Fedora Strategy 2028: February/March Planning Work and Roadmap 'til Flock

What’s this all about?

The Fedora Council’s primary responsibility is to identify collective Project goals and set things up so they happen. It’s been a long time since our last official big strategic push, and we’re now working on the next one.

This has taken longer than I’d hoped, but after work this summer, discussions on this site, and now a week of intense collaboration at last week’s Fedora Council Hackfest, we have a draft of the high-level part of the plan. From this, we’ll move step by step, to identify specific desired outcomes and design the projects, initiatives, and programs we’ll need to realize them.

Fedora’s basic purpose remains as explained in our Foundations, Mission, and Visionwe envision a world where everyone benefits from free and open source software built by inclusive, welcoming, and open-minded communities , and everything we do is meant to get us there.

Navigate by the North Star…

For the next five years, we’ve chosen a simple target: double the number contributors who are active every week. This serves as a “North Star” metric (or “Guiding Star”, if you prefer) — a basic gauge that what we’re doing is having the results we need. If we accomplish this, we’ll be closer to our vision — and have a stronger base for whatever the next step might be.

Our planning framework

How will we get there? We’re using a framework called a logic model. If this is new to you or you’d like a refresher, see Theory of Change: how we plan (and explain our plans!).

The Council presents our first draft!

Based on the work we’ve done so far, the Council has developed a list of high-level goals — the impact we’re looking for in order to reach the North Star target, which in turn moves towards the ideal world we envision.

At the hackfest, we focused on taking this list and building the Impact column — and we added a new-to-Fedora-process column: Objective.[1] Think of these Objectives[2] as “an important thing we want to achieve as part of this” and Impact as “why we believe this specific Objective will move us towards our North Star”.

By the end of the week, we all felt satisfied with this first draft of this part of the Fedora Strategy 2028 logic model:

I apologize for the tiny text — I couldn’t get the whole thing to format nicely here as a table. Expand the following sections to see the Objectives and Impact areas we worked on:

Fedora is for everyone...

Fedora is for everyone

Focus Area: Accessibility

Objective Impact
Fedora websites and docs use the current best-practices for a11y. Expanded access to information allows people we were not able to reach before to enter the Fedora world.
Fedora Linux Editions use the best-available open source a11y tech People who benefit from assistive technology can become Fedora Linux users.
Our project tooling follows best a11y practices. Users who benefit from assistive technology can become Fedora Project contributors.

Focus Area: Reaching the World

Objective Impact
Fedora Linux is available pre-installed on more systems from more vendors. Reduced barrier to entry lets more people try Fedora.
Fedora Linux is widely available in cloud providers and CI services. Fedora is perceived as the default developer platform.
Fedora maintains a strong network of thriving local communities around the world. As a distributed project, we are more diverse and resilient.
Fedora leads in Linux distribution development...

Theme: Fedora leads in Linux distribution development

Focus Area: Community Sustainability

Objective Impact
Everyone in Fedora has a mentor, and everyone in Fedora is a mentor. Better onboarding. Growth for current contributors. Continuity of expertise and reduced lottery factor.
We have insight into community health and trends through meaningful metrics. We know where we’re succeeding — and what needs help. By measuring more than git commits, we can highlight more than coding.
Modernize our communications tooling. An easier, more friendly face for the project, and higher quality discussion for all, leading to increased involvement and engagement.

Focus Area: Technology Innovation & Leadership

Objective Impact
Fedora is a popular source for containers and Flatpaks. Fedora is a trusted source of software beyond our own base OS, reaching more users and potential contributors.
Immutable variants are the majority of Fedora Linux in use. New way of doing things brings excitement and new energy.
We integrate programming language stack ecosystems. Developers choose Fedora Linux because we provide a straightforward programming environment that works the way they expect.
We build on the success of Fedora.next...

Theme: We build on the success of Fedora.next

Focus Area: Editions, Spins, and Interests

Objective Impact
Each Edition has a story for each release. Better fuel for marketing, leading to more interest and press.
It’s trivial to create and maintain a new Fedora Spin or Remix. Niche environments bring enthusiasm greater than their size — and allow experimentation which may power future growth.
More (active) SIGs, fewer images. We grow strong special-interest teams and networks into other communities — neuroscientists, roboticists, teachers…

Focus Area: Ecosystem Connections

Objective Impact
Better collaborative workflow with CentOS Stream and RHEL. More resources for us — and a bigger potential contributor base through EPEL and similar.
Get people working on Amazon Linux directly involved in Fedora as an upstream. More resources for us — and another source of resources and vested-interest paid contributors.
Collaborate on tooling, practices, and offerings with peer distros and upstream projects. Make Fedora less alien to people from other projects. Working collaboratively multiplies our efforts.

About this draft

Some of these goals are new efforts. Others are about expanding existing work, or revitalizing something that was more active in the past. And some are significant changes in emphasis.

Our chosen objectives are not necessarily perfect right now. That’s okay — we’ve got time to revise and adjust. But I think we have a good start, from which we can work on the rest: filling from the high-level columns on the right to the increasingly-pragmatic columns on the left. If we stay at the abstract level too long, we’ll never get to the real work.

As we do each step, we’ll continually validate: for each item in a column, are the items linked on the column to the left necessary and sufficient? And, are the items on linked on the right the outcomes we want and expect? Are there important things we want to do that don’t connect into something on the right side? If so, we might be missing something!

What if something is missing?

First, it’s important to note that while this plan will give us a focus for our efforts, it isn’t everything. There are (at least) two major categories of project effort that we’ve intentionally left out:

  • Regular, fundamental work. Release engineering, marketing, documentation, support, design, etc., etc., etc. Just because these don’t have a row on the plan doesn’t mean they aren’t important. In fact, to the contrary, we’ll need these things to execute on most of the planned initiatives.
  • Anything someone wants to do that isn’t really aligned. If this were a business, we’d probably want to shut down efforts that don’t fit the goal. But, since we’re volunteers, we all do what interests us. And if that interest doesn’t particularly fit this plan, that’s perfectly fine. As long as it’s within Fedora’s overall scope (and not harmful in some way!), go for it. The project overall will allocate resources in support of the big plan, but there’s room for exploring in other directions too.

One area the Council discussed in particular: Improving Onboarding. This is clearly important, and directly aligned with the North Star target. We’ve decided for now to consider this one of the fundamentals, something underneath lots of other parts, not an initiative on its own. But I’m open to finding some way of making this in particular more prominent.

There may be some really big things that are missing, and if so, I hope to discover them and correct in the next several weeks.

Next Steps

After the Council meeting, we’ll have time for public discussion and comment, and then the Council will agree on an final version.[3] In the next few weeks, I’ll post topics about each of the Objectives, and we’ll have a month of discussion and adjustment for each. If necessary, we will discuss and make adjustments to the draft during our regular Council meetings or in Council tickets. This should be complete by the end of April.

Next, we’ll find leaders for each initiative — an Executive Sponsor from the Council and a Lead (who does not need to be a council member). We’ll work on defining the Outcomes — the Key Results, and then fill in Actions and Resources — for the latter, both what’s available and gaps that need to be filled for success. This should be done by the end of June.

We’ll then organize this into an overall schedule — with eighteen big areas we’re looking at, we don’t expect everything to happen all at once! We expect that some of these will be Community Initiatives[4] in the formal sense, while others may simply be projects — or ongoing programs.

The entire Logic Mode and overall plan should be finalized in time to present at Flock at the beginning of August.

Fedora Strategy 2028: a topic index for our planning process serves as a central point to keep track of all of the related posts I’ll be making between now and then.


  1. If you’re following along from earlier posts, you might recognize many of these. As we worked, it became clear that what we had was good but not in the right form for Impact. ↩︎

  2. And, in somewhat related news: we’re renaming what were previously “Council Objectives” to “Fedora Community Initiatives” — more on that to come! ↩︎

  3. “Final” doesn’t mean “permanently fixed forever”, of course — over the next five years, we’ll adjust and correct. But final in the sense of “this is the current plan”. ↩︎

  4. formerly known as “Council Objectives” ↩︎

7 Likes

North star seems to be a bad metaphor for a world-wide community since it only works in the northern hemisphere. Where is the guidance for LATAM?

Regarding “Fedora is for everyone”: Using abbreviations like “ally” makes it harder for non-native speaker to understand texts. Is using/providing Basic English considered current best practice?

Also regarding “Fedora websites and docs use the current best-practices for ally”. With Fedora’s First foundation, it would be nicer to lead instead of follow accessibility best practices.

Regarding “Modernize our communications tooling” - the impact of “higher quality discussions” are depend more on the behavior/usage of the tooling and less on the tooling itself. To get better communications, I believe better training is needed but it will also reduce involvement. Wanting more and better discussions does not seem to be realistic.

Regarding “Fedora is a popular source for containers and Flatpaks” - as a leader, it should be the popular source.

Regarding “We integrate programming language stack ecosystems” - we do this already, don’t we? This seems to be missing a number of how many stacks we want to integrate.

Regarding “More (active) SIGs” - the Impact mention special-interest teams instead of groups.

The objective to make it easier to create and maintain new Spins and remixes to have niche environments and at the same time reduce the amount if images (“fewer images”) seem to contradict themselves.

Regardding “Better collaborative workflow with CentOS Stream and RHEL.” and “Collaborate on tooling, practices, and offerings with peer distros and upstream projects.”, these objectives are very fuzzy. The other objectives are mostly more specific.

Regarding “Double the number of Fedora contributors active every week” - doubling the number every week is a very ambitious goal. Maybe “Double the number of Fedora contributors active during a week” is more clear.

1 Like

In Matthew’s defense, I talked him out of a footnote that said “or Southern Cross for those of us in the Southern Hemisphere.” I get what you mean, though. I think as a metaphor, it’s well-understood enough, but we can try to find a more global term.

No. We are not accessibility experts and shouldn’t pretend to be. We should follow the best practices identified by those who are. We can lead in how well we follow them.

Right now, we shove language stacks into RPMs. This is about finding ways to integrate those without necessarily requiring 100 RPMs to be built. That doesn’t add value to the ecosystem.

I can see that. The goal is to make it easier to create new Spins and remixes where appropriate, but to reduce the images where they don’t make sense. For example, the Python Classroom Lab makes a lot of sense as a live image that can be used in classroom settings. It should be easier to make things like that. On the other hand, the Comp Neuro lab may be more useful as a curated software set that doesn’t require a specific installer image. We shouldn’t require (in practice, not in policy) teams to create images when that doesn’t make sense for their use case.

Your rewording is the correct interpretation. The “every week” applies to the contributor count, not the doubling. It’s probably better as “Double the number of weekly active contributors”, which phrases it like metrics used in other parts of the software industry (e.g. “monthly active users” of websites)

5 Likes

I’m open to different phrasing. I like the current phrasing because it puts the emphasis on “Fedora contributors” , while the rephrase (to my ear at least) emphasizes “weekly active” instead. I could go for something like:

Double the number of Fedora contributors
measured by weekly activity

But I’m kind of inclined to just leave it for now — I can live with the possibility of an ambiguous misreading. Plus, if we accidentally get doubling-growth week-over-week, I won’t complain. :classic_smiley:

2 Likes

It is probably a good idea to prefer simple language. However, I don’t think we would use that particular conception of it, which is really a constructed artificial language based around a very limited list of words and simple grammar.

@till — on the specific points: the discussion on each of them over the next few months should both provide expanded explanation for what each means and give opportunity to sharpen them.

I think the header “leads in linux distribution development” is confounding me bc I don’t read the items below it as being development of a distribution. (Distribution development to me means letting people make their own distributions like remixes / etc.) They seem to be community maintenance and technical leadership things. So I think maybe a less ambiguous heading like “Fedora is a Linux leader” or something like that…? Not sure.

The other thing is I was kinda disappointed that there isnt a high level thingy for containers and development. We have the containers (as in Containers · GitHub) community as a close peer / sibling… and the idea would be, generally, the OS platform is freed, but we want containers to be free as well… see A new conceptual model for Fedora – Fedora Community Blog for the idea there.

To me, “Linux distribution” covers what Fedora, Debian, Ubuntu, openSUSE, Arch, etc. create., while if we say just “Linux leader” here it tends to narrow things towards kernel type features. As much as I dislike their stats page, I basically mean “things that might be covered by https://distrowatch.com/”. I’m open to a different way of saying it, though. Maybe it’s the word “development” that needs a tweak? I think we originally had this as “Fedora leads in the distro space”…

I think we have aspects of this in two places: first, under the theme above, in the technology focus area, we have “Fedora is a popular source for containers and Flatpaks.” But if I understand correctly, you’re talking particularly about the tooling projects — that would fit in the “Ecosystem Connections” focus area at the end of the list… specifically, " Collaborate on tooling, practices, and offerings with peer distros and upstream projects."

But maybe you are suggesting something closer — bringing those things more officially under the Fedora umbrella? I’m open to that, but we’d need those projects to want that…

1 Like

Another possibly-missing Objective under “Innovation and Leadership in Technology” might be: updating our packaging workflow, from things like rpmautospec to bigger ideas like Source Git and Packit.

I, for one, would like to see us move to a packaging model where the spec file (or equivalent) is clear of all boilerplate — something like 3 lines for a simple package. But, this might be an initiative we do separately rather than something that fits this plan. (Making packaging easier and more friendly to upstreams could be a way to increase active participants. But I’m not sure there’s a straight line to the “North Star”.)

1 Like

Hey Matthew,

Re: wording confusion in distro leadership

“Fedora leads in the distro space” would not have confused me as much as the current title so I think I prefer that language. I think “development” probably is the key confounding word. It definitely makes sense as a category it’s just the words I’m hung up on, lol.

Re: development/containers:

OK yes, the ecosystem connections bit is really what I’m looking for, specifically with containers and developers. In the past we’ve had a focus on developers and their use case…

Yeh, it would be nice if the Containers projects wanted more of an explicit partnership there. I don’t think it makes sense to pull them under the Fedora umbrella because they are cross platform and distro agnostic (although, to be clear, there’s a clear preference / priority / advantage for using with Fedora & friends because that’s where we mostly all came from!) So I think we should state the implicit partnership we have going on there kind of more explicitly I think? And better - I don’t know the right word, cross-promote maybe isn’t really the right one but gives the sense - between the projects to better tell the higher level narrative that the silly F model I made outlines. Be friends out in public :slight_smile:

I think there could be a straight line to the north star in this - in that, for example, podman has growing leadership and influence in the container space, if Fedora Workstation is the premiere development workstation explicitly stated to work with podman where you get the best / latest build-in (which I think is the case, isn’t it?) and we actually promote that, could that bring in some Linux curious dev to choose Fedora over other options to work with Podman? (And IIRC on cross platform when you install Podman you’re getting it on Fedora CoreOS? So we’re getting Mac and Windows users of Fedora that way already :slight_smile: )

3 Likes

A post was split to a new topic: Fedora Strategy 2028: What is a contributor? What is a contribution?

I disagree with the “doesn’t add value” part. Granted, it’s a chore if the ecosystem is not broad enough yet. Instead of ignoring the added value of having everything in RPMs, automation should be the answer.

My take: “doesn’t add value” is too strong, but it does not add nearly as much value as it could given the effort, and we have trouble demonstrating the (real!) value that I agree does exist.

It may be that “better RPM automation” is the best answer for us[1]. We intentionally phrased this in a broad way. But I think that must include allowing us to step-back and explore other ways we could address the problem and possibly do even more (and how we can get there from what we have).

But also, let’s talk about this one in specific in more depth when we get to the details of that objective. :classic_smiley:


  1. I’d love to see a typical next-generation spec file be just ~3 lines of metadata, and any other lines be by definition exceptional… ↩︎

1 Like

@smooge asked a really good question that I think is so important that I’ve split it into its own topic:

1 Like

One thing I think is missing from the next 5 years goals is to reduce Fedora carbon footprint by identifying areas where we might being wasting resources and optimize software we distribute to end user to optimize users electricity usage.

For example, I was about to write on devel to ask if the EncourageI686LeafRemoval change introduced in F37 could become the default in F39 or F40. We’re building thousands of packages in Koji for ix86 architecture by default, but the vast majority of those are never used/distributed to end users, since we’re not producing ix86 images anymore (as far as I understand).

I also remember a discussion about delivering optimized kernels to users based on supported features of user’s CPU, but I don’t remember is it was deliberated it was impossible or not.

I think there might be more areas which I don’t know in Fedora infrastructure where we can optimize resources and since it’s becoming more and more important to reduce carbon footprint it should be included in Fedora goals for the next years.

5 Likes

That’s easy to fix, just use the term “Lodestar” or “Polestar”. Done.

The tooling encourages more discussion, and does enhance the quality of the discussion. Case in point the use of Discourse vs. Mailing Lists. I won’t go into all the differences, because they are obvious.

2 Likes

For me, being more. workstation rather than server focused, the most important thing is consistent, reliable performance of all desktop environments. Linux is still a geek tool that is unlikely to attract new end-users in large numbers. Any little thing I want to install/tweak ends up needing hours of web searching and mind-reading. I have wasted untold hours fighting with bugs and performance issues. I think the every 6-month relrase cycle smacks of the same featuritis as MS and Apple. I would prefer fewer new features and more attention to making what already exists reliable.

Also, read Stephen Covey on mission statements and then ask yourselves how any of these things can help an employee or contributor decide what to do when a novel or unforseen situation arises.

1 Like

The 7 Habits of Highly Effective People guy? I did a search on what he has to say about mission statements, and everything I came up with is about personal ones, and has guidance like “It reflects a strong connection with your deep inner life and recognition of your deepest and best self.” Is there something in specific you were thinking of?

We’re not planning on rewriting the basic Fedora mission statement or vision at this time — we went through that recently and I don’t see a strong need to revisit. We know where we’re going. Is there something specific you would suggest?

I think our mission, vision, foundations — and the north star statement here provide good guidance for how to approach a novel situation — and, we are working down to specifics from there, as I’ve outlined.

I think we’ve been doing a six month release cycle longer than Apple or Microsoft. (Especially if you trace back to Red Hat Linux before Fedora.) But, I think this misunderstands what we do. Fedora is fundamentally an integration project. We don’t directly develop most of the software we include — rather, we strive to make it all work together in a coherent way for end users, and for others who want to build solutions on a reliable base.

I know Linux can be frustrating some times — but so can all software. The unique thing is that this is our software — not Fedora’s, not Red Hat’s, but collectively a shared world of tools and applications we are building together. Part of Fedora’s role is addressing those bugs and performance issues. That’s a crucial way we give back.

But, this is big. There are many moving parts coming from so many different people, all with their own ideas. In the open source world, change is happening. We have found that integrating it on a regular six-month cadence provides a good balance: faster than that (as in a “bleeding edge” rolling release) and users have to accept large changes at any arbitrary time. Slower than that and change accumulates and accumulates, making it harder and harder to do the integration work and making the leap between versions a big deal for end users.

We have downstream distros — projects that (as in our mission!) build on what we do, and release at a slower cadence. These include Red Hat Enterprise Linux (on a three year cadence) and Amazon Linux (cadence TBD). Fedora ELN and CentOS Stream are Red Hat’s way of making that large jump easier for developers, vendors, and hopefully users. You may find that one of these downstreams matches your personal tastes more. That’s okay!

Or, you could consider running Fedora Linux “N-1” — that is, stay on a release one back from the current. You still get change, but with a buffer for the newest stuff to settle down. And you can help us! Make documentation better. Find the pain points and do what you can to smooth them.

Because — bringing this back to the topic at hand — while we want more users, our goal for this plan is to bring more people into the project of building this amazing, collaborative thing. That will allow us to serve more users, make things better and more polished, and ultimately reach a world where everyone benefits from free and open source software built by inclusive, welcoming, and open-minded communities.

3 Likes

Overarching target to grow active contributors is universal goals and objectives, in my view, and i’m thinking about how we can channel our efforts into the goals.

We would need to extrapolate meaningful success criteria (instead of saying metrics, for now) for each group. This is to match underlying efforts to achieve towards the goals.

Within my remit in Docs team and my interest area (Modularity/EPEL), I think there are a few success criteria (backed by stats we can access) to focus on.

  • Number of views: QuickDocs page
  • Number of documents converted using the latest template (style, header, metadata)
  • Number of download: dnf stat (modular and so on)
  • Number of badges awarded for knowledge sharing (like Sensei)
  • Number of PR
  • Number of visual guide created (on top of text rich content): we need lightweight alternative to Fedora Classroom, so people can craft a short GIF animation in lieu of full-blown video content, which is hard to make)

Using a little bit out of topic analogy in personal fitness, pace (min/km) is my motivator to run consistently and how to keep everything else in check - hydration and recovery. What I’m trying to convey here we need specific goals in each group for concerted efforts. Raw power equals a number of active contributors. Knowing numbers that matter isn’t for competition, but for motivation.

1 Like