Open ... And Shut The more the Linux Foundation broadens its mandate beyond its core mission of "fostering the growth of Linux", the more it risks stepping on the toes of its most ardent supporters.
This tension was on full display earlier this week when The Register reported an apparent conflict between the Linux Foundation's support for OpenMAMA, an API used in financial messaging systems, and AMQP, a financial messaging system wire protocol.
While some, like Chris Swan, were quick to point out that the two can actually be complementary, it seems certain that the Linux Foundation's support for particular open-source projects will create conflicts with its sponsors.
But need it?
Since its creation from the merger of the Free Standards Group and Open Source Development Labs in early 2007, the Linux Foundation's mandate has shifted, if subtly. Whereas today the Linux Foundation declares that it "foster[s] the growth of Linux" with an emphasis on "promoting Linux and providing neutral collaboration and education", back in 2008 this part of its mission came last, not first, and was much more constrained:
- Protecting Linux by sponsoring key Linux developers and providing legal services
- Standardising Linux and improving it as a platform for software development
- Providing a neutral forum for collaboration and promotion
The first two areas were clearly focused on the Linux Legal Defence Fund, employing Linus Torvalds and supporting the Linux Standards Base, among other things. But it's the third area that has expanded over time, and is now the the organisation's primary concern. Originally intended to create conferences, it has become a way for the Linux Foundation to support open source more generally, and not always directly related to Linux.
OpenMAMA is a middleware-agnostic messaging API. It's not Linux. It's not part of Linux. In fact, though it currently supports Linux on x86 systems, a near-term goal is to also support Windows and Solaris. This isn't about Linux.
So why is it being hosted by the Linux Foundation? According to the OpenMAMA website, "the Linux Foundation provides a neutral home for the industry and the open-source community to collaborate on OpenMAMA". Which is great. But it still doesn't explain why the Linux Foundation should host it instead of, say, The Apache Software Foundation or some other neutral, non-Linux-centric body.
In part this is answered by the Linux Foundation's executive director, Jim Zemlin, who in an email to me suggested that one of the questions the foundation asks itself before taking on a project is: "Will this activity increase the adoption of Linux or open source?" (emphasis mine).
Over the years, the Linux Foundation has proven to be an exceptional steward for Linux development, and an active, informed voice for open-source development, generally. It is therefore not surprising that the New York Stock Exchange, original developer of OpenMAMA, would turn to the foundation to host the project, particularly given NYSE's involvement with the organisation over the years. Nor is it surprising that the foundation would step in to fill the void on open-source leadership of projects that have a relatively close connection to the operating system and its adoption.
The financial services industry is a huge adopter of Linux. Helping solve some of its middleware woes seems like a good way to broaden the kernel's adoption, too.
However, this doesn't obviate the fact that even within Linux development, the foundation's support for particular projects can rankle its sponsors at times. Consider MeeGo (now Tizen). While I was serving as COO at Canonical, the primary sponsor of the Ubuntu Linux operating system, I remember the frustration we felt when the foundation announced that it would assume leadership for MeeGo, a competitive OS for mobile devices. To me, that felt like the Linux Foundation was picking a winner.
It still does, though MeeGo-turned-Tizen has been anything but a winner.
My friend and Linux Foundation executive director, Jim Zemlin, emailed me to give more colour on how the Linux Foundation thinks about things:
We do not view the world of Linux as a zero-sum game within the different market segments where Linux is used. That has been a fundamental tenant of open source. The investment that happens in projects such as Tizen or MeeGo benefit Linux in general as a substantial amount of the code created there ends up in the upstream projects that are used to create a host of Linux based products and services. In fact some of these projects have an explicit "upstream first" policy in order to encourage sharing.
The reality is that we no longer operate in a winner takes all world. After all, that is the whole point of open source. And that is why the Linux Foundation is able to balance the interests of the hundreds of companies and thousands of individuals who participate in our many programs.
As a rule, the Linux Foundation does a great job of this. Under Zemlin's reasoning, the foundation isn't so much picking winners as it is trying to ensure Linux wins by making sure there is a centralised neutral place to focus mobile, server or desktop Linux, including adjacent technologies that directly impact Linux (like OpenMAMA). And since it's impractical for the foundation to do generic R&D, it necessarily focuses on a particular project but invites all to share in the development load and results.
Does the Linux Foundation step on the toes of its sponsors at times? Absolutely, whether intentionally or not. But in the non-zero sum world that Zemlin describes, this is both to be expected and probably appreciated. The Linux Foundation, after all, is hell-bent on growing the Linux and open source industries, and not any particular vendor therein. I'd hazard a guess that for every time the foundation upsets a sponsor, there are nine more times that praise it.
That's just the nature of the open source beast, a fact that Red Hat, Canonical, and other supposed Linux Foundation competitors realise and embrace. ®
Matt Asay is senior vice president of business development at Strobe, a startup that offers an open source framework for building mobile apps. He was formerly chief operating officer of Ubuntu commercial operation Canonical. With more than a decade spent in open source, Asay served as Alfresco's general manager for the Americas and vice president of business development, and he helped put Novell on its open source track. Asay is an emeritus board member of the Open Source Initiative (OSI). His column, Open...and Shut, appears twice a week on The Register.