Linux, musical road-dogging, and daily life by Paul W. Frields
FAB reprint.

FAB reprint.

I was asked to reprint here on my blog a posting I made to FAB last week regarding the status of the target audience discussions and their place in the wider scope of improving the Fedora distribution. The Fedora Board’s discussions about target audience are meant not as an end but a means to an end. By setting down in black and white some of the answers to questions frequently asked by contributors, we can begin to share an understanding about where we are going in the future, and how we need to adapt the processes by which we make the Fedora distribution to reach those goals.

I’ve seen this process cast as “figuring out what we’re doing” or “what we are” but those characterizations are somewhat missing the mark. The real question is “Where do we want to be in the future?” and “How can we get there?” The Board’s mission is to lead in those areas and contribute to the Fedora Project and distribution being the best they can be.

I’ve made a couple trivial edits where I found a missing word, and included hyperlinks instead of footnotes for easier reading:

In our Thursday meeting, the Board talked at length about the lengthy discussions that have been happening on this list, which have been both spirited and, as always, very helpful. Continuing to make the best possible Fedora distribution is a top priority to everyone who works on it. We all want to see the Fedora Project succeed as the leader in advancing free and open source software, and the Fedora distribution is how we put our best work in front of a wide audience twice a year (and at all times in between!). And of course we want the Fedora Project to continue to be a vibrant community where contributors pursue a variety of goals, sharing our core values of Friends, Freedom, Features, and First.

The specific discussion about making the best possible distro has focused in on target audience. This is a sensible (and arguably overdue) step, because to provide a Fedora that satisfies someone, we first have to know who that person is, and differentiate them from the mass of “everyone,” and also we need to be clear in what we expect them to be able to do. We found four defining characteristics that we believe best describe the Fedora distribution’s target audience: Someone who (1) is voluntarily switching to Linux, (2) is familiar with computers, but is not necessarily a hacker or developer, (3) is likely to collaborate in some fashion when something’s wrong with Fedora, and (4) wants to use Fedora for general productivity, either using desktop applications or a Web browser.

This target audience does not [represent] a major shift away from what most people in the Fedora community believe. Having a target audience also does not preclude any feature development that goes beyond that audience.

By having an audience in mind, we as a community can prioritize resources, and at the same time make it possible for people who want to concentrate on other audiences to build community around those efforts. Fedora teams already are making progress on this, and one example is our Fedora QA team — which opens a schedule of  community test days for every release, and provides information on hosting them. Thanks to our remixing tools, anyone can put together test day spins to facilitate the testing. Anyone who is interested in technical goals, whether they are part of the target audience focus or not, has a place and resources in the Fedora Project to help achieve them.

The Board members and I believe that making the experiences of getting, using, and contributing to Fedora better for the target audience will also improve them for our close community as well. This is not an either-or proposition, but a win-win. In essence, the target audience is much larger than the group of people in Fedora who are, say, subscribed to this list; or who develop features, collateral, and other content for Fedora; or who do great Fedora advocacy work, whether through speaking, writing, or in any number of other ways. When we improve Fedora from the perspective of that superset of people, we are also very likely improving it for our core contributor community as well.

Those people may end up telling others about their experience, and thereby expand our actual user base even beyond its substantial current size. That’s a very fortunate consequence of making a better product, but our goal is not to simply target “everyone,” which isn’t a reasonable goal given finite resources. We wouldn’t be unhappy if more people started using Fedora casually, even outside our audience, but at the same time we want to continue to build something designed for the people we think we can reasonably please.

We don’t delude ourselves that the target audience definition is now “done.” What we have now is simply a shared understanding of where to start, and we can start adding definitions of tasks and expectations, to understand an example or profile of our target audience — what this person wants, understands, needs, is like. That work is going to require input from people who know more about user design than any single person in one meeting. I’m hoping people here can constructively help us draft this profile on the wiki, or use other collaboration tools to create a better shared understanding of this profile. Mairin Duffy recently posted her take on starting this work and I’d like to see even more definition added to the profile.

Of course, this doesn’t magically happen overnight. In fact, it can’t happen at all without coming together as a community to address the nuts and bolts of actually fixing things that are broken. Part of the miracle of the Fedora community is that we aren’t afraid to admit a failure, understand it, and fix it and move on. Hoping to contribute to solutions, the Board discussed some of the brokenness, issues we often hear from people who do fall within our target audience, including people who are in our large community of contributors. Among those were frequency and reliability of updates and upgrades. If we want to attain the goal of making our audience happy, as a community we need to do a better job of not breaking their systems or causing them to doubt the quality of the software they’re receiving.

This is a topic that will bear further discussion, obviously. Together we need to figure out the best ways to balance our desire for advancing free and open source software (i.e. the Fedora Project mission) with the provision and promotion of a platform that our target audience can confidently use. I’ve referred to this in the past as “update discipline” but not in a flippant manner. I don’t mean “discipline” in the sense of reward/punishment or anything like that — rather, in the sense of a community dedication to doing things well, consistently. At least one set of ideas has been written up already to brainstorm on the problem, and while I think there’s still work to do to figure out a solution, I think there’s already quite some consensus that the problem exists, is important, and is worth trying to solve.

Finally, the Board talked about the proposal for “unfrozen Rawhide,” which Jesse Keating offered at a Fedora event this summer. Just like many of our contributors, we’ve felt the pain of having an uninstallable Rawhide, which negatively affects everyone’s ability to more efficiently deliver new code and features. In essence, Rawhide has been too often “eating babies” indiscriminately, and we need to improve its contribution to our develoment ecosystem. The Board feels that Jesse’s proposal not only has the potential to help us achieve a more installable Rawhide, but if it’s managed correctly we could have a Rawhide that more of our core contributors could actually use during and prior to test phases — while not undoing our ability to allow and encourage innovation and new ideas.

So in summary the three points that came out of Thursday’s Board meeting — target audience, better update discipline, and a more useful Rawhide — are all topics that we intend to discuss further here, and at the FUDCon in Toronto. We’ll have a majority of both the Board and FESCo representatives together then, along with a wide selection of our community, to help put together a roadmap for the next evolution of Fedora. I’m very excited about that event, and can’t wait to take advantage of the opportunity to help move Fedora forward. And although it goes without saying, as always I’m very grateful to be working with such an impassioned and dedicated community of contributors.


    1. @Gianluca: That’s hilarious! I don’t read LH but hey, parody++. I always try to keep in mind that not upsetting anyone means you’re probably not doing anything. 🙂

  1. “update discipline”
    I think there’s already quite some consensus that the problem
    exists, is important, and is worth trying to solve.

    Once again, I have to heavily disagree. And as I’m not the only one, I don’t think speaking of a “consensus” makes sense. In fact I suspect there’s a vocal minority complaining about the updates and a silent majority that is perfectly happy with them.

    From the proposal you linked:

    Batched updates must occur and appear to the user no more often than once per week

    That is a catastrophe. Whenever the current irregular update pushes, for whatever reason, have a whole week between them, the result is a huge push with dozens of packages. This takes very long to download and install and is much more likely to disrupt your work than a small push. It’s much easier to deal with one small push a day than a huge one every week. More frequent, smaller pushes also make it easier to identify the culprit if something is wrong.

    Plus, I really don’t see what this would achieve. Those people who really want to update only once a week can already choose to do so. Why should we force this model on everyone?

    We should shoot for reliably daily updates, not weekly ones!

    System updates should only:
    – fix critical bugs or security vulnerabilities
    – provide hardware enablement
    Application updates may add new features even in a stable release at the upstream brand’s discretion as long as someone can be held responsible for fixing problems

    There are several things I don’t like about this:
    * What defines what is a “system update” and what is an “application update”?
    * The policy for “system updates” is throwing out Fedora’s main point of differentiation compared to other distributions, so it sounds like “throwing the baby out with the bathwater”. The kernel might slip through under “provide hardware enablement”, but what other packages which were previously kept up to date would be affected by that policy?
    * Why should it be upstream’s decision whether we ship feature upgrades of an application to a stable release? Most upstreams probably don’t even have an opinion there. This is a distribution decision and should be handled consistently across the distribution.

    I think we need a stronger policy to strongly encourage (“SHOULD” type policy) version upgrades where they don’t break things (incompatible changes, feature regressions etc., I’ll put up a more detailed proposal to link to ASAP), while still discouraging those that do break things (but IMHO this part is already working well, the problem is with packages not getting updated for no good reason).

  2. @Kevin: You may have misread my post and inferred that the brainstorm to which I linked was already favored. But I don’t think a service-pack approach is the answer for a community-driven release. I do think that page makes some valid points about how the very people we try to attract are negatively affected by poor update processes. Broken dependencies, changing APIs, and inconsistent approaches to updates all make for a negative experience by all Fedora users.

    To some degree these are also affected by the fact that most productive contributors can’t live on Rawhide consistently because of the way it’s managed. Since for 1-2 months out of the cycle Rawhide isn’t usable by mere mortals (or even some superhumans), the only way to get brand-new updates is to push them back into old releases, breaking compatibility and/or user experience. That’s why I feel that any new approach needs to be coupled with improvements in how we manage Rawhide, so that (1) maintainers who want to constantly push out the latest updates regardless of compatibility can do so in a branch that people can actually use, and (2) we still retain a place where things are allowed to be broken more routinely if necessary.

    There is value in Fedora being able to provide updates quickly during a release’s lifetime, but not in the inconsistent way we do it now. It could very well be that only a subset of packages needs to have a more strict approach, and that edge packages — the kind used only by a small audience — have more leeway. You’re welcome to write up and make a proposal of course.

  3. Allen Halsey

    For target user, please include the children, friends, parents, and significant other’s of the linux enthusiast. I think it is common for for a Fedora Linux enthusiast to by the sysadmin for others.

    I helped my wife create a poster for an event hosted by her employer using Inkscape on my Fedora box. Now she logs into her own account whenever she needs to. My 5 and 7 year old have their own accounts and enjoy logging in and seeing the desktop background they’ve chosen. They play Tux Racer, Alex the Alligator, and web-based games. I’ll start teaching them Inkscape soon. My mom and dad just need web and email. I plan on setting them up with Fedora soon. Remote admin is one of Fedora’s strong points.

    Please include these non-RHCT level users in the target user profile.

    1. @Allen: Thanks for your thoughtful comment. My family, too, uses Fedora, including my wife, my daughter (8), and my son (almost 6). Keep in mind that by concentrating on one or more target users, we are not looking to make decisions that don’t help other people. On the contrary, the things that we can do to make Fedora work better for our target audience will often make for a smoother experience by the people around them as well, such as friends and family who use Fedora. By keeping a focus on our end goals, we can produce a better distribution that will inevitably make a lot more people happy!

Comments are closed.