4 txlf2026 2026-11-06 2026-11-07 2 00:05 https://pretalx.com https://pretalx.com/media/txlf2026/img/TXLF-transparent_vo3FkXn.png US/Central Atrium Expo Hall 2026-11-06T10:00:00-06:00 10:00 07:00 Come meet our sponsors and our community of open source projects in the exhibit hall. txlf2026-104612-expo-hall en false https://pretalx.com/txlf2026/talk/7FGJTN/ https://pretalx.com/txlf2026/talk/7FGJTN/feedback/ Big Tex Lightning Talks 2026-11-06T16:45:00-06:00 16:45 00:30 Lightning talks are short, focused presentations designed to share an idea, project, or perspective in five minutes or less. No prior signup is required, just show up to this session and get in line. txlf2026-104615-lightning-talks en false https://pretalx.com/txlf2026/talk/GKNSHM/ https://pretalx.com/txlf2026/talk/GKNSHM/feedback/ Balcones Workshop 2026-11-06T09:00:00-06:00 09:00 02:50 This hands-on tutorial will take you from Vi beginner or intermediate user to a power user, able to leap tall text buildings with a single bound. Unleash the full power of that blinking cursor, become more efficient at your job, and fame and fortune will surely follow. Most people use about 10% of what Vi is fully capable of. Awaken to your text editor’s full potential. Who Should Attend This Presentation? Anybody that works on UNIX and Linux systems would benefit from fully learning Vi, an ubiquitous and standard editor. Topics include: - History and design principles - The command language syntax - All movement commands - Many text alteration commands - Working with registers (“clipboards”) - Working with multiple files - Undoing and redoing changes - An introduction to the Vim editor This is a hands-on workshop. Attendees should bring a laptop to follow along with the in-class exercises. Please download a copy of “Alice’s Adventures in Wonderland” from Project Gutenberg, which we will use to practice Vi motion, editing, and text transformation commands. https://www.gutenberg.org/cache/epub/11/pg11.txt txlf2026-103345-power-editing-with-vi-basics Linux Aleksey Tsalolikhin en false https://pretalx.com/txlf2026/talk/CEYSED/ https://pretalx.com/txlf2026/talk/CEYSED/feedback/ Balcones Workshop 2026-11-06T13:30:00-06:00 13:30 01:20 Ready to stop hoping your security holds and start verifying it does? Forget security theory—join our practical application lab. Attendees will walk away with ready-to-use configuration files and deployment examples for Security-Enhanced Linux (SELinux), Keylime, and USBGuard, enabling you to construct a truly robust, attestable, and compliance-ready Red Hat Enterprise Linux (RHEL) environment. Learn this advanced security skill set, including: **SELinux:** Move beyond basic permissive and enforcing modes to architect and troubleshoot custom Mandatory Access Control (MAC) policies, significantly limiting the damage from exploited processes. This will help you detect compromised services and privilege escalation. **Keylime:** Remote attestation and integrity monitoring protects against unauthorized modifications to the bootloader, kernel, or critical system files. It provides cryptographic, continuous proof that a host has not been tampered with—ideal for detecting sophisticated, hardware-level rootkits and ensuring compliance. **USBGuard:** USB device authorization policy stops malicious USB devices (like "Rubber Ducky" or “BadUSB” keystroke injectors or firmware attacks) from executing automatically and blocks unauthorized removable media, preventing both data exfiltration and malware introduction via physical ports. Bring a laptop to follow along in a web based lab or use the git to run it on your own machine or homelab. txlf2026-97819-rhelevant-security-practices-selinux-keylime-usbguard-hardening-lab Security Joshua Loscar (aka Jscar) en false https://pretalx.com/txlf2026/talk/Q7CPTX/ https://pretalx.com/txlf2026/talk/Q7CPTX/feedback/ Balcones Workshop 2026-11-06T15:00:00-06:00 15:00 01:20 SQL is a powerful language for working with relational databases such as MySQL, PostgreSQL, SQL Server, and Oracle. This is an 80-minute introduction to writing SQL database queries. Please load a copy of DBeaver Community Edition (free, open-source) from https://dbeaver.io/download/ on your Mac, Windows, or Linux laptop to work along with the presentation. We will use the sample database that is included with DBeaver. We will start with simple SELECT statements to retrieve data, INSERT to add data, use UPDATE to modify it, and DELETE to remove it. We will then move on to using WHERE to narrow your database searches, grouping & ordering for readability, and using built-in functions. This is a great way to learn how to use a relational database. txlf2026-97860-structured-query-language-101 Databases Dave Stokes en false https://pretalx.com/txlf2026/talk/DQ3XW7/ https://pretalx.com/txlf2026/talk/DQ3XW7/feedback/ Stadium Workshop 2026-11-06T09:00:00-06:00 09:00 01:20 Are you new to local AI workflows? Looking to better understand LLMs and Retrieval-Augmented Generation (RAG)? Don’t want to spend a fortune on a hosted solution? In this tutorial, we will learn how to use Lima to create a repeatable local Linux environment geared towards working with local LLMs and RAG. We will use Ollama to run a local language model, add a small set of local documents, generate embeddings, store them in a local vector database, and then use retrieval-augmented generation to answer questions from that document set. It is designed for beginners who want to get hands-on with local LLM development. Attendees will leave with a configured lab, reusable examples, and a clearer understanding of how to build, test, and expand local RAG workflows. txlf2026-101389-build-a-local-rag-lab-with-lima-and-ollama Artificial Intelligence mrjandro en false https://pretalx.com/txlf2026/talk/MVXENR/ https://pretalx.com/txlf2026/talk/MVXENR/feedback/ Stadium Workshop 2026-11-06T10:30:00-06:00 10:30 01:20 Running a model locally is easy. Turning that model into multiple task-specific services — each with its own security boundary, system prompt, and restart policy — is where most guides stop. This workshop picks up where `ramalama run` leaves off. RamaLama gives you container-native local AI inference with one command: pull a model from an OCI registry, detect your GPU, serve it. Podman quadlets turn that into a systemd-managed service that starts on boot, restarts on failure, and logs to journald, but one model can do many jobs! By placing lightweight agent containers in front of a single RamaLama model server you will build task-specific AI services that behave like any other daemon on your Linux box. No Kubernetes required. Prerequisites: Laptop with a browser. Familiarity with Linux command line helpful but not required — we explain every unit file directive. txlf2026-102510-ramalama-podman-quadlets-build-task-specific-ai-services-with-systemd Artificial Intelligence David Duncan en false https://pretalx.com/txlf2026/talk/88EDHE/ https://pretalx.com/txlf2026/talk/88EDHE/feedback/ Stadium Workshop 2026-11-06T13:30:00-06:00 13:30 02:50 Image based Linux OSes have become increasingly popular. From GnomeOS to Bazzite there are a lot of ways to build and maintain your images. This workshop will walk you through how to improve a standard transactional deployment to be hardened, sealed, and verifiable including hardware-rooted attesation by utilizing Unified Kernel Images (UKI), bootc, and composefs. We will walk through the entire lifecycle of an OCI-native operating system: 1. Defining the image: Customizing a container image tailored for your needs. 2. Leveraging Composefs: Understanding how bootc ingests OCI containers and materializes them into a secure, verifiable, read-only filesystem. 3. Sealing the OS: Bundling your configuration into a Unified Kernel Image (UKI) to establish strict lifecycle and update guarantees rooted in hardware security. 4. Booting and Verification: Deploying your newly minted image and validating its tamper-resistance and deployment integrity. By the end of this lab, you will walk away with a fully functional, self-generated sealed image template and the practical confidence to implement container-like immutability across your infrastructure. Prerequisites for Attendees: - Familiarity with Podman and basic Linux command-line navigation. - A Linux machine capable of running a virtual machine (KVM/QEMU, qemu-img, virtiofsd, openssh-clients) txlf2026-102121-confidential-computing-verifiable-and-sealed-os-images Cloud Native Sean Thrailkill en false https://pretalx.com/txlf2026/talk/KBRFMJ/ https://pretalx.com/txlf2026/talk/KBRFMJ/feedback/ Bevo Workshop 2026-11-06T09:00:00-06:00 09:00 02:50 This workshop provides a practical introduction to Software-Defined Radio (SDR) using the USRP radio and the open-source software tool-chain (C++, Python, UHD, and GNU Radio), running on Linux. We will discuss the hardware and the architecture of the USRP radio, and how to get starting using it, installing the development environment, and programming the USRP from C++ and Python, using the UHD API, and using the GNU Radio framework. Several real-time demos and examples of wireless systems running on the USRP radio will be shown. Attendees should gain a basic foundation and practical understanding of how to configure, program, and operate the USRP to implement various wireless systems and applications. txlf2026-101959-introduction-to-software-defined-radio-sdr-with-the-usrp-radio Other Neel Pandeya en false https://pretalx.com/txlf2026/talk/XG9E9J/ https://pretalx.com/txlf2026/talk/XG9E9J/feedback/ Bevo Workshop 2026-11-06T13:30:00-06:00 13:30 01:20 Cedar is an open-source authorization policy language and CNCF Sandbox project for expressing fine-grained permissions in policies that are human-readable and designed for automated analysis. In this hands-on workshop, attendees will use Cedar to authorize privileged Linux operations invoked through sudo. Using a local VM and the Tarp browser client, participants will test and modify policies controlling who may view logs, check service status, or restart a systemd service. The demo also shows context-aware rules—such as allowing local operations while denying SSH requests—and how root-owned fixed command mappings prevent argument injection, arbitrary commands, and unauthorized shell access. txlf2026-103382-build-a-safer-sudo-with-cedar Other Michael Schwartz en false https://pretalx.com/txlf2026/talk/EF3G9A/ https://pretalx.com/txlf2026/talk/EF3G9A/feedback/ Bevo Workshop 2026-11-06T15:00:00-06:00 15:00 01:20 When yet another Linux kernel vulnerability lands on the front page, what are your options? If you are lucky, it was patched months ago and is already in your distro kernel or latest stable LTS release. Worst-case, the patch doesn't exist yet and you're going to be exposed until one exists and you can reboot your fleet. What if I told you... there is another way to mitigate these vulnerabilites and you don't even need to reboot your machine! To do this at Cloudflare we reached for BPF-LSM. In this hands-on workshop, you'll build your very own defensive BPF LSM program that can be used to disarm some of these attacks. Starting from a "hello world" LSM program, you'll write and load your own BPF-LSM code on your laptop and watch it deny an operation in real time: blocking a syscall path the same way a live vulnerability mitigation would. We'll cover how the Linux Security Module framework and BPF fit together, how LSM hooks differ from raw syscalls, how to return "permission denied" from your own policy, and how to verify it actually works. Prerequisites: A laptop capable of running the provided VM image (setup instructions TBD). Comfort with C and the Linux command line is helpful. txlf2026-103495-mitigating-all-the-vulns-with-bpf-lsm-a-tutorial Security Chris Arges en false https://pretalx.com/txlf2026/talk/J3DRGC/ https://pretalx.com/txlf2026/talk/J3DRGC/feedback/ Atrium Expo Hall 2026-11-07T10:00:00-06:00 10:00 07:00 Come meet our sponsors and our community of open source projects in the exhibit hall. txlf2026-104611-expo-hall en false https://pretalx.com/txlf2026/talk/NZSHTY/ https://pretalx.com/txlf2026/talk/NZSHTY/feedback/ Big Tex Keynote 2026-11-07T09:00:00-06:00 09:00 00:50 TBD txlf2026-104613-keynote en false https://pretalx.com/txlf2026/talk/MZY8DZ/ https://pretalx.com/txlf2026/talk/MZY8DZ/feedback/ Big Tex Talk 2026-11-07T10:00:00-06:00 10:00 00:50 Linux has been one of the most active open source project since it's creation in 1991. Having 9 changes an hour, for the last 10 years or so, presents an interesting challenge for traditional safety analysis methodologies. The [ELISA]([url](https://elisa.tech/)) (Enabling Linux in Safety Applications) project has been exploring approaches to help with the analysis of linux being used in safety critical systems. This session will go through some of the approaches being investigated to date and welcomes discussion on other approaches to consider. txlf2026-99970-enabling-safe-usage-of-linux-in-systems-progress-to-date Linux Kate Stewart en false https://pretalx.com/txlf2026/talk/HU8TDL/ https://pretalx.com/txlf2026/talk/HU8TDL/feedback/ Big Tex Talk 2026-11-07T11:00:00-06:00 11:00 00:50 It's been a big year for RISC-V and Linux: we have new fast hardware, drivers, and bugs! This talk is meant to give an overview of the state of the ecosystem and community. I want to go over big releases of this year (RVA23 hardware, Fedora 44 images and the Omni Kernel) and how you can learn more, engage with the ecosystem, and contribute back. While there is a lot going on in the RISC-V universe, this will focus on Linux capable hardware and that broader ecosystem. There is a lot to share and I also want to learn what the audience thinks are high priority issues that we can tackle together. The beginning of this talk will be brief, and the remainder will be much more interactive with audience polls and questions. txlf2026-101451-linux-and-risc-v-2026-edition Linux edolnx en false https://pretalx.com/txlf2026/talk/E97QXC/ https://pretalx.com/txlf2026/talk/E97QXC/feedback/ Big Tex Talk 2026-11-07T14:00:00-06:00 14:00 00:50 IT is full of late nights, urgent tickets, and systems that always seem to fail right before a holiday weekend. Over time, that pace catches up with you. I learned that lesson the hard way. In this session, I’ll share what burnout really looks like from someone who’s lived through it and how I found my way back to a healthier balance. We’ll talk about the patterns that lead us there, from over-commitment and constant context switching to the pressure to “always be learning.” Then we’ll look at practical ways to protect your energy while still loving your work: setting better boundaries, automating what drains you, and finding community outside the console. You’ll leave with real strategies you can use to make IT a sustainable career instead of a sprint. Because the best admins, developers, and engineers aren’t the ones who work nonstop. They’re the ones who learn how to rest, recharge, and still show up ready for what’s next. Learning Outcomes: - Recognize early warning signs of burnout in IT and DevOps roles. - Identify habits and boundaries that support long-term career health. - Apply small, practical changes to rebuild focus and motivation. txlf2026-96692-from-bash-to-burnout-staying-sane-in-a-24-7-tech-world Other Eric Hendricks en false https://pretalx.com/txlf2026/talk/SUKNZ9/ https://pretalx.com/txlf2026/talk/SUKNZ9/feedback/ Big Tex Talk 2026-11-07T15:00:00-06:00 15:00 00:50 The Southern California Railroad Museum is one of the biggest transportation museums in the United States. One major feature of this museum is its Signal Garden. This allows guest to press a button and see signals work. Behind the scene is a Linux Raspberry PI controller. This talk discusses how the 100 year old railroad signals are interfaced with the PI and the unique challenges that had to be overcome. For example, the "Wig-Wag" signals use two huge electromagnets which are continually turned on and off. Any time you turn off a big coil, you get a huge EMP which will fry unprotected machinery. Additionally there is the problem with dealing with systems that are over 100 years old for which the documentation was lost sometime in the 1930's. The signal garden presents some unique challenges. Such as the time I started to upgrade the computer, opened the cabinet, saw the black widow making a nest on the main logic board, slammed the door shut and got some spray to debug the computer. txlf2026-100468-experience-using-linux-to-control-the-southern-california-railroad-museum-s-signal-garden Other Steve Oualline en false https://pretalx.com/txlf2026/talk/U9C3Q8/ https://pretalx.com/txlf2026/talk/U9C3Q8/feedback/ Big Tex Talk 2026-11-07T16:00:00-06:00 16:00 00:50 At IBM, we have over 9,000 Open Source Engineers who work in over 900 open source organizations. As the people who review all of the open source contributions for compliance, we have legal oversight and must document everything. Until recently, we have required engineers to fill out word processing templates and then have our compliance team review each document one line by line to look at every release we do. This is painful. In this talk, we introduce the new pipeline for IBM open source compliance review that uses docassemble.org, an open source tool to guide engineers to answer questions to generate the needed documents, and Granite, an local open source AI model, to help compile, categorize, and streamline the feedback loop for our compliance team. This allows them to focus on the content and legal questions instead of the process. We’ll tell the story of how we got here, what we’ve learned, and even have some simple code to help you get this off the ground for your processes. Leverage docassemble, and you'll be shocked on how quickly you'll find real ROI and freedom to tackle real problems versus managing paperwork. txlf2026-101114-streamlining-enterprise-scale-open-source-governance-with-docassemble-granite Other JJ Asghar en false https://pretalx.com/txlf2026/talk/YK79SJ/ https://pretalx.com/txlf2026/talk/YK79SJ/feedback/ Big Tex Talk 2026-11-07T17:00:00-06:00 17:00 00:50 In a corner of the industry where an institution of government contractors and political pressure rules, what happens when the wrong person builds the right thing, an unsanctioned structure with unsanctioned resources, when institution isn't watching? In this session, you'll hear about a bored Navy wife butting up against a wartime IT support strategy that involved two Excel spreadsheets and a phone. You'll hear about what she discovered while she built solutions: iterative development as natural evolution in broken systems, open source as a legitimate response to bloated enterprise solutions, frustration and constraint as motivators, and the innovation that emerges when a witch with an old Linux box and time on her hands has nothing to lose. txlf2026-99368-cauldron-ops-how-a-bored-navy-wife-out-shipped-enterprise-software Other KL Grady en false https://pretalx.com/txlf2026/talk/QU7BME/ https://pretalx.com/txlf2026/talk/QU7BME/feedback/ Big Tex Closing Remarks 2026-11-07T18:00:00-06:00 18:00 00:15 We'll wrap up the conference with a few final thoughts and highlights. Join us as we close out the event together. txlf2026-104616-closing-remarks en false https://pretalx.com/txlf2026/talk/PE8DWC/ https://pretalx.com/txlf2026/talk/PE8DWC/feedback/ Big Tex After Party 2026-11-07T19:30:00-06:00 19:30 03:00 TBD txlf2026-104617-after-party en false https://pretalx.com/txlf2026/talk/BUPFTA/ https://pretalx.com/txlf2026/talk/BUPFTA/feedback/ Lil Tex Talk 2026-11-07T10:00:00-06:00 10:00 00:50 Docker made containers usable and it has the mindshare, but it isn't the only containerization option we have! This talk is a working tour of Podman: what it is, how it fits into (and differs from) the Docker workflow, how to use it in real environments, and a couple of short war stories. We'll start from the basics: what a container is and the history of containers. Then we'll walk through some of Podman's features. From there, we'll get into practical matters: building images, running services, user-space networking, file-ownership, and a bit more! This isn't a sales pitch or a call to use Podman over Docker, it's just an overview of an alternative project from an enthusiastic user! txlf2026-102321-podman-a-practical-tour Cloud Native Hobbs en false https://pretalx.com/txlf2026/talk/Z7UKXE/ https://pretalx.com/txlf2026/talk/Z7UKXE/feedback/ Lil Tex Talk 2026-11-07T11:00:00-06:00 11:00 00:50 Running enterprise-grade Kubernetes platforms like OpenShift provides invaluable career skills, but translating that architecture into a home or departmental lab can feel like shoving a square peg into a round hole. Datacenter luxuries like hardware load balancers, enterprise storage arrays, and endless pools of RAM don't exist under your desk. Thankfully, the upstream OKD project brings that exact same cloud-native power to your local hardware—if you know how to navigate the infrastructure hurdles. This presentation is a practical survival guide for successfully deploying OKD in resource-constrained environments. We will map out the architectural prerequisites that trip up most deployments, focusing heavily on configuring wildcard DNS subdomains and strict reverse DNS (PTR) records for node discovery. We will look at strategies for handling cluster ingress using lightweight open-source load balancers like HAProxy, and contrast persistent storage methods, comparing single-node local storage layout against multi-node NFS shared exports. Attendees will leave with a clear, realistic blueprint for spinning up a functional OKD cluster without melting their hardware or their power budget. Key Takeaways: - Sizing and Footprint: Navigating the trade-offs between traditional High Availability topologies, hyper-converged virtualized clusters, and Single Node OKD (SNO). - The Networking Triad: Step-by-step logic for implementing the API load balancer, configuring wildcard application routes, and getting PTR records right. - Storage Demystified: Configuring robust persistent volume backing using standard local LVM provisions or shared NFS storage mappings on Fedora CoreOS nodes. txlf2026-102513-enterprise-kubernetes-at-home-a-homelabber-s-guide-to-okd Cloud Native Thomas Cameron en false https://pretalx.com/txlf2026/talk/FQTX9Y/ https://pretalx.com/txlf2026/talk/FQTX9Y/feedback/ Lil Tex Talk 2026-11-07T14:00:00-06:00 14:00 00:50 AI assistants can generate infrastructure code quickly, but speed alone does not make infrastructure safe, repeatable, or maintainable. The interesting problem is not how to get an AI agent to write Terraform, Compose, Ansible, or Kubernetes manifests. It is how to design infrastructure projects where an agent can make useful changes without bypassing the engineering practices that make infrastructure-as-code reliable. This talk presents a practical framework for using agentic tools across the infrastructure lifecycle: build, update, test, deploy, and destroy. Using Docker Compose-based applications as an approachable example, we'll examine how familiar IaC practices such as pinned dependencies, modular and DRY configuration, environment separation, secrets management, idempotency, automated testing, and reviewable changes become even more important when an AI agent is modifying the repository. We'll also look at the agent-facing side of the problem: repository instructions, scoped context, validation commands, tests, and explicit guardrails that give an agent enough information to work effectively without giving it unrestricted authority over production-impacting decisions. Although the examples use Docker Compose, the patterns apply broadly to Terraform/OpenTofu, Ansible, Kubernetes, Helm, Pulumi, CI/CD systems, and other infrastructure-as-code environments. The goal is not better prompt engineering. It is better infrastructure engineering: building repositories where humans, automation, and AI agents can all make changes through the same predictable, testable, and reviewable lifecycle. txlf2026-99474-agentic-infrastructure-lifecycle-management Artificial Intelligence Cody Lee Cochran en false https://pretalx.com/txlf2026/talk/SYKJDG/ https://pretalx.com/txlf2026/talk/SYKJDG/feedback/ Lil Tex Talk 2026-11-07T15:00:00-06:00 15:00 00:50 What if you could ask an AI "is my house good for solar panels?" - and get a real answer, without sending your address leaving your machine? In this talk, we'll build a fully local AI agent stack on Linux that evaluates solar potential for any Texas property - no cloud APIs, no OpenAI key, no data leaving your network. We'll wire together Ollama running a local LLM, a custom MCP (Model Context Protocol) server exposing tools for geocoding and solar estimation, the free NREL PVWatts government API, and PostgreSQL for local storage. The result: an agent you can ask "what would solar look like on my roof?" and get a real, human-readable answer grounded in real data. We'll cover: - Running a local LLM with Ollama on Linux - Building an MCP server that exposes real-world tools to any AI client - Turning raw government solar data into plain-English savings estimates - Persisting and querying results in local PostgreSQL - Running the whole stack as a systemd service - no always-on laptop required Texas gets 200+ sunny days a year and some of the most volatile electricity prices in the country thanks to ERCOT. This is a practical talk about what local, open source AI can actually do - on your hardware, on your terms, using a protocol you can reuse for your own tools tomorrow. As the industry shifts from expensive frontier models toward cheaper, optimized inference this talk shows what's possible when you run capable AI agents directly on hardware you own. txlf2026-97727-don-t-mess-with-my-data-local-ai-mcp-on-linux Artificial Intelligence Rajesh Alwar en false https://pretalx.com/txlf2026/talk/HQWWGD/ https://pretalx.com/txlf2026/talk/HQWWGD/feedback/ Lil Tex Talk 2026-11-07T16:00:00-06:00 16:00 00:50 AI agents are evolving from conversational assistants into systems that can execute shell commands, modify files, call APIs, start containers, and interact with production infrastructure. Giving an agent these capabilities can unlock powerful automation—but it can also turn a bad model decision, prompt-injection attack, or compromised tool response into a serious security incident. This session explores how Linux security primitives can be used to contain agentic workloads before they are trusted with real systems. We will build a practical threat model for an AI agent with access to the operating system and examine risks such as excessive privileges, unsafe command execution, credential exposure, unrestricted network access, resource exhaustion, and accidental modification of host files. The session will then demonstrate how to reduce these risks using rootless containers, Linux namespaces, cgroups, capability dropping, seccomp profiles, read-only filesystems, AppArmor or SELinux policies, network restrictions, ephemeral credentials, and detailed audit logging. A live demonstration will compare the same agent running in an overly permissive environment and in a hardened Linux sandbox. We will observe which dangerous operations succeed, which are blocked, and how the security controls affect the agent’s behavior. Attendees will leave with a practical architecture and security checklist for running AI agents that can safely interact with Linux systems, containers, APIs, and infrastructure. txlf2026-103342-when-ai-agents-get-shell-access-securing-agentic-workloads-on-linux Artificial Intelligence Pragya KeshapAkhilesh Keshap en false https://pretalx.com/txlf2026/talk/M8H3K9/ https://pretalx.com/txlf2026/talk/M8H3K9/feedback/ Lil Tex Talk 2026-11-07T17:00:00-06:00 17:00 00:50 Somewhere in Dune's backstory, a group of war-weary humans wrote a commandment into their bible: thou shalt not make a machine in the likeness of a human mind. It's one of the most quoted lines in the entire saga. The real reason behind it wasn't just "the machines got too smart." Humans had quietly handed those machines their judgment, and they woke up one day having forgotten how to make decisions for themselves. Sounds familiar? This talk tackles a very current, very specific fear: the fear of change, now supercharged by AI, LLMs, and coding assistants. Not the constant "tech moves really fast" fear engineers have always carried. This one is different. This one whispers that your hard-won expertise might be losing value in real time, that the field now moves faster than you can track, that you're losing the craftsmanship you spent a career building. If you've felt that in the last few years, you're not imagining it, and you're not alone. Let's use Dune as a lens to examine how a society can answer that exact fear in two different, equally tempting, equally wrong ways. Then we'll look at what actually works better. Expect an honest, occasionally funny, deeply Dune-flavored case for how technical people can hold onto their expertise and relevance. You don't have to reject the tools. You don't have to hand over your judgment to them either. No doom-posting (but yes, Dune-posting), no hype, no easy answers. Just a clearer way to think about this fear going forward. txlf2026-102418-thou-shalt-not-make-a-machine-in-the-likeness-of-a-senior-engineer Artificial Intelligence Major Hayden en false https://pretalx.com/txlf2026/talk/FM99UQ/ https://pretalx.com/txlf2026/talk/FM99UQ/feedback/ Balcones Talk 2026-11-07T10:00:00-06:00 10:00 00:50 We spend millions protecting the servers and the data. We do countless threat modeling exercises. We employ armies of red-teamers to prove that our systems are secure from various technological threats. But at the end of the day, have we lost sight of what we are really protecting? Ultimately, "what" we are protecting boils down to "who" we are protecting. If we've lost sight of the humans who ultimately are utilizing these systems and entrusting our technological ecosystems with their precious data and identities, then it's easy to become cynical as well as Trapped In The Weeds of what is technically possible versus what is sustainable and desirable as outcomes for our Security Programs. When security has come to be seen as the "Department of No," that's a good sign that we have lost our way and need to chart a better course for keeping our users, developers, and ourselves safe. This means getting past our biases, stopping blaming the user, and getting on the same page as those we need to secure. This talk is a call to action to get away from our keyboards and get into real conversations with our colleagues that can help us transition away from being bottlenecks to being the solvers for efficiency, and the team that helped make the safe way also the easiest and most sustainable way. txlf2026-100221-protect-the-humans-at-all-costs-why-technology-alone-can-t-solve-security Security Dwayne McDaniel en false https://pretalx.com/txlf2026/talk/FRFKTU/ https://pretalx.com/txlf2026/talk/FRFKTU/feedback/ Balcones Talk 2026-11-07T11:00:00-06:00 11:00 00:50 The last few years Free and Open Source software has come under attack! Large and smaller projects have been flooded with "Slop Contributions". Frustrated maintainers have had enough of AI contributions. To makes things worse cyber attacks specifically malware have skyrocketed in the Open Source world. Malware has increasingly been spotted in new areas of the FOSS supply chain. In this talk you will hear about what maintainers are doing to gain back control from the "Slop Contributions". I will also discuss the latest Linux Malware landscape and what projects are doing to protect users. txlf2026-101404-attack-of-the-ai-bots Security Eduardo Robles en false https://pretalx.com/txlf2026/talk/UXUS8R/ https://pretalx.com/txlf2026/talk/UXUS8R/feedback/ Balcones Talk 2026-11-07T14:00:00-06:00 14:00 00:50 Two very different scenarios, joined by KDE. What do you do when your favorite Enterprise Linux distribution takes out KDE? What do you do when you want a touch friendly desktop for your tablet or phone? Join Troy as he explains how he, and the Fedora KDE SIG, made KDE for Enteprise Linux better than ever, as well a Fedora Plasma Mobile an upcoming place to watch. For Enterprise Linux he'll explain what end users want and how it is being provided. For Plasma Mobile he'll talk about what state it is currently in, use cases, pain points (usually hardware related) and his world famous demo's. txlf2026-102453-kde-from-enterprise-to-mobile Distributions Troy Dawson en false https://pretalx.com/txlf2026/talk/XVYKEX/ https://pretalx.com/txlf2026/talk/XVYKEX/feedback/ Balcones Talk 2026-11-07T15:00:00-06:00 15:00 00:50 Transactional systems offer a middle ground between Fedora Atomics, and their image based updating, and read-only filesystems and Traditional Fedora, with it's read/write filesystems, as they can be configured to do either, offering great flexibility. In a read-only context, transactional Fedora offers a minimal system, and workloads are handled through podman or docker containers, for server usage. On the desktop, toolbox, distrobox, flatpaks, and potentially other application delivery mechanisms can be used, while still providing the possibility of local modifications to the read-only filesystem, for things that just make sense to be delivered via RPM. In a read-write context, Transactional systems offer Atomic updating, offering the possibility of more reliable updating, more stable systems over time, and the ability to "rollback" snapshots, in case of issues. or for troubleshooting purposes. This talk will discuss the basic ideas behind Transactional Updates, how they're used in openSUSE, and what it takes to bring them to Fedora. txlf2026-101695-transactional-systems-in-the-fedora-ecosystem Distributions Neal GompaShawn W Dunn en false https://pretalx.com/txlf2026/talk/PJLVGM/ https://pretalx.com/txlf2026/talk/PJLVGM/feedback/ Balcones Talk 2026-11-07T16:00:00-06:00 16:00 00:50 It's a question almost every Linux user, developer, or platform engineer has asked. Behind that seemingly simple question lies one of the most important, and least understood, engineering tradeoffs. Every application, container, and cloud workload depends on software that has been packaged, tested, maintained, and released by a Linux distribution. Yet most engineers never see the work required to transform upstream source code into software that is stable enough to trust in production. Every package update must balance security, compatibility, regression risk, reproducibility, dependency management, architecture support, release criteria, and long-term maintenance. What appears to be "slow" is often deliberate engineering designed to reduce risk for millions of users. In this session, we'll follow a package from its upstream release to its arrival in a Linux distribution, exposing the decisions, tradeoffs, and engineering practices that happen along the way. We'll explore why maintainers backport fixes instead of always upgrading, why regression prevention often outweighs feature velocity, and how release engineering quietly protects the software ecosystem that nearly every modern application depends on. Whether you build applications, manage infrastructure, contribute to open source, or operate production systems, you'll leave with a practical mental model for understanding how Linux software reaches your systems, and a deeper appreciation for the engineering judgment required to balance innovation, stability, and trust at scale. Linux distributions aren't slow. They're deliberate, because reliability is built one engineering decision at a time. txlf2026-104608-why-isn-t-my-distribution-shipping-the-latest-version-yet Distributions Autumn Nash en false https://pretalx.com/txlf2026/talk/UMCX99/ https://pretalx.com/txlf2026/talk/UMCX99/feedback/ Balcones Talk 2026-11-07T17:00:00-06:00 17:00 00:50 5BSD explores a different approach to operating system security. Rather than starting every process with broad ambient authority and then restricting it through layers such as namespaces, seccomp, LSMs, and cgroups, 5BSD begins with no ambient authority. Every privilege is represented by an explicit capability granted to a process, and no process can acquire authority it was not intentionally delegated. Built on FreeBSD, 5BSD extends the operating system with a capability-oriented runtime while preserving compatibility with existing software, including Linux applications through the Linux ABI layer. Traditional UNIX applications continue to operate through familiar interfaces, while capability-aware applications can take advantage of new system calls and kernel services designed around explicit authority. The system is bootstrapped by the Oracle, the privileged system initializer. Oracle launches workloads, constructs their initial capability sets, and delegates only the authority required for their intended function. After startup, workloads operate using the capabilities they possess rather than relying on ambient privileges. Services are discovered through Serviced, 5BSD’s capability-aware service registry, while kernel facilities such as Keyvault and mac_abac provide cryptographic services and attribute-based access control without exposing unnecessary privilege to applications. This talk introduces the architecture behind 5BSD, including capability bootstrapping, authority delegation, capability-aware system calls, and compatibility with existing UNIX and Linux software. We will follow a workload from process creation through capability initialization, demonstrate how capabilities are transferred and enforced, and examine how the system behaves when software attempts operations outside its delegated authority. Attendees will gain a practical understanding of capability-oriented operating system design, how 5BSD integrates capabilities into a BSD-derived kernel without abandoning compatibility, and how explicit authority provides a foundation for building systems with stronger isolation and more predictable security properties. txlf2026-98189-the-design-and-implementation-of-the-5bsd-operating-system Distributions Kory Heard en false https://pretalx.com/txlf2026/talk/XQUDW7/ https://pretalx.com/txlf2026/talk/XQUDW7/feedback/ Stadium Talk 2026-11-07T10:00:00-06:00 10:00 00:50 As a sysadmin, I spent years managing my Proxmox homelab the hard way - juggling browser tabs, editing config files by hand, running the same commands over and over. Every gap in my workflow was a small daily frustration. Then I stopped waiting for someone else to fix it and built ProxUI: a self-hosted Proxmox management UI built by a sysadmin, for sysadmins. This talk is not about being a developer. It's about recognizing that the tool you need probably won't be built by a company - because your use case is too specific, too personal, and too real. Every feature in ProxUI exists because I needed it: a single-page cluster dashboard, multi-cluster switching, automated cloud-ready qcow2 image builds, LXC device passthrough without editing config files, and more. Each one started as something I was doing by hand. I'll walk through the gap-driven development process: how frustration becomes a feature, how a sysadmin with minimal software engineering background ships a real tool and giving to the open source community. I'll also cover the infra around the project — how to build and store cloud images, how to run a live demo environment, and how to deploy ProxUI in your own homelab in minutes. If you've ever thought "someone should build this" — this talk is for you. txlf2026-101964-stop-suffering-how-i-built-proxui-to-fix-my-homelab Infrastructure Alexander Shut en false https://pretalx.com/txlf2026/talk/SGK7R7/ https://pretalx.com/txlf2026/talk/SGK7R7/feedback/ Stadium Talk 2026-11-07T11:00:00-06:00 11:00 00:50 Everyone knows 3-2-1 backups, but that "1 offsite copy" is the part that trips people up. Rather than pay for a storage subscription, why not enter a pact with a friend? You'll both grab your stuff from the cloud, store them on a big drive, and sync a copy to each other's big drives—fully encrypted and inaccessible to the public internet. This is the true story of how I walked a friend through this experiment and became their Fireproof Backup Buddy. Their most important stuff is attached to my NAS; my stuff is on a drive attached to their computer. Flood, fire, ransomware, deeply stupid technical decision: we're both covered, and neither of us pays a grudging backup bill. This talk will walk through setting up a system with: - Restic encrypting for diff-based backups - Tailscale (or just Wireguard) for reliable remote access - rclone keeping your Google stuff up to date (reality check) Will I be able to restore a real file from my friend's computer, during the talk? Maybe! Probably. Could be fun. txlf2026-100451-recruit-your-friends-into-a-backup-cult-encrypted-off-site Infrastructure Kevin Purdy en false https://pretalx.com/txlf2026/talk/U97UWX/ https://pretalx.com/txlf2026/talk/U97UWX/feedback/ Stadium Talk 2026-11-07T14:00:00-06:00 14:00 00:50 This talk serves as an introduction to the Incus project. Incus provides a powerful way to manage application containers, operating system containers and virtual machines in a single interface. While most people turn to Docker for containers or Proxmox for a more traditional hypervisor, Incus is an overlooked alternative whether it's a homelab or a production environment. txlf2026-101574-incus-containers-vms-and-more Infrastructure Nathan Thornhill en false https://pretalx.com/txlf2026/talk/B8VMKA/ https://pretalx.com/txlf2026/talk/B8VMKA/feedback/ Stadium Talk 2026-11-07T15:00:00-06:00 15:00 00:50 The cloud is just someone else's computer but you data is your own. In this talk I will discuss common misconceptions about today's Internet that leave people unknowingly giving up their right to privacy, like Third Party Doctrine and Social Media, how to protect their data at home with self hosted infrastructure using ZFS and Samba 3 and how to keep access to your things secure even when crossing borders. I work in cybersecurity and travel with my devices. I've been through countless border crossings, airport inspections and in insecure work environments where I must protect my data while maintaining minimal downtime. In this talk, I will demonstrate my self hosted infrastructure that allows me to travel in peace knowing that my network is secure, my data is intact, my privacy is maintained and my downtime is minimal. txlf2026-100545-the-right-to-rm-rf-building-a-global-self-hosted-backup-and-restore-platform Infrastructure Justin Campbell en false https://pretalx.com/txlf2026/talk/AGAAFR/ https://pretalx.com/txlf2026/talk/AGAAFR/feedback/ Stadium Talk 2026-11-07T16:00:00-06:00 16:00 00:50 In this session, we’ll dive deep into the concept of High Availability (HA) and explore how it benefits both customers and engineers/admins alike. The goal is to equip you with a clear understanding of how to plan and implement a highly available system that ensures uptime and resilience. We’ll start by distinguishing High Availability from Disaster Recovery—two concepts that often get mixed up but serve different purposes in the realm of system reliability. Next, we’ll explore various types of High Availability setups, from basic redundancies to more sophisticated failover mechanisms, and walk through the initial planning steps required to design an HA system that suits your needs. Then, we’ll shift our focus to the core of your application: data. We’ll talk about: - Best practices for storing and managing files in an HA environment. - SQL and NoSQL data: What are the differences and considerations when building for high availability? - How to make your data resilient to failures and ensure it’s always accessible. Finally, we’ll wrap up with a hypothetical example, followed by a real-world case study of an HA system I’ve architected that’s in production and actively supporting thousands of users. By the end of this session, you’ll have a solid understanding of how to design, implement, and manage a high-availability solution that keeps your services reliable and your customers happy. txlf2026-99605-effective-ha-high-availability-architecting Infrastructure Carl Thompson en false https://pretalx.com/txlf2026/talk/PRLRSR/ https://pretalx.com/txlf2026/talk/PRLRSR/feedback/ Stadium Talk 2026-11-07T17:00:00-06:00 17:00 00:50 Every sysadmin knows the five stages of outage grief, and they all end at DNS. Yet for something that runs the entire internet, a clear and approachable explanation of it is surprisingly hard to find. In this talk, Jonathan will walk through a little history of DNS, what nameservers actually do, what TTLs are, and everything from caching to record types to email. Come learn the backbone technology you blame every day, so next time you can say "it was DNS" with confidence. txlf2026-102521-it-s-not-dns-it-can-t-be-dns-it-was-dns Infrastructure Jonathan Wright en false https://pretalx.com/txlf2026/talk/QHU83G/ https://pretalx.com/txlf2026/talk/QHU83G/feedback/ Bevo Talk 2026-11-07T10:00:00-06:00 10:00 00:50 Competing with tech giants like Microsoft and Google may seem impossible for smaller companies with limited resources. However, Nextcloud has proven that an open-source business model, remote work, and a culture of open communication and innovation can provide a path to success. This presentation will explore how organizations can leverage open-source development, a distributed workforce, and agile innovation strategies to build competitive alternatives with a fraction of the budget and workforce. Attendees will learn how a transparent, community-driven approach leads to sustainable growth, fosters rapid innovation, and builds user trust. txlf2026-101952-building-a-microsoft-and-google-alternative-with-1-of-the-resources-the-power-of-open-source-remote-work-and-innovation-culture Development Frank Karlitschek en false https://pretalx.com/txlf2026/talk/XWJUHQ/ https://pretalx.com/txlf2026/talk/XWJUHQ/feedback/ Bevo Talk 2026-11-07T11:00:00-06:00 11:00 00:50 Nobody uses Perl for anything important any more... right?!? The answer to this question may surprise or even SHOCK YOU. As Monty Python says... "I'm not quite dead yet!" Learn how Perl is used to: - Predict storm surges & save lives - Implement cutting-edge AI models - Perform genetic analysis for medical breakthroughs - Keep the Internet running - And much more! txlf2026-102472-perl-is-dead-right Development William Braswell en false https://pretalx.com/txlf2026/talk/SCFZC8/ https://pretalx.com/txlf2026/talk/SCFZC8/feedback/ Bevo Talk 2026-11-07T14:00:00-06:00 14:00 00:50 A large share of High-Performance Compute (HPC) work runs on Linux: the schedulers, bare metal nodes, job queues, and MPI processes. The operating model is usually reactive. An engineer submits a job, waits hours or days, and reads a log file after the damage is done. By then the compute is spent, the data is stale, and the decisions that depended on it are late. This talk is a real enterprise case study in flipping that model from reactive to proactive. We built a Linux-native automation layer that connects a chat-based control plane to underlying bare metal HPC infrastructure, serving several thousand internal engineering users running millions of simulations per year. Shell scripts and system-level integration drive event-driven pipelines and feed data into a control plane that gives engineers timely, actionable signals before compute time is wasted. We combined pre-submit validation, to catch the failures that never needed to run, with runtime inspection of solver state to identify stalled or diverging jobs while they are still cheap to kill. Near real-time alerts moved job status out of log files and into chat where users already work. The result is hundreds of thousands of compute hours saved per year, tens of thousands of engineer hours reclaimed, and millions of dollars in cost savings. This talk shows that proactive HPC operations at scale are an automation problem you can already solve on the Linux stack you have. txlf2026-101775-from-log-files-to-live-signals-rethinking-hpc-operations-on-linux-bare-metal Automation Borislav Sabotinov en false https://pretalx.com/txlf2026/talk/D9AYRJ/ https://pretalx.com/txlf2026/talk/D9AYRJ/feedback/ Bevo Talk 2026-11-07T15:00:00-06:00 15:00 00:50 You’ve written some Ansible playbooks — but how do you really know they work before you push them to production? In this session, we’ll explore how to test and validate your Ansible automation safely and repeatably using Test Kitchen with lightweight virtualization tools like Docker, Podman, Vagrant, or QEMU. We’ll cover how to set up a local “kitchen” for experimenting with playbooks, write simple verification checks, and integrate tests into your CI pipeline. You’ll learn how to: Build reproducible test environments for your Ansible code Validate roles and tasks automatically before deployment Catch errors early and increase confidence in your automation Whether you’re new to Ansible or managing complex infrastructure, this talk will show you how to cook up reliable playbooks without burning down your servers. txlf2026-101697-testing-ansible-playbooks-in-your-kitchen Automation Robby Callicotte en false https://pretalx.com/txlf2026/talk/MNVVLN/ https://pretalx.com/txlf2026/talk/MNVVLN/feedback/ Bevo Talk 2026-11-07T16:00:00-06:00 16:00 00:50 Follow me and enter the world of VistaTorn, a bleak and dreary country full of inadequate search, elongated wait-times, hardcoded permissions and no observability as far as the eye can see (which is not very far). In this place rises a hero, a visionary, a beacon of innovation... and that hero is YOU! Learn how to navigate the murky waters of search, with no semantics or vector databases in sight. What tools should you implement to upgrade and conquer these streams of data? Within the castle, work to improve the throughput of services and investigate any and all latencies that arise through instrumentation of standardized and trusted observability tools. Join me on this journey to conquer your stack and save VistaTorn! txlf2026-102208-how-to-upgrade-your-system-like-a-jrpg Observability Sophia Solomon en false https://pretalx.com/txlf2026/talk/BMNQWK/ https://pretalx.com/txlf2026/talk/BMNQWK/feedback/ Bevo Talk 2026-11-07T17:00:00-06:00 17:00 00:50 In enterprises or organizations users and their 2nd factors are managed centrally by you, the IT department. This works fine with logins to central services like web portals, the VPN or your SSH servers. But when it comes to the login at notebooks it comes to some additional challenges you need to overcome. Even in 2026 - notebooks can go offline. There are many different technical ways for multi factor authentication with different underlying cryptographic mechanisms. Symmetic keys or assymmetic keys. Some can work offline, some are hard to manage, some do not work offline in a sensible way. In this talk we are looking into different authentication mechanism be it x509 certificates, OTP or FIDO2 to name a few with their pros and cons. Then we check how the open source multi factor management system privacyIDEA implements different ways of authenticating offline at your user's notebooks - be it Linux or Windows - with OTP tokens or with FIDO2 devices. In 2019 I held a talk at TXLF about privacyIDEA in general as an on prem multi factor solution. We are now looking deeper into management of FIDO2 devices and management for local logins to Linux and Windows machines. You will get in depth knowledge how you can secure your notebooks with reliable authentication hardware like Yubikeys or other FIDO2 devices centrally managed in your own on prem privacyIDEA system. txlf2026-98902-centrally-managed-multi-factor-authentication-for-all-your-users-notebooks Security Cornelius Kölbel en false https://pretalx.com/txlf2026/talk/JZGLEW/ https://pretalx.com/txlf2026/talk/JZGLEW/feedback/