Objective Review: More (active) SIGs, fewer images

We’re working on Fedora Strategy 2028 — our next five-year plan. We are now reviewing those Objectives and their associated Impact. Read this guide for details on the current planning phase.

This Objective is part of the Theme “We build on the success of Fedora.next” and the Focus Area Editions, Spins, and Interests. For general discussion of this focus area, please see the topic Fedora Strategy 2028: Focus area review (Editions, Spins, and Interests).

Objective and Impact

Objective: More (active) SIGs, fewer images.
Impact: We grow strong special-interest teams and networks into other communities — neuroscientists, roboticists, teachers…

Based on people’s reactions in the initial strategy draft thread, the terse phrasing we used for this Objective isn’t exactly self-explanatory. Maybe we can find a better way to say it in this discussion. I kind of like the short, “punchy” version, though — and maybe that plus some short explanation is enough.

SIGs — Special Interest Groups — have been a part of Fedora since very early on. Successful SIGs become thriving sub-communities of their own, and bring new people who share similar interests into the project. Continuing this and building more infrastructure and support for these teams will clearly grow the active contributor base.

So what’s this about “fewer images”?

Well, as currently set up, we don’t give much guidance. Here’s the current documentation: Fedora Project Wiki: Creating a SIG — something I wrote in 2016 and which hasn’t really been updated or cared for. It basically says:

  1. Make a wiki page
  2. Make a mailing list (or don’t), and create a ticketing system if you want
  3. Maybe a regular IRC meeting? I dunno?
  4. Good luck!

This minimalist approach was a reaction to a previous process which had turned out to be too heavyweight in practice. But it probably veers too far in that direction: it doesn’t really offer any concrete support or anything to do.

One thing a SIG might do is create a new Spin — a bootable Fedora Linux OS deliverable showcasing software related to their particular area (for example: robotics, neuroscience, graphics & design, audio production, security). When we did the big Fedora.next website redesign, we decided to call these particular variants “Fedora Labs”.[1]

This is a pretty intensive, heavyweight thing to do, and requires resources every Fedora Linux release. And the thing is… they’re not really being used. The software that’s packaged by SIG teams certainly is, but not very many people are consuming it that way. I’d make graphs, but none of the Labs even breaks 0.1%. That isn’t terribly surprising: the computer-lab use case isn’t common anymore, and there’s no other particular reason to install the Lab instead of picking your favorite desktop environment and installing the relevant software on top of that.

But, the people working on the Design Suite and Robotics Lab and CompNeuro are awesome. We should figure out some other way for them to offer useful curated collections of software to users — maybe as groups in software installation tools, maybe something else. We don’t want to break these teams by taking away one of the few things we offer as a focus for SIG energy without anything better to offer.

I hope this makes sense now — what do you think?

Our goal now

For this Objective and related Impact, validate that:

  1. If the Impact is achieved, it’s reasonable to expect an increase in active Fedora contributors.
  2. Success in the Objective logically results in the intended Impact.
  3. That link is reasonably sufficient — that is, it represents everything needed to have the Impact.
  4. While there might be other ways to have similar Impact, the chosen Objective is the right one for Fedora right now.
  5. The wording is precise and clear. The Objective is concrete, and the Impact is (at least a little bit) inspirational. Together, they fit into this Focus Area.

Bonus. If you can improve the longer explanatory paragraphs at the top of this post, that’s helpful too!

As outlined in the roadmap, this post will close in one month.


  1. That’s because at that time, it was still not uncommon to have physical classrooms full of computers which you might take physical media into and literally boot up the Design Suite in order to teach Inkscape. Previously, these kinds of deliverables were all in a list with the other Spins, which mostly focused on a particular desktop technology — KDE Plasma, Cinnamon, Xfce, etc. That made these relatively popular desktop environments harder to find, so it seemed like splitting them out would make things more clear… but in practice people have been asking confused questions about the terminology ever since, so we’ve officially abandoned it. ↩︎

3 Likes

Totally onboard with this one. I think we make way too many images that no one ever uses. Right now we only really have comps groups to manage package groups, but perhaps we could make that easier to manage or come up with a better process.

Note that there are still cases you do need an image, like if you want people to try your desktop env live before installing or if you have a small device you need to copy an image to a microsd card.

Not that we need to have details here yet, but crazy ideas: a live media that lets you use ostree to switch between any desktops you want to try out live. Anyhow.

4 Likes

I wonder to what degree some ISOs could be replaced with better marketing for the specific packages that would be in these labs. If people are using the specific solutions these SIGs are working on even though it’s not in the from of using the lab, I think it could still wok. Would be interested to hear from some of these teams.

1 Like

Members in SIGs have some affiliation with Labs. Among several SIG and Labs related to scientific computing, I see consolidation of those could bring our objective into action. 69 SIGs exist. If we map out SIGs to Labs (not always related), we could discover potential areas of consolidation.

Regardless of whether physics, biology or astronomy, researchers (either academics or private institutions) use a set of database, big data framework, data analysis and authoring tools for validation of test and publication. They are pro-FOSS users, in general.

  • Spark, hadoop, kafka
  • R libraries
  • Latex, Jupyter and so on

SIGs could be consolidated into that dimension - base tools for scientific research (regardless of domain).

Until we have sizeable number of users (can’t say what is sizeable, but 1% of total user base?), consolidation could help more active SIGs that share common use cases and outcome (user story and publication).

One specialist Lab that does not come into any of above category is security. Without security lab, I have no way to discover tools for networking and security audit in one place. What we need is tutorials using security lab.

About fewer images, I like the idea of @kevin

Just an example to take my spin on it, we could think about a live media that allows users to;

  • Pick your choice of desktop environment
  • Select domain or use cases: scientific research, pen-tester, visual FX, graphic design and so on
  • Choose a grouplist (either dnf or graphical package manager like GNOME software or Discover): we just need to regroup a list of software for grouplist package groups
  • For EPEL users in enterprise, we could have similar categorization using grouplist
  • Customization: we could have further categories if needed

Afaik, this is what classical Fedora Everything ISO always provided.

I admit it has been quite long time since I last installed Fedora from scratch, and I don’t remember exactly what are the default choices in the netinstall image now, but there should still be this feature in one of the images we build?

I think the main issue which caused us to turn to respins was that not all configurations could be expressed as package groups to install. There are actions taken through the install process, which are not easy to reproduce via package installation.

I believe it is a bad thing to have such hard to reproduce things. And I think in last years we made an effort to reduce those steps. That’s why I think it is a right time to raise this topic again and try to see if reasons to create custom respins are no longer there.

And additionally, nowadays we can also use things like Ansible System Roles for configurations which dnf install alone can not solve.

1 Like

Thanks for sharing insights I wasn’t aware about Everything ISO until today.

I had a go with Everything ISO using the ‘Fedora Custom Operating System’, which give a huge selection of additional software groups. It is a completely customizable OS. Blown away.

In addition, it is interesting to see Ansible use cases with system roles.

1 Like

This topic was automatically closed after 25 days. New replies are no longer allowed.