michaeldehaan.net

Hmm, Fedora uberthread. Thanks Fedora Planet for the heads up. So, what is Fedora? Hmm, good question.

I think it is really good as a developer distribution due to compatibility and upstreamability with some really good bits (RHEL and/or CentOS) but is losing (I’d rather say not capitalizing on) marketshare on the desktop/normal-computer-user front because it hasn’t defined itself. It is not so much “of and for itself” as people have a lot of different directions they want to take it. There is a lot of value in enabling someone to suprise you, and by no means should we ever squash that. Many will say we should not be concerned with “loss” and that numbers are not measurable, etc. All fine, I agree.

While it is exclusionary to suggest it become say, an end user distribution, I see no reason why Fedora, with the properly applied levels of focus, could not be better than OS X in nearly every user capacity and become an unstoppable wave — we know open code (with solid QA) can be superior due to infinite potential team size and moves faster when those models can be applied — WHEN those models can be applied. For Linux this is the right next step after winning the server space if World Domination is the goal. It would be a noble effort. Though it almost requires what everyone in Fedora wouldn’t really accept — assigned tasks from project managers and a lot of organization. Some would help, but not everyone, as they are all here for different reasons. That “this is your work” model doesn’t work in OSS space so much, which is why I’d venture you see Ubuntu doing it in semi-closed fashion. I think the problem is largely one of time. Most solid OSS contributions I’ve seen come out of either (A) someone’s at work time (server stuff mostly) or (B) weekend time. It would eat us alive to ask everyone to spend non-work time on things that don’t scratch their own personal itches. It wouldn’t work.

Remember Red Hat began as a distribution where it was aggregating community bits that made them usable — it required that level of direction, I’d say. I’m not sure what Canonical is doing is wrong.

We just find in this the question of this — what is Fedora’s identity if we cannot direct all of our tribe as one? It’s a place for people who need other packages to get packages to develop things. It’s a forge pointing towards CentOS and RHEL. It enables people to suprise us and bring that into other distributions that people also use.

Thus I’d probably say this — define Fedora as that — the Forge. Concentrate on the embedded space (where Fedora is widely used already), the Server space (Red Hat dominates here), Fedora Hosted, and EPEL (perhaps one of the coolest things ever that never gets enough press). Make your development tools and processes really awesome, powerful, and flexible. Go with that. Keep GNOME working and all that — for the people that work on Fedora.

I think that’s more interesting than playing the “me too” game on the desktop. Now what that actually means, I’ll leave that as an exercise for the reader. But for the last several years, that’s exactly what I saw Fedora as — it was a great place to develop software that needed consumers and other software to work off of it with each other, and it was an excellent bridge to other derivative distributions. Though I’d say the community I feel the most apart of is the user community and the upstream community of individual products, Fedora is the hub that makes those projects possible.

Meanwhile, someone really really should try to do a fully open OS X killer. That would be beautiful. Though I’m not sure that’s what we should be doing with Fedora, due to (as Spot said) many of us having reasons of contribution that fall in different directions. That’s cool.

Keep being a framework for us to surprise ourselves. Concentrate our efforts on making that happen in cooler, better, and faster ways. Work on our own infrastructure and make it kick even more butt than it already does. Lower barriers to entry into the development project. Reach out to new developer audiences that have traditionally not contributed. Stuff like that.