FluConf 2025

The Program

FluConf is a fully online event featuring sessions in a variety of formats. Our program will feature pre-recorded videos, live-streamed talks, and blog articles.

We're providing a schedule to allow sufficient time for reading, viewing, and follow-up discussion with our contributors on the fediverse, even though many of the sessions can be enjoyed asynchronously.

We'll update the details of individual sessions on an ongoing basis as their resources become available online. We suggest that you follow our account on the fediverse for regular updates throughout the weekend.

Day one
Saturday, February 1st (UTC)

Enable JavaScript to view the schedule in your local time zone

Zen and the Art of Multicast - An Inquiry Into Technology and Values

By Brett Sheffield (he/him)
@dentangle@chaos.social
www.librecast.net

Format: Pre-recorded video hosted on Librecast.net, with transcript available

Visit session URL âž”

Day one (2025-02-01)
- UTC

A conversation about technology "quality", and how changing our lens and making Human Rights a fundamental requirement changes outcomes and how we view quality, technology, and the world around us.

How we measure quality depends entirely on our values. Putting human values rather than shareholder value at the heart of our system requirements leads us to fundamentally different designs for our software, our networks and our society.

Whether we are designing, building, or procuring technology we need to be clear about our values before we start. By requiring concepts like privacy, accessibility and consent from the outset we can discard many common antipatterns, protocols and approaches, leading us to novel solutions that better support our needs.

The Librecast Project is working at the intersection of Human Rights and Technology. We believe that our fundamental rights depend on universal group communication without borders, censorship or netblocks, and we are working to build the Next Generation Internet to support that vision.

See full session details

A tale of two compilers

By Wim (he/they)
@wim_v12e@scholar.social
limited.systems

Format: Live presentation via Zoom, recording hosted on archive.org

Visit session URL âž”

Day one (2025-02-01)
- UTC

I propose a talk about two compilers I have created, to investigate compiling Fortran programs to Uxn, a tiny stack-based VM with 64 kB memory and 8-bit instructions created by 100 Rabbits (https://100r.co/site/uxn.html)

The talk focuses on two languages: venerable Fortran; and Funktal, a strict, statically typed functional language for Uxn; and on their compilers: a general purpose, "serious" source-to-source compiler for Fortran, and a custom, "frivolous" compiler from Funktal. The link between the two will become clear in the talk.

The main reason for doing this is to show that modern compilers and programming languages can be frugal and resource constrained: there is no need for huge amounts of memory, disk space, network bandwidth and compute power purely to create a program, and for most programs also not to run it.

See full session details

A luddist criticism of "Open Source"

By tante (he/him)
@tante@tldr.nettime.org
tante.cc

Format: Live presentation on BigBlueButton, recording hosted on archive.org

Visit session URL âž”

Mirror available on Youtube

Day one (2025-02-01)
- UTC

People interested in Fluconf will also be interested in Open Source or to be more precise Free/Libre Open Source Software. Everyone has been using Open Source for many years but many from this community have shifted to trying to run their personal infrastructures on non-proprietary pieces of software and some even hardware.

This is already an important shift towards freeing ourselves and each other from corporate dominance but is it really doing enough? Are our licenses protecting the values we actually care about?

Coming from a luddite background I want to dive into a a bit of a critical reading of the values that we use codified mostly as licenses: Are they really enough? Which aspects are they missing and why? And what are the consequences of those omissions?

I'll try to end up looking at mechanisms of integrating luddist principles in software projects. Can they offer additional safeguards?

This sessions doesn't claim to have all the answers. But hopefully some good questions and a few ideas of where to go.

See full session details

The Degrowth of Social Media

By KC Legacion (he/they)
@kclegacion@scholar.social

Format: Watch-party hosted on BigBlueButton, original video on archive.org

Visit session URL âž”

Mirror available on Vimeo

Day one (2025-02-01)
- UTC

In this short film, a master's graduate presents his research on social media and the climate crisis via an online livestream. All is going well until supernatural glimpses into the near future interrupt the presentation.

See full session details

Forking the historical text: Incarnations, autofabulation, and beyond

By Erica (ERN) (she/they)
@riveraerica@writing.exchange
riveraerica.com

Format: Article

Visit session URL âž”

Day one (2025-02-01)
- UTC

Drawing on forking practices in software development, sampling practices in audio production, my work with collage art, and the speculative archival work of Saidiya Hartman, I've been trying to develop new artistic forms more sensitive to the networked nature of our histories and futures. Enter incarnations: a literary form in which you rewrite a historical text and set it in a far future, and write it from the perspective of your future self. The form also functions as a strategy for intervening in processes of "copyright"-making by challenging traditional conventions of authorship and (self-)publishing. Finally, incarnations also fall under a new, umbrella genre of speculative writing I call autofabulation, conceived as an explicitly anti-colonial, anti-imperial, and anti-oppressive artform intended to allow its practitioner to actively work towards liberation. This event will also include generative prompts that encourage participants to engage in their own forms of speculative archival work, including—but not limited to—writing their own incarnations.

See full session details

Co-op Cloud: Grow Your Own International Democratic Tech Federation

By Calix (they/them)
@coopcloud@social.coop
coopcloud.tech

Format: Live Owncast stream, recording now available via peertube

Visit session URL âž”

Day one (2025-02-01)
- UTC

The Co-operative Cloud is an international, democratic federation with members representing developers, administrators, and users of our system for setting up community technology infrastructure.

Since 2020, our tools and configuration have been used by worker co-operatives, trade unions, artistic collectives, friend groups, other libre software projects, & home-hosters to provide privacy-supporting, user-empowering platforms to their members and supporters. Co-op Cloud has been used on recycled servers and upcycled laptops as well as traditional "cloud" providers, helping extend the useful life of otherwise-forgotten hardware, & squeeze the most out of cheapest hosting providers' cheapest services.

In this livestream (hosted on a Co-op Cloud Owncast instance) we'll get into detail on the governance side of Co-op Cloud: the issues we saw with other models of community-building and software maintenance which led us here, our inspirations in building the Co-op Cloud Federation, the resources that have supported us along the way, & some tales of the rough and the smooth of our experiment so far.

We hope this session, including Q&A, will be useful to anyone involved, in or considering starting, a public-interest technology project.

See full session details

Day two
Sunday, February 2nd (UTC)

Improving Diversity and Inclusion in the Open Source Community

By Rowland Mosbergen (he/him)
@rowlandm@disabled.social
www.practicaldiversity.org

Format: Live presentation on BigBlueButton, recording hosted on archive.org

Visit session URL âž”

An accessible version of the presentation's slide content is available on Google Docs

Day two (2025-02-02)
- UTC

Many open source research communities want to be more diverse, equitable and inclusive.

However many DEI strategies and plans fail because they don’t understand:

  1. that most DEI frameworks should talk in concrete terms, not in abstract terms,
  2. what is systemic discrimination and why it is like poison,
  3. that systemic and subtle discrimination accumulates over time,
  4. what Intersectionality really means and how it adds more nuances that need to be addressed, and
  5. that providing opportunities, creating safe spaces, and sharing power are the antidotes to discrimination.

This presentation will explain these concepts and show how you can make a difference as an individual, organisation, or a community in practical ways that can be implemented easily.

It will also show examples of how these practices have embedded into open source communities, fellowships, conferences, recruitment and internship programs - and the impact that they have had.

See full session details

Ambient CI engine: an introduction

By Lars Wirzenius (he/him)
@liw@toot.liw.fi
liw.fi

Format: Live presentation on BigBlueButton, recording hosted on archive.org

Visit session URL âž”

Day two (2025-02-02)
- UTC

Ambient is a new CI engine being developed that is local-first, where CI runs in a VM under QEMU. It makes it safe and secure to run CI for projects where you can't necessarily trust the code to behave well.

It is very early days for Ambient. At this stage, it is a command line program you run manually, not a service that reacts to changes in a version control repository. The intent is that Ambient will support use as a service in the future.

This talk will discuss the motivations that led to Ambient and major decisions for its software architecture. The talk will explain why the developer thinks Ambient is good for running CI for other people's software.

See full session details

Techno-logy, luddism, feminism (and what about AI ?)

By Khrys (she/her)
@Khrys@mamot.fr
khrys.eu.org

Format: Article

Visit session URL âž”

Day two (2025-02-02)
- UTC

A rough translation in english of the talk I held in french at Capitole du Libre in 2024.

See full session details

Neo-Luddism in a developing country can be odd

By ohio olarte (he/him)
ohio.araw.xyz

Format: Article

Visit session URL âž”

Day two (2025-02-02)
- UTC

In this article, I aim to talk about my personal experience using neat tools in FOSS, how it affects my choices when obtaining a new piece of technology, and how it might (or might not) affect the society at large, especially if it is to be pitted against the "convenience" delivered by Big Tech products and services, all within the context of where I live in, the so-called Philippines.

See full session details

Phosh 2024 in Retrospect

By Guido (he/him)
@agx@ruhr.social
phosh.mobi

Format: Article

Visit session URL âž”

Day two (2025-02-02)
- UTC

Phosh is group of people that care about freeing the mobile devices in our pockets from proprietary operating systems. It is also a graphical shell targeting mobile devices running a (close to mainline) Linux kernel. This currently includes the PinePhone and Librem 5 as well as devices initially shipping with Android (like the OnePlus 6/6T or Pocophone F1). It is shipped in Linux distributions like postmarketOS or Debian/Mobian and can also be used for devices that currently don't have a mainline Linux option by repurposing the Android kernel via e.g. the Droidian project. It builds on well known components from the Linux ecosystem like GTK, GNOME, the wlroots compositor library and others (see https://phosh.mobi/posts/phosh-overview/).

In this post we give a short look back on what happened in the project during the last year to give more people a way to use their devices as they wish for as long as they want, free from tracking and surveillance.

See full session details

Scaling Community Conversations and Decisions

By benjamin melançon (he/him/ours)
@mlncn@social.coop
agaric.coop

Format: Live presentation via BigBlueButton

Visit session URL âž”

Day two (2025-02-02)
- UTC

How can a group of thousands of people talk about and decide anything? How's this 'community' concept supposed to work at scale, even in theory?

Any Free/Libre Open Source Software project will have elements of do-ocracy (rule of those who do the work), but not all decisions should devolve to implementors. A better ideal is that decisions should be made by the people who are most affected.

Particularly when a decision strongly impacts more than those who carry it out, we need better ways of making decisions that give everyone their say. This starts by letting people by heard by everyone else. Fortunately, we can scale conversations and decisions in a fair and truly democratic way.

  • Learn why meritocracy ("rule of those with merit") is a bogus and harmful concept.
  • Add sociocracy and sortition to your vocabulary and understand how these esoteric concepts can help our communities scale.
  • Get introduced to (and give feedback on) Visions Unite, free software for mass communication mediated by the participants.

The most important things we have to do need to be done together. And our work together is most powerful when we make decisions about it together.

See full session details

Safety Critical Machine Learning Does Not Exist

By simplymath (he/him)
@simplymath@social.cryptography.dog
simplymathematics.github.io

Format: Recording of live BigBlueButton session now hosted on archive.org with slides

Visit session URL âž”

Day two (2025-02-02)
- UTC

Machine learning models -- deep neural networks in particular -- have performed remarkably well on benchmark datasets across a wide variety of domains. However, the ease of finding adversarial counter-examples remains a persistent problem when training times are measured in hours or days and the time needed to find a successful adversarial counter-example is measured in seconds. Through an evaluation of many pre-processing techniques, adversarial counter-examples, and neural network configurations, the conclusion is that deeper models do offer marginal gains in survival times compared to more shallow counterparts. However, we show that those gains are driven more by the model inference time than inherent robustness properties. Using the proposed methodology, we show that modern machine learning models are hopelessly insecure against even the most naive of attackers.

See full session details

The Swiss Cheese Model: How Infosec Must Learn From Pandemic Response

By noah (he/him)
@text@posting.institute
covid.tips

Format: Article

Visit session URL âž”

Day two (2025-02-02)
- UTC

As a digital security professional and a pandemic activist it has been tremendously revealing to me to see how my professional community has responded to the pandemic. After a brief year of "safe mode" conferences and online trainings, the "back to normal" urge overrode many people's threat models and notable figures in the community began joking about getting covid at conferences, or client engagements, or work trips and so on. This proposal is something of a manifesto aimed at reminding the security community (and indeed the technology community) about our commitments to Defence in Depth, and drawing comparisons between still-successful pandemic interventions and how we can apply these same techniques to information security...and a plea for a new kind of cybersecurity community, one that aims to work in solidarity with our users rather than in spite of them, one that strives to prevent digital as well as physical social murder.

See full session details

Daring to dream: Reclaiming our connections and building our resilience

By Anna e só (they/them)
@anna@friend.camp
notapplicable.dev

Format: Article

Visit session URL âž”

Day two (2025-02-02)
- UTC

Reflecting on attending FOSDEM 2024 as a disabled keynote speaker, coming to terms with the realization that we can’t move forward if we keep focused on just working on improving the parts of a whole and never address the whole; radical change won’t happen if we keep limiting ourselves to dream mediocre dreams. We have to dare ourselves to dream bigger, higher, better. We have to take care of each other.

See full session details