First, thanks to Greg for an excellent, thoughtful post on Fedora’s goals. I remember well — and I’m sure Greg does too — the FUDCon in Raleigh in January 2008 where members of the Fedora community sat down to try to distill “what Fedora stands for” into a powerful message. The result was the freedom, friends*, features, first mantra — guiding values that we’ve enshrined on our Foundations page.
The ability of any Fedora contributor to scratch his or her own itch is one of the strengths of our community. It has yielded many exceptional contributors, brought volunteer leadership to many of our project teams, and produced a steady enough stream of young contributors to merit an annual scholarship recognizing their achievements. One of the jobs of every Fedora project leader is to balance contributors’ ability to scratch their itches effectively, even when they’re not in concert with everyone else. That facet of the Fedora Project is a deliverable for our objective of building FOSS communities, and we need to continue to provide it continually and continuously.
At the same time, there’s a distinct difference between a project that has this capability of traveling forward in many directions at once, and a product with such a capability. And another one of our objectives is to provide a top-tier Linux-based general computing platform. We provide a myriad of free software packages that allow people to define a combination they want to personally use. Since we provide 100% remixability, anyone can produce a personally satisfying combination from that universe of installable stuff. That’s a more or less autonomic feature of Fedora these days, of which the many official spins produced each release are evidence.
The Fedora Project can easily encompass, for example, a dozen hosted code projects that each manage virtual infrastructure with a slight twist, or SIGs that want to provide completely different desktop environments. That flexibility isn’t as easy or desirable with a single product. Finite resources mean choices, because one product can’t do all things equally. (And in complete fairness, even a big-tent project like Fedora can’t do completely opposite things equally well — such as promote rapid advancement at the same time as long-term stability.)
That’s why, for instance, we don’t offer seven email clients directly in the default Fedora spin, nor three different mail servers to process their mail, nor three desktop environments in which to run them.** Community test, QA, and documentation efforts five times their current size couldn’t effectively test and note all those combinations of software! And the contributors doing those jobs in Fedora are just as important to our community as any others. So a shared set of priorities are important. That’s also reflected in aspects such as our packaging guidelines, which on the surface seem to restrict autonomy. But in reality they’re an expression of our community’s shared priority for scalability. These guidelines produce better software security and interoperability, as well as encourage forward momentum in all the upstreams whose software we distribute.
That point — scalability — brings me to another complete agreement I have with what Greg wrote about sacrificing bedrock principles of free software methodology in pursuit of a worthy goal; the ends do not justify the means. Fedora provides and champions an honest, sustainable, scalable system of upstream collaboration. And one of the best ways for us to demonstrate how well that system works is by putting an exceptional distillation of it into users’ hands. That experience is not just about what appears on screen when they boot a Live CD or USB image, but what happens afterward as they use it daily — and how we can use that experience to on-ramp curious, willing people into participating in sustainable, meaningful free software practices. There is still betterment we can apply to the way that users (which includes, not excludes, our contributors) experience Fedora, and I’m interested in continuing to explore them, as we’ve done for many releases.
So in short, I’m interested in focusing on a set of priorities in the Fedora distro, and empowering a wide set of alternative capabilities in the Fedora Project, following the core values of freedom, friends, features, and first. Which I think is just what the doctor (DeKoenigsberg) ordered.
* “Friends,” as opposed to “folks”… less folksy, but more accurate.
** Of course, the user, once at the helm, has complete freedom to do any of these things!