Chaos Computer Club - recent events feed

Variational Autoencoders (VAEs) were first introduced as early concept learners in the vision domain. Since then, they have become a staple tool in generative modeling, representation learning, and unsupervised learning more broadly. Their use as analogues of human cognition is one of the first steps towards the understanding of more complex cognitive models leading up to models of human brain function and behavior. As part of a series of talks on cognitive science and deep learning at the realraum in Graz, this presentation will focus on the role of VAEs in cognitive science research. Topics: - Supervised vs. unsupervised learning - Deep Learning basics: classifiers and backpropagation - Autoencoders: architecture, training, embedding, and generative modeling - Variational Autoencoders: statistical latent space, and the reparametrization trick - Training VAEs: loss functions, optimization, and the KL divergence - Concept learning: VAEs in cognitive science https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.realraum.at/realraum-october/talk/LHH3M9/

See you next year! Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/KAB39N/

Take a trip back through time with me as we learn about how bridges happened. Why did we do it at all? Whose idea was it to hook up the then fledgling matrix.org network to the biggest IRC network? How many databases were shredded in the line of duty? This talk features a journey back through time as we look at bridges and integrations that have been built on Matrix over the years, and what we can hope to learn from it. There will be laughter, there will be eyebrows raised, there will be content involving the infamous IRC bridges and there may be a tear or two. There is plenty to dig into here as we go behind the scenes on what it takes to maintain a bridge for the global Matrix community, as well as how the standards of appservices have evolved over the years Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/D83FKE/

The Cyber Resilience Act is a new European regulation that has the main goal to increase European cyber security and resilience, through accountability. A lot has been said about the impacts of the CRA on open source, in particular towards non-profit foundations, but what about organisations - such as Element - that operate complex licensing models with a mix of monetised and non-monetised products? In this talk we endeavour to shine a light on our thinking regarding CRA compliance for our products, as well as implications for the wider ecosystem of vendors and communities building on Matrix. We also introduce our roadmap of communications around the CRA, aimed at those using and building on top of our products. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/AS8JHQ/

Messaging platforms offer to protect user privacy via a variety of features, such as disappearing messages, password-protected chats, and end-to-end encryption (E2EE), which primarily protect message contents. Beyond such features, "untraceable communication" tools for instant messaging protect users from network attackers observing transport layer metadata, which can reveal who communicates with whom, when, and how often. However, unlike E2EE, the effectiveness of these tools depends on large anonymity sets, making widespread user adoption critical. This talk presents a research study with 189 users of messaging apps about their perceptions of "untraceability" as a concept, as well as their opinions on the widespread availability of tools for untraceability. The study explores their perceptions of "untraceability'' from a broad conceptual standpoint; rather than focusing on a particular tool or implementation, we analyze how users reason about what features should be incorporated by two fictitious messaging platforms, Texty and Chatty, to prevent third parties from "knowing who communicates with whom". The results point to a critical gap between how users and privacy experts understand untraceability, as well as tensions between users that see untraceability as a protection to individual privacy and users that see it as a threat to online safety and criminal accountability. Beyond untraceability, I discuss how this research is relevant to the design of messaging platforms that promote privacy as a central value. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/DDKP3V/

Join us for an introduction to the Element Server Suite Community, the simplest way to deploy a Kubernetes-based Matrix 2.0-ready stack, maintained by Element. This workshop will deploy all the components of the suite: a Synapse homeserver with Matrix Authentication Service enabled, a Matrix RTC backend for calls, and an Element web client. You'll be able to try it on your local laptop using your Docker daemon, or deploy it on a virtual machine with a single-node Kubernetes setup. Whether you're new to Kubernetes and Matrix or looking to expand your server capabilities, this session will give you the foundation to get started and experience the latest Matrix features on your own machine. This workshop introduces you to the Element Server Suite (ESS), a streamlined and user-friendly way to deploy a Kubernetes-based Matrix 2.0-ready stack. No prior Kubernetes experience is required ! The suite is designed to be simple and accessible for both beginners and experienced users. During the session, you’ll walk through a hands-on deployment using one of three options: - your own Kubernetes cluster - a virtual machine with a single-node k3s cluster - or your laptop running KinD (Kubernetes in Docker). You’ll gain insight into the architecture of the suite, learn how to set up a self-hosted Matrix server, and explore how ESS can be customized to meet the needs of your community. By the end, you'll have a working Matrix 2.0 environment with a Synapse homeserver, Matrix Authentication Service, RTC backend, and Element web client - all ready to use! Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/8N8HZS/

This talk will introduce the Helen Nissenbaum's theory of Contextual Integrity as a framework for understanding privacy in messaging platforms. Contextual Integrity views privacy not as keeping information secret, but as making sure information flows in ways that match people’s expectations in a given context, or in other words, what feels appropriate to share, with whom, and for what purpose. For example, if Alice shares her live location with Bob through a messaging app, she likely expects the app to use her location only to deliver it to Bob. But if the app also uses her location to target ads, she may feel that her privacy was breached. The problem isn’t that the location was shared, but that it was shared in a way that didn’t match the context or her understanding of how the information would be used. I will explain the theoretical framework with examples of how it can be adapted to identify and explain privacy expectations of particular messaging features, and discuss how it can be applied to interoperable messaging to identify potential privacy concerns. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/PEJYHG/

[NeoBoard](https://neoboard.io/) is an open source real-time collaborative whiteboard built on the Matrix protocol. It functions both as a widget within chat-focused Matrix clients like Element Web, and as a standalone web app that acts as a lightweight, whiteboard-centric Matrix client In this talk, we’ll share how we migrated NeoBoard’s real-time collaboration feature from a peer-to-peer WebRTC implementation to MatrixRTC with a LiveKit backend, a shift that was key to scaling to hundreds of concurrent users. We’ll cover the challenges we faced, the lessons we learned, and the architectural decisions that made the transition smooth. We’ll also touch on what’s still missing and what’s next for both NeoBoard and MatrixRTC. NeoBoard’s use of the Matrix protocol provides a replicated data store for CRDT-based events exchanged between all users collaborating in a whiteboard session. These custom events encapsulate shape and drawing data and are shared via the room timeline, enabling clients to reconstruct a consistent, persistent view of the board over time, even across network disconnections or when joining later. However, while this model ensures eventual consistency, it falls short for real-time interactions that require low-latency feedback, such as live cursor movement or simultaneous drawing. For these use cases, a dedicated real-time transport layer is essential. Using WebRTC, we initially added a real-time layer to NeoBoard that enabled low-latency peer-to-peer communication between users. However, this relied on a full-mesh topology, where each client maintains direct data channels with every other participant. This architecture quickly becomes unsustainable as the number of users grows, due to quadratic bandwidth and connection overhead. Additionally, WebRTC depends on ICE negotiation using STUN and TURN servers to establish connectivity across NATs and firewalls, which is often unreliable and can introduce additional latency and server load. The resulting variability in connection quality and resource usage made it difficult to deliver a consistent experience at scale. Given the success of Element Call's adoption of MatrixRTC and LiveKit, we decided to move forward with an alternative real-time data layer based on these technologies that, given enough backend resources, can scale indefinitely. In this talk, we’ll cover: - The architecture and design of NeoBoard’s real-time layer - An overview of relevant MatrixRTC spec proposals and their role in our migration - Implementing MatrixRTC with a LiveKit backend, including: - Session membership management - RTC focus handling - Infrastructure considerations and requirements - Live demo - Key challenges and what’s next Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/TM3ASG/

At Oodrive, our mission is to offer a secure and sovereign solution to manage our partners' most sensitive data. Building on this foundation, we are developing a secure collaboration platform that fully aligns with these values. Collaboration cannot happen without communication, therefore we decided to integrate a secure chat in our product. We chose the Matrix protocol to achieve that goal. Now that a first version is out and available to our partners, we want to share with the community how we integrated matrix in our product, and the challenges we overcame. Namely : - How we provision user accounts in our matrix instance. - How users authenticate transparently in the chat. - Make a demonstration to illustrate the seamless integration Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/A9YW3B/

Most of the discussions around privacy laws have been focused on the protection of data and condition of processing. But some others rights are starting to be litigated more widley and will soon play a bigger role for IT systems managers. This talk will focus on the GDPR article 16 (right to rectification), and how the current Matrix spec is inadequate for servers operators aiming to let people change their username, since MXID can't be changed (as of this proposal writing in June 2025). Based on the ongoing work done by the [Fedora community](https://fedoraproject.org/) on the [Fedora Username Change project](https://communityblog.fedoraproject.org/call-for-volunteers-regarding-the-fedora-username-change-project/), on recent legal cases in Europe, we will see why the adoption of [MSC 2787 (Portable identites)](https://github.com/matrix-org/matrix-spec-proposals/pull/2787) should be more than a nice to have, and how it intersect with questions of privacy and anti-discrimination. The talk will also touch on the topic of SSO configuration and downstream consumers best practice to deal with such changes. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/PTAPZE/

Tammy is a multiplatform Matrix Messenger designed to be redefined by you. Want a different room list layout? Need a messaging interface tailored to a specific use case? Tammy’s extensible architecture makes that not only possible - but easy. In this talk, we’ll showcase how Tammy empowers developers to create radically customized Matrix experiences through its extension system. We’ll walk you through how we’re using Tammy to build Timmy, a TI-Messenger variant tailored for a very specific user group, with a completely different look and feel - all without forking or rewriting the core client. Expect a live demo, some under-the-hood insights, and a glimpse into the roadmap: spaces, audio/video, Matrix 2.0 and more. Whether you want a more focused UI, a minimalist mobile mode, or something wildly experimental, Tammy gives you the tools to build it. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/3YMJFP/

In this talk I will present my journey in developing a decentralized and secure communication platform for a university with thousands students and faculty members. Using Matrix and Element as foundation I tailored the solution to meet academic needs, ensure data privacy and integrate with some systems, bots, widgets etc. The session will cover: - Why Matrix was chosen over other platforms. - Technical stack and deployment model. - Customization of Element and federation decisions. - Challenges (including authentication, onboarding, moderation and compliance). - Real-world impact and feedback. - Future plans including bridging, widgets and improvements. This case study shows how Matrix can empower large academic communities and help other institutions explore privacy-respecting and scalable alternatives to traditional messaging platforms. By the way, it is real story. :) Universities face growing challenges in balancing seamless communication, data privacy and platform sovereignty. This talk presents the story of how I built a Matrix-powered communication platform based on Element for a university community with thousands students, faculty and staff. I will share my hands-on experience deploying a self-hosted Matrix homeserver, using the Element client for academic use and integrating the system with some infrastructure. You'll learn about: - The decision-making process: Why Matrix and Element? - Technical architecture and deployment at scale. - Real-world challenges in onboarding, federation, moderation and support. - Custom feature additions and user adoption results. - Lessons learned and recommendations for other institutions or organizations considering Matrix. This session is ideal for technologists, sysadmins, public sector decision-makers and open-source advocates exploring decentralized alternatives to Slack, Teams or WhatsApp in education or large organizations. So yep, feel free to ask! Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/MLHCY7/

fairkom has developed a [matrix]-based school messenger for the Department of Education of the German state of Rhineland-Palatinate to serve up to half a million users. Schulchat RLP has one code base for all platforms and is available as app on Apple and Google Play Stores. What makes this solution unique is the deep integration of the identity and user management system from the "Bildungsportal RLP", ensuring seamless use for teachers, students, and even parents - from primary to vocational schools. For Schulchat RLP we have added a sophisticated address book into the messenger with role based communication templates and adapted the client application, based on FluffyChat, to meet pedagogical requirements, such as enabling active read confirmations. We will be presenting the project lifecycle, challenges with E2EE, SSO, hosting and how we manage user feedback and daily operations. fairkom is a silver member of the [matrix] foundation and offers custom messaging solutions and scalable hosting. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/QTBJP3/

A thank you from the organisers. We hope you enjoyed your day! about this event: https://talks.python-summit.ch/sps25/talk/QW7U3F/

Our Lightning Talks are open to everyone 😊 about this event: https://talks.python-summit.ch/sps25/talk/VG8HT8/

Networks are all around us, shaping phenomena like epidemics, communication, and transportation. In this talk, we will explore how real-world problems can be analyzed and solved using graph-based methods and simple algorithms. Drawing from examples such as trade networks, corporate structures, and historical data, I will demonstrate how network analysis reveals insights that would otherwise remain hidden. Using NetworKit (and NetworkX), we will analyze real-world datasets to answer questions like: What does the core-periphery model reveal about trade networks? Could we have predicted that Moscow will become Russia's capital? How do corporate hierarchies differ from interaction hierarchies within organizations? Throughout the talk, I will introduce key concepts in network analysis and showcase Python as a tool for research. Attendees will have access to all datasets and code, enabling them to replicate the analyses and apply these techniques to their own projects. This session is designed for Python enthusiasts with an interest in data science, networks, and/or applied research. about this event: https://talks.python-summit.ch/sps25/talk/GZHTJX/

Activation functions are fundamental elements of deep learning architectures as they significantly influence training dynamics. ReLU, while widely used, is prone to the dying neuron problem, which has been mitigated by variants such as LeakyReLU, PReLU, and ELU that better handle negative neuron outputs. Recently, self-gated activations like GELU and Swish have emerged as state-of-the-art alternatives, leveraging their smoothness to ensure stable gradient flow and prevent neuron inactivity. In this work, we introduce the Gompertz Linear Unit (GoLU), a novel self-gated activation function defined as `GoLU(x) = x Gompertz(x)`, where `Gompertz(x) = exp(−exp(−x))`. The GoLU activation leverages the asymmetry in the Gompertz function to reduce variance in the latent space more effectively compared to GELU and Swish, while preserving robust gradient flow. Extensive experiments across diverse tasks, including Image Classification, Language Modeling, Semantic Segmentation, Object Detection, Instance Segmentation, and Diffusion, highlight GoLU's superior performance relative to state-of-the-art activation functions, establishing GoLU as a robust alternative to existing activation functions. about this event: https://talks.python-summit.ch/sps25/talk/PNL9LP/

Germany’s digital health infrastructure has long been built on the principles of sovereignty, open standards, and interoperability. With the electronic patient record (ePA) and KIM for secure data and document exchange, Germany established a federated ecosystem using standards such as FHIR and OIDC with sectoral identity providers. The recent introduction of the TI-Messenger (TIM) extends this ecosystem to real-time communication — now connecting over 25 million insured citizens through the RISE TI-Messenger, directly integrated into their ePA app (ePA FdV). Matrix serves here not as the foundation of sovereignty, but as a loyal enabler within an existing sovereign architecture. Operating own Matrix instances (e.g. via the RISE Cloud) ensures data locality and governance under German regulatory control, while federation allows secure inter-institutional communication. This approach balances federated flexibility and controlled trust domains, distinguishing Germany’s model from purely centralized or global platform solutions. The talk discusses the integration of Matrix within the sovereign health stack — ePA, sIDP, KIM, TIM, and OIDC — and explores risks and resilience in large-scale federation. It illustrates how Germany’s health communication layer evolves from document exchange to conversational care, preserving sovereignty while embracing interoperable, open technologies. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/CJSBUZ/

NATO organised a TaskForce-X Baltics event where we very quickly contracted, deployed and integrated telemetry and video streams from uncrewed platforms. This required a lot of coordination and this was done through our on-prem Matrix capability. I would like to share my experience by using Matrix to support this very demanding operational project. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/7ZXT8V/

Trixnity is rapidly emerging as a powerful, flexible, and fully asynchronous SDK for building Matrix applications using Kotlin Multiplatform. What many don't realize: Trixnity is already in widespread use - potentially powering apps usable by over 50 million people. It’s embedded in multiple Matrix-based messengers, though you’d only know it by digging into app license disclosures. This talk shines a light on Trixnity’s hidden but significant role in the Matrix ecosystem and why more developers should pay attention. This session also presents the current state of Trixnity in 2025, highlighting key milestones, recent developments, and its growing adoption in real-world applications. Licensed to the public under https://creativecommons.org/licenses/by-sa/4.0/ about this event: https://cfp.2025.matrix.org/matrix-conf-2025/talk/BCLRR3/