evaluating our original perceptions.
The end of the honeymoon
Looking back, I realize that, in looking askance at Canonical, what I and almost everyone else have been doing is reacting to our own expectations as much as what anyone at Canonical has said or done. When we have condemned Canonical hypocritical, we have been voicing frustrations over our unfulfilled expectations as much as objecting to inconsistent behavior.
To state the obvious, Canonical does not exist to meet our expectations. Like any other company, it exists to make a profit.
Yes, Canonical falls short of our hopes and ideals. Perhaps, though, we should stop waiting for Canonical and Ubuntu to lead us into the promised land of universal acceptance, and regard it as just another company using FOSS for its own purposes. At times, it will be an ally of the community. At other times, it will do what's best for it alone.
Such an attitude may not be as much fun as unchecked outrage. But it would definitely be much fairer.
Re: Canonical moralityUnfortunately, nothing you say changes the fact that Canonical is not a pure FOSS company.
Canonical moralityMark Shuttleworth explained in a video, rather apologetically, that it was not suitable to withdraw his financial support for Ubuntu at that time. His endeavor was to make Ubuntu self-supporting. This is high virtue.
As an Ubuntu One member, for example, i have more storage and services at no cost, than i shall ever use, let alone breaking into the paid categories.
Let us have some heart, some love. What Ubuntu is doing is brilliantly conceived, and strategic for universal benefit. It seems a stylistic thing to be hardcore hard-hitting cynics, to belong to the day's religion, Science, owned by Money.
Linux surpasses that in visionary fashion, giving, giving, wealth all the way. This is love. Money can never do the work of love. It is more a weapon than a tool. Humanity has a long way to evolve. And Mark Shuttleworth sees where we have to go. He exemplifies love and happiness.
I agree with your observation of Canonical. I discovered Linux in 2007 and at this time canonical sent CDs ubutu 7.04. I remember putting Inquiring on the motivation of a entreprise to send free CD around the world a CD operating system based of Ubuntu. But that with Ubuntu I'am discover the world of Linux. But this is not the system I have adopted. Over the years I have tested a large number of distribution stopped my choice on Pardus, Vectorlinux on desktop and laptop PC-BSD with Triskel 4.5. These are the main system I use. I continue to test other systems for fun.
I always had doubts about Canonical and the time confirmed my doubt. I also believe that the open source community is a bit naive with respect to the GNU. The BSD license is even more naive licence for me. Apple used the FreeBSD kernel for built his Operating System without giving anything back to the BSD community. A large number of business uses of the source code for their own purposes and benefits of good faith free of the OpenSource community naive they use for their benefit. Is not it true that Ubuntu has never contributing one line of code to the Linux kernel!
The open source community also has a cultural problem, free is not an absolute and to pay $ 5 or $ 10 for a Linux system is not a shame. The open source is not a synonym to free. I subscribe to the LinuxPromagazine, Admin magazine because is a fair price for both exellent magazine. I also contribute to the operating system I use. The community will have to learn contributed financially to their preferred Distro. The open source also needs to be funded like any organization it is a non-profit or not.
Sorry if my english is average
FOSS and ProfitsI agree with you in that we should re-evaluate our perception of Canonical and accept its nature. My frustration, however, is less with Canonical itself (or with other similar companies) as with the implications that its behavior has for FOSS and its relationship to for-profit corporations. How compatible are the interests of the free software community with the pursuit of profit as the primary goal? Is it possible to have a dense field of high profile for-profit firms that actively and genuinely advance the values of free software even if these enter in contradiction, at times, with the maximization of profit? Or do we have to accept that only a few big corporations whose main sources of income are free software products can find niches that allow them to be profitable enough so that they do not have to choose between profitability and free software? I do not have answers to these questions, but I think that your post highlights a recurrent problem in the community: we devote a lot of attention to what different actors do as if their behavior depended exclusively on their preferences and values. But free software is part of a wider environment that presents opportunities and limits to what the free software community can do, and these constraints are of different types and strengths for different actors. Are there structural limits to the growth and direction of free software? If so, how should we adapt our behavior in order to continue advancing freedom? Or should we change our goals? I believe that we should spend more time reflecting on how FOSS fits in the wider economic and social environment so that we can avoid wars that we cannot win or expectations that we cannot meet (and this is a general point, I am not at all suggesting that Canonical could not act differently).
A major setback for the Linux desktop.
Improved support for GPU in virtualization.
News site for the openSUSE community falls victim to a Wordpress exploit.
The source code is available online.
One out of three virtual machines on Microsoft Azure Cloud run Linux.
The form factor of the board makes it a drop-in replacement for Raspberry Pi.
Makes it easier for customers to move workloads into container-centric applications.
SUSE’s answer to container-centric operating systems.
Linux 4.9 is the biggest release in terms of number of commits.
The latest version of the official RHEL clone is here.