~/conference

DevConf.CZ 2026

open source contributors and maintainers, platform engineers, kubernetes engineers, devops engineers, site reliability engineers, security engineers, ai and ml engineers

distroless container builds and automation; rootless containerization with Podman; GitOps for virtual routers (vRouter Operator); DevWorkspace and Devfile cloud workspaces; state-first end-to-end CI testing; eBPF and OpenTelemetry network observability; local small language model deployment and fine-tuning; SBOMs and dependency stack visibility; live-patching LLM facts with ROME; post-process OCI image layer optimization; integrating post-quantum crypto modules; secure automated CVE remediation pipelines

titleauthorlength
Keynote: How I learned to stop worrying and love CVEsValentin Rothberg, Stef Walter35m
Open source generates ~50,000 CVEs this year (~130/day) largely driven by AI, so the response is full automation: Hummingbird builds minimal distroless…
Talk in English at D105 (capacity 300) on
These bootc are made for mailin'Michael Scherer36m
Builds an image-based mail stack using bootc images (Fedora Image Mode) and Stalwart for SMTP/IMAP, fronted by Caddy and run with Podman; builds are automated…
Talk in English at D0206 (capacity 154) on
Hacking Your Smartwatch: Building an Open Wearables EcosystemJozef Mlich36m
Open-source wearable tooling: PineTime, Bangle.js, AsteroidOS and Amazfish are shown with concrete specs and tradeoffs — PineTime: Nordic SoC 64 MHz, 64 KB…
Talk in English at E105 (capacity 70) on
Architecting Integrity: Responsible Software Evolution in the AI EraPreethi Thomas, Clément Verna30m
Claims a repeatable five‑step “AGENTS” workflow grounded in a Harvard agent framework and applies it to a CI/CD prototype with an orchestrator agent, an…
Talk in English at E105 (capacity 70) on
Zero copy migration from VMware to KubernetesMartin Nečas27m
Forklift is a Kubernetes operator that migrates VMs to KubeVirt using virt-v2v for guest conversion and NBDKit for mount-and-convert, with options VDDK…
Talk in English at D0207 (capacity 90) on
Scaling Cloud Development Environments on K8s: Enterprise-Scale Adoption LessonsIlya Buziuk34m
Kubernetes-based cloud dev environments are treated as Kubernetes custom resources, using the DevWorkspace Operator and the Devfile spec to standardize…
Talk in English at D0207 (capacity 90) on
Stop Re-Downloading Your Container Images: Content-Based Layers with ChunkahClément Verna25m
Tool called chunkah post-processes OCI images to regroup files by package so a 5 MB Vim update no longer forces re-downloading a 500 MB instruction-based…
Talk in English at D0206 (capacity 154) on
Hands on: How to engage Gen Z and Alpha to Free SoftwareBonnie Mehring34m
Two FSFE programs target kids: Ada & Zangemann (ages ~6–12) is a Creative Commons working book plus movie and GitHub repo used in classrooms and adopted by the…
Talk in English at E105 (capacity 70) on
The risks of AI for skill formation, and how to mitigate themMário Fernandes31m
A controlled study has 50 participants learning the lesser‑known Python async library Trio with or without an in‑IDE AI assistant (Cosmo backed by ChatGPT)…
Talk in English at E105 (capacity 70) on
Harnessing tenure variety within a teamKarolina Kula21m
Anecdote: a pull request sits open three weeks after a junior rewrites it four times while seniors demand architecture and naming changes. Cited numbers…
Talk in English at E105 (capacity 70) on
Podman 6.0 — The Next Evolution of Rootless ContainerizationDavid Darrah30m
Podman 6.0 replaces CNI, slirp4netns and BoltDB with Netavark, Pasta and SQLite3, moves networking to nftables, supports multiple IPs per container, and…
Talk in English at D0207 (capacity 90) on
Controller-Native Trace Propagation for Delivery AnalyticsJosiah England37m
Implements controller-native trace propagation on a Tekton/Conflux delivery platform: controllers inject W3C trace context into Kubernetes resource annotations…
Talk in English at D0206 (capacity 154) on
Stop Looking for the Perfect Prompt: The Design-First Workflow for Coding AgentsMark Kemel33m
Argues the "perfect prompt" is a myth and prescribes a design-first, stage‑one workflow: write a Markdown spec as external memory, use Cursor/Claude Code…
Talk in English at D0207 (capacity 90) on
OpenShift CI: What if we stopped retesting everything all the time?Petr Muller35m
OpenShift CI uses prow jobs and Tide retest to speculatively re-run PRs when base changes, but 2/3 of PRs merge without any retest and only 0.8% see additional…
Talk in English at D0206 (capacity 154) on
How AI helped us ship updates in a Linux distroTomáš Tomeček, Laura Barcziová37m
Ymir is an end-to-end agentic service that produces merge requests to backport CVE fixes into RPMs but never auto-merges them; agents run in OpenShift pods…
Talk in English at D105 (capacity 300) on
From Laptop Chaos to Fedora Cloud: Quadlets and ContainersDavid Duncan33m
Shows how a messy laptop stack (Postgres, Redis, Grafana, agents) becomes a reproducible Fedora Cloud Edition deployment by converting each service into a…
Talk in English at D105 (capacity 300) on
Being a product owner in an open source projectEwoud Kohl van Wijngaarden35m
Compares concrete upstream/downstream workflows: Foreman, Pulp, Candlepin feed a Satellite downstream with 3-month upstream / 6-month downstream release…
Talk in English at E105 (capacity 70) on
Surgery on a Brain: Live-Patching Facts in LLMs with ROMEJakub Reš, Matej Olexa35m
Rome edits a specific factual association by computing a Rank-One Model Editing update: locate a single transformer layer (often around layer 13), compute a…
Talk in English at D0207 (capacity 90) on
The practical implementation of Documentation as CodeChristian Lölkes35m
Argues for treating docs like code: store docs in Git for full traceability (Git blame/bisect), version and release docs alongside code with semantic…
Talk in English at D0206 (capacity 154) on
Openstack Continious Performance tests (CPT)Supraja Manda18m
Automated pipeline uses Prow CI to trigger Ansible-based cluster installs (JetBrew) and run Browbeat workloads against Nova, Cinder, Neutron and Barbican…
Talk in English at D0206 (capacity 154) on
Inside MoE Optimization: A Profiler-Guided Tour of torch.compile and vLLMParshant Sharma24m
Shows a concrete pipeline that combines torch.compile + vLLM with hand‑written custom ops and CUDA graphs, yielding a fused MOE kernel that drops ~15–20 kernel…
Talk in English at D105 (capacity 300) on
Building Reliable AI Assistants with Small, Self-Hosted Language ModelsJose Angel Morena Simon34m
Built a production assistant that runs entirely on self-hosted models using a decomposed pipeline: four-layer architecture (action classification, tool…
Talk in English at D105 (capacity 300) on
Network Observability with eBPF and OpenTelemetryDominik Süß34m
Demonstration of the OpenTelemetry eBPF Instrumentation (OBI) network subsystem shows eBPF programs on hosts tracking per-flow bytes and direction, exporting…
Talk in English at D0207 (capacity 90) on
GitOps for the Free Cloud: Bridging Kubernetes and OpenStack with ORCMartin André34m
ORC is a Kubernetes operator: a set of CRDs and controllers that declaratively manage OpenStack resources using OpenStack Resource Controller (ORC) built on…
Talk in English at D0207 (capacity 90) on
An accessible approach to quantum algorithms with QFrame and QrispFintan Bolton33m
Eclipse Qrisp is an open‑source Python client that expresses quantum algorithms as usual logic and arithmetic; the layer QFrame builds high‑level quantum…
Talk in English at D0207 (capacity 90) on
Building a data analytics platform on strong opinions held looselyIvica Kolenkas35m
A three‑to‑four‑person team builds and operates a self-service data analytics platform using Terraform, Snowflake, dbt and Airflow, serving ~620 active users…
Talk in English at D0206 (capacity 154) on
Enhancing Upper Limb Rehabilitation with Unsupervised Machine Learning ModelsPriscila Gutierres28m
A pipeline that ingests eight joint-angle streams from a Unity AR/VR puzzle game, extracts median and variance into 16 features, applies PCA to 3 components to…
Talk in English at D105 (capacity 300) on
How to Analyze Terabytes of Data from GitHub Archive at High SpeedAliaksandr Valialkin35m
GitHub Archive data at https://gharchive.org contains about 1.3 billion events and roughly 2 TB of JSON‑lines per year and wide events can exceed 400 fields…
Talk in English at E105 (capacity 70) on
On QUIC, HTTP/3 and tech stack unificationGabriel Michael Homa35m
HTTP/3 moves transport into UDP and combines TLS so connections resume across network switches via connection IDs, removes head‑of‑line blocking so multiple…
Talk in English at E105 (capacity 70) on
Local LLMs on low end hardware - a practical perspectiveJames Freeman35m
A second‑hand workstation with 32-core, 64GB RAM plus an RTX 3060 12GB GPU costs roughly the same as a Mac mini and runs local LLMs (Qwen 3.5, Gemma 4 via MoE)…
Talk in English at D105 (capacity 300) on
Get Shit Done with Get Shit Done - Spec-Driven development in AI timesKrzysztof Przekwas38m
Spec-Driven Development treats the spec as the executable source of truth so code becomes a byproduct; frameworks demonstrated include Open Spec, Superpower…
Talk in English at E104 (capacity 72) on
The Triangle of Compromises: Podman Quadlets Solve the Industrial Edge DilemmaAndrii Melashchenko26m
Industrial edge use cases for 256 MB–1 GB devices get a concrete alternative: Podman Quadlets run OCI containers as systemd services with a 16-line .container…
Talk in English at E112 (capacity 156) on
New security features in systemdZbigniew Jędrzejewski-Szmek35m
systemd shifts privilege work from setuid binaries into services and API calls: a clean-environment executor spawns privileged actions, empower adds…
Talk in English at E112 (capacity 156) on
Secret Tips to Reduce Insider Threat Risks with Behavioral AIKarel Kotoun15m
Claims that 71% of data loss is insider-driven and that behavior is predictable lead to a concrete stack: ingest Teams, Edge/Chrome, DLP telemetry and…
Talk in English at A113 (capacity 64) on
Improving OpenShift Cluster Updates, Step by StepDavid Hurta14m
OpenShift cluster updates get concrete fixes: implement pre-update checks that run static and dynamic probes, add Prometheus queries to flag affected clusters…
Talk in English at A113 (capacity 64) on
Hardening the Source: Scaling Branch Governance with Gemara, ComplyTime and AmpelMarcus Burghardt36m
Demo shows an automated governance pipeline that turns branch protection into verifiable evidence: it uses Gemara + ComplyTime + Ampel to define policy as…
Talk in English at E112 (capacity 156) on
Hardening OS Distribution: Verifiable and sealed OS with bootc and composefsColin Walters35m
Describes an OCI-native OS path that consumes OCI images, materializes them with composefs, and boots a signed unified kernel image (UKI) verified by fsverity…
Talk in English at E112 (capacity 156) on
Applying CI/CD Patterns to Image Mode for RHELAlessandro Rossi27m
Shows a concrete CI/CD pipeline: a Containerfile builds an OCI image, a Tekton pipeline (examples for GitHub Actions, GitLab CI, Jenkins, Ansible exist) signs…
Talk in English at E112 (capacity 156) on
Lost in Transliteration: Why strlen("Dvořák") Returns 8Avinal Kumar31m
An exploration of character encodings and how C libraries handle them, starting with ASCII, code pages, and Unicode and then focusing on the C runtime…
Talk in Czech at E104 (capacity 72) on
Beyond the Screen: A Deep Dive into Linux Accessibility for DevelopersVojtech Polasek35m
Demo code and slides live on GitHub and show concrete fixes: GTK4 emits focus events into the AT-SPI tree, Orca reads them via the Orca screen reader, and…
Talk in English at E104 (capacity 72) on
Running events, calendars, and keeping on top of CfPs for your communityAndrew Burden07m
KubeVirt runs a simple, repeatable system: track CFPs 4–8 months ahead, order swag ~5 months before events, and surface deadlines weekly via an events wiki and…
Talk in English at A113 (capacity 64) on
From Sysadmin to Software Engineer: A Non-Traditional Path into TechKaja Prokopova18m
A sysadmin background translates into concrete habits: debugging discipline, failure-mode thinking, and respect for blast radius, all carried into an associate…
Talk in English at A113 (capacity 64) on
The VEP Police: An AI Agent That Governs an Open-Source ProjectItamar Holder33m
An AI agent built with LangGraph state machine, MCP and LLMs autonomously discovers and monitors enhancement proposals across GitHub, performing parallel…
Talk in English at D105 (capacity 300) on
From CLI to MCP in 20 MinutesYaacov Zamir31m
Shows an automated path from the kubectl-mtv CLI to a working MCP server that reuses CLI parsing and help text to expose commands like get-plans over MCP, uses…
Talk in English at E104 (capacity 72) on
Integrating PQC in OpenSSL & Firefox via Loadable Modules for Cryptographic AgilityAkif Mehmood, Nouman Ali khan29m
Demonstrates a working integration: OpenSSL 3.2 + nginx uses a Rust Aurora provider that loads external PQC libs, and a Firefox fork uses a Rust qryptotoken…
Talk in English at E112 (capacity 156) on
Confidential Containers: The Next Era of Cloud Data SecurityAndrea Bozzoni, Roberto Carratala29m
Confidential Containers put hardware TEEs into Kubernetes so encrypted memory and attested execution can protect data in use; implementation uses Coco, Trusty…
Talk in English at E112 (capacity 156) on
bpftrace: Evolution from a command line tool to a complex observability frameworkViktor Malík35m
bpftrace moves from quick one‑liners to a maintainable observability language: started in 2016, GitHub shows only ~10k unique lines so far but new features…
Talk in English at E112 (capacity 156) on
Skiff - OCI image analysis utilityDan Čermák15m
skiff is a read only OCI image introspection CLI built on podman's container libraries that reports compressed vs uncompressed layer sizes and pinpoints the…
Talk in English at A113 (capacity 64) on
The TPM as an everyday security toolRoberto Hueso Gomez28m
Your machine already has TPM 2.0 (required on most PCs since ~2016) and the talk shows concrete uses: reading 23 PCR registers and locking keys to PCR 1 and 7…
Talk in English at D105 (capacity 300) on
Building Ultra-Low Latency APIs in Node.js with Native Modules and Worker ThreadsDeepesh Nair, Prathamesh Shirsat27m
Claims measurable wins: replacing hot JS paths with N-API (Rust/C++) yields image blur 20 ms → 0.4 ms and whole-request latency from ~32 ms (blocked) or ~27 ms…
Talk in English at E104 (capacity 72) on
End of OpenID in FedoraMichal Konecny24m
Fedora removes legacy OpenID and migrates services to OIDC, with OpenID decommissioned early May 2026. A separate Epsilon separated instance is built for…
Talk in English at E104 (capacity 72) on
From SBOM to Dependency Stacks: Making Software Structure VisiblePeter Štefunko15m
Prototype converts an SPDX 2.3 SBOM into a readable dependency layout by removing cycles, topologically sorting packages, mapping each package to upstream…
Talk in English at A113 (capacity 64) on
It Was Fast on My Machine: What Production Reveals About Web PerformanceAditya Patil, Gargi Kshirsagar25m
Local-fast 120 ms responses routinely jump to 2.8 s in production because of real-world layers: DNS, CDN misconfiguration, differing cache hit vs cache miss…
Talk in English at E104 (capacity 72) on
Engineering Intent: Solving the AI Productivity ParadoxJiří Daněk, Andriana Theodorakopoulou36m
Cites an IDC stat that 70% remediated up to 40% of generated code, argues fast codegen creates “AI slop” that passes CI but breaks architecture, and recommends…
Talk in English at E104 (capacity 72) on
Usable Cybersecurity Assessment ToolsMariia Bakhtina, Yasemin Acar12m
EU-funded CCAT moves four academic tools toward production-grade usability so organisations can make verifiable compliance decisions. TLS-Scanner probes…
Talk in English at A113 (capacity 64) on
Kernel Module Management in Software-Defined StorageKonstantin Maksimov14m
Kernel storage extensions for GPFS consist of three kernel modules that implement VFS hooks, inode and file operations, architecture-specific code and POSIX…
Talk in English at A113 (capacity 64) on
Escape Hyper‑V: The New Agentless Path to OpenShift VirtualizationElad Hazan15m
Migration Toolkit for Virtualization (Forklift) now connects to Hyper‑V without agents, using WinRM remoting for live VM discovery and SMB or iSCSI disk…
Talk in English at A113 (capacity 64) on
Beyond Pull Requests: How Non-Code Contributions Scale Open Source ProjectsSakshi Nasha, Kris Freedain16m
Non-code work measurably scales open source: community advocacy, docs edits, RFC comments and GitHub issue triage move projects forward and onboard…
Talk in English at A113 (capacity 64) on
Beyond the Linear Trap: Scaling CI/CD with State-on-DemandVáclav Brož35m
Modern CI/CD stalls on linear end-to-end flows; claim: treating state as first-class lets tests start from any business state, cutting validation from ~100…
Talk in English at D0206 (capacity 154) on
For a Czech Open Source AllianceAlix Guillard35m
Argues for forming a Czech open source business alliance to fill the current gap: no market study exists, a previous association dissolved in 2020, and…
Talk in English at E105 (capacity 70) on
Bridging Embedded Linux and Real-Time Control with Arduino QŠtěpán Bechynský30m
Board combines Debian Linux + STM32 real-time MCU on a single Arduino UNO Q (Qualcomm A‑series application core plus STM32), requires USB Power Delivery, and…
Talk in English at E105 (capacity 70) on
How to Survive Your First Tech Talk (Without a Kernel Panic)Mario Fernandez29m
First conference talk goes wrong: rushed prep, overcomplicated slides, and a kernel panic about two minutes in while roughly 80 engineers watch; a co‑presenter…
Talk in English at E105 (capacity 70) on
vRouter-Operator: Bringing GitOps and IaC to Virtual Network Functions in K8sDate (Yu-Chiang) Huang33m
vRouter-Operator converts router configurations into Kubernetes CRDs and uses a controller to render templates and push updates into VMs via QEMU guest agent…
Talk in English at D0207 (capacity 90) on
Observability’s Sixth Sense: Detecting Anomalies in MetricsDima Kozlov35m
Claims that a single unified anomaly score per timeseries replaces hundreds of brittle thresholds, produced by models that run as online (incremental) or…
Talk in English at D105 (capacity 300) on
The Climb Nobody Expected: From ~#1200 to #3 rank on the PyTorch LeaderboardSaiesh Prabhu24m
A corporate engineering team details how it climbs roughly 1,197 leaderboard spots to reach the top three by changing how work happens not what is built…
Talk in English at E105 (capacity 70) on
User Defined Networks: Gateway to Secure Multi-TenancyJatan Malde, Ami Desai32m
Replaces flat pod networking with namespace-scoped or cluster-scoped UDNs that give per-namespace CIDRs, support PUDN vs SUDN (primary for pod IPs, secondary…
Talk in English at D0207 (capacity 90) on
High Density VMs in OpenShift: The Journey to Native K8s SwapIgor Bezukh, Itamar Holder31m
Describes a concrete engineering path: a custom wasp-agent that uses OCI runtime hooks and the OCI update binary to work around systemd cgroup reconciliation…
Talk in English at D0207 (capacity 90) on
Discovering the Magic Behind OpenTelemetry InstrumentationJose Gomez-Selles31m
Argues manual instrumentation gives richer, actionable context than relying solely on automatic hooks (monkey patching, JVM bytecode, eBPF), and demonstrates a…
Talk in English at D0206 (capacity 154) on
Fine-tuning a small model for style/vibe (a Kimi distillation and beyond)Misha Ramendik33m
Claims that a Kimi K2 vibe can be transplanted to a much smaller model are tested by distilling from a 1T-parameter Kimi into a 1.5B Granite 4 Hybrid using…
Talk in English at D105 (capacity 300) on
How Community Collaboration is Fixing the Future of Tech RegulationRoman Zhukov34m
By 2026 the EU Cyber Resilience Act (CRA) explicitly excludes most free and open source software, after community intervention that adds the new role of open…
Talk in English at D0206 (capacity 154) on
Is Your AI Truly Open ? Cutting Through the HypeAlfonso Cancellara37m
Defines concrete checks: use the Foundation Model Transparency Index, Linux Foundation Openness Framework and SPDX 3.x model sheet to verify released artifacts…
Talk in English at D105 (capacity 300) on
Who Dropped the Packet? Solving K8s Network Mysteries in Real-TimeNeeraj Bhatt, Ramesh Sahoo35m
Operator loads eBPF programs in the kernel, exports metadata (not full packets) to a flow pipeline that enriches records with Kubernetes API info and stores in…
Talk in English at D0207 (capacity 90) on
Event-Driven Infrastructure: Autonomous Operation and Unreliable NetworksBo Maryniuk32m
Claims an event-driven, state-machine architecture that lets agents make deterministic local decisions when controllers or networks fail; demonstrates a…
Talk in English at D0206 (capacity 154) on
What can A2A do for you and what can you do for itJiri Podivin34m
A2A is an open, Protobuf-defined protocol now stewarded by the Linux Foundation that standardizes agent metadata, messages, agent card discovery and task vs…
Talk in English at D105 (capacity 300) on
No Status Updates, How I Do One-on-OnesSerhii Turivnyi36m
Manager role focuses on people not tasks, removing blockers and building trust; one-on-ones stop being status reports and become coaching sessions that surface…
Talk in English at E105 (capacity 70) on
How Parenting a Neurodivergent Teen Refactored My Management StyleChristian Trautman28m
Parenting an 11-year-old with Autism and ADHD forces concrete fixes that map to engineering management: explicit checklist and whiteboard chart routines for…
Talk in English at E105 (capacity 70) on
To InfiniBand and Beyond!Orel Misan, Sebastian Scheinkman32m
KubeVirt VMs receive real InfiniBand VFs via the SR-IOV Network Operator plus the IB SR-IOV CNI, using VFIO on Mellanox ConnectX‑7 cards at up to 400 Gbps in…
Talk in English at D0207 (capacity 90) on
Architecting High-Density Performance Testing with Container-Based SimulationShubham Bansal30m
Replaces 4,000–5,000 VM-style agents with thousands of lightweight containers orchestrated on a few powerful hosts using Podman on a few powerful hosts and…
Talk in English at D0206 (capacity 154) on
No Unsigned Models in My Cluster: Bringing Container Trust to AI Models on K8sAyushi Tiwari, Rahul Sharma27m
Practical pipeline that signs model directories with OMS v1.0 (keyless cosign-style flow, Rekor transparency logging), injects a verification init container…
Talk in English at D0206 (capacity 154) on
The Open Source Path: from Consumer, to Critic, to Contributor, to now MaintainerThameez Bodhanya16m
Outlines a concrete path from Consumer to Critic to Contributor to Maintainer with examples: Ubuntu’s “toaster” installer, running Kubernetes on bare metal…
Talk in English at A113 (capacity 64) on
Fantasy Game ConsolesLukáš Zapletal15m
Compares three concrete fantasy consoles and what you can actually build: Lua-based Pico-8 ($15, free with export limits) forces an 8K token limit and exports…
Talk in English at A113 (capacity 64) on
Symmetric Memory in PyTorch: 10x Faster GPU Communication for AIRohit Singh Rathaur34m
Introduces symmetric memory for zero‑copy RDMA in PyTorch using NCCL 2.29 one‑sided APIs, claiming up to 10× lower latency for GPU communication and sub‑10 ms…
Talk in English at D105 (capacity 300) on
Identical Testing Environments from Laptop to CI with tmt and Testing FarmPetr Šplíchal, Cristian Le25m
Describes concrete changes that make CI runs reproducible: tmt now drives Testing Farm scheduler so a local `tmt run` can provision the same Artemis VM, apply…
Talk in English at E112 (capacity 156) on
Most impacting features of newest JDKsjiri vanek39m
Surveyed developers most want practical changes: demos show calling C, Go and Rust via the foreign function & memory API (stable since JDK22) with native…
Talk in English at E104 (capacity 72) on
LLM Agents Gone Wild (And How to Tame Them with Quarkus)Martin Stefanko39m
Demo shows a Quarkus + LangChain4j agentic module that generates implementations at build time and wires agents by annotations, uses an agentic scope map keyed…
Talk in English at E104 (capacity 72) on
systemd-sysext in Prod: What We Learned Extending /usr Without a Package ManagerBrian Exelbierd, Daniel Zaťovič33m
Flatcar has been shipping systemd-sysext overlays since 2022 and runs Kubernetes, Docker and containerd as overlay images so the cluster can upgrade Kubernetes…
Talk in English at E112 (capacity 156) on
Transforming SBOMs from Compliance Burden to Security AssetMartin Sikora, Ales Raszka35m
Automated SBOM lifecycle is integrated into Konflux, producing machine-readable SPDX JSON for every production build and signing artifacts with cosign; a…
Talk in English at E104 (capacity 72) on
Survival of the Safest: Automating LLM Defense with Genetic ProgrammingPetr Kaška29m
An open-source framework performs systematic red‑teaming against 25 prompt‑based attack types and uses genetic programming to evolve small editing rules that…
Talk in English at E104 (capacity 72) on
PKI problem: who we actually trustAndrey Bondarenko16m
Default device CA bundles contain ~150 root certificates so any one of many governments, post offices or telcos can sign for any domain. Real incidents include…
Talk in English at A113 (capacity 64) on
RHDH Local as an AI Agent: Zero-Config Platform SetupFortune Ndlovu28m
RHDH Local gains an agentic onboarding CLI that scans Git repos, detects tech stacks, maps dependencies to catalog entries, recommends plugins and writes…
Talk in English at D105 (capacity 300) on
Why Your Community Needs a Python Pizza EventAnežka Müller10m
Micro-conference format that runs 10-minute talks back-to-back with 2-minute switches in a single-track, one-day schedule, designed to hit ~100 attendees with…
Talk in English at A113 (capacity 64) on
Validating Freedom From Interference in Automotive LinuxPablo Ridolfi35m
Red Hat’s Rivus testbed runs Syzkaller in QEMU entirely inside the QM Podman container (manager and executor in VM‑less mode) while running KSelftests and a…
Talk in English at E112 (capacity 156) on
From Podman to Prod: Building Trusted Container Images with Konflux on OpenShiftVladimir Sokolenko16m
Moves a simple Flask app from a local Podman build to a Konflux Tekton pipeline on OpenShift, using a Fedora 44 base, buildah/builder and hermeta to prefetch…
Talk in English at A113 (capacity 64) on
Beyond Jira: Decode that Control ChartMatěj Týč23m
Claims that Jira Control Chart cycle plots are misleading for Scrum and that cycle time is useful only after translation to Scrum specifics. Demonstrates with…
Talk in English at A113 (capacity 64) on
Years of Wear in Minutes: Benchmarking Aged FilesystemsJan Jurča15m
Benchmarks show rapid-aging methods that simulate months of filesystem wear in minutes, implemented as Filestorm, an open-source C++ tool available via COPR…
Talk in English at A113 (capacity 64) on
Dopamine, Dunning-Kruger & a Life in Tech: Confidently Wrong (& That's Okay)James Freeman14m
Begins with a McArthur Wheeler lemon‑juice anecdote and claims that the Dunning‑Kruger gap makes people most confident when they know the least and that the…
Talk in English at A113 (capacity 64) on
Quarkus in Action…​in Action!Martin Stefanko37m
Quarkus claims much faster cold starts and smaller runtimes by shifting work to build time and already shows public benchmarks; the book builds a running…
Talk in English at E104 (capacity 72) on
Local package layering on bootc systems with DNF5Evan Goode26m
Design for a DNF5 plugin called dnf rebuild that stores desired packages in packages.input.yaml, runs pre/post hook Containerfiles, builds a derived OCI image…
Talk in English at E112 (capacity 156) on
Foreman, Katello and bootable containersAdam Růžička35m
Foreman and Katello act as the stable content and provisioning hub for bootable container workflows: images (CentOS Stream 10 example) contain 480 RPMs, 66…
Talk in English at E112 (capacity 156) on
I compiled WebAssembly to lambda calculus! A journey into esoteric programmingMark Lagodych14m
A compiler converts WebAssembly 1.0 (floats excluded) into pure lambda calculus, emitting lambda programs that interpret WASM semantics including numbers as…
Talk in English at A113 (capacity 64) on
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.