[{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/ai/","section":"Tags","summary":"","title":"AI","type":"tags"},{"content":" Why Should I Use AI for the Development of Yocto-Based Embedded Systems? # The development of embedded systems with Yocto can be complex and time-consuming.\nThis is problematic:\nSignificant development and time effort flows into the BitBake-based build system instead of features, bug fixes, and security patches. Well-trained Yocto developers are rare. Internal Yocto training within the team is time-consuming. Requirements in the area of IT security (supply chain security) bind development and time resources: Security patches. Security requirement engineering/management. Threat modeling. Prioritization conflict between new features and IT security best practices. AI can:\nSignificantly simplify and accelerate working with the BitBake build system. Agent Skills have multiple benefits: They are the foundation for reproducible and scalable acceleration of development as well as training material for internal Yocto training. How Can AI Simplify and Accelerate the Development of Yocto-Based Embedded Systems? # Automation of Routine Tasks: AI can automate repetitive tasks such as creating recipes, configuring build environments, and testing images. This saves time and reduces errors.\nOptimization of Build Processes: AI can analyze and optimize build processes to shorten build time and increase efficiency.\nError Detection and Resolution: AI can identify potential errors in recipes or configurations and suggest fixes.\nKnowledge Management: AI can serve as a knowledge database that stores and provides best practices, common problems, and solutions.\nWhat Are Agent Skills? # Agent Skills are special abilities or modules that AI agents use to perform specific tasks. These skills can be considered as pre-built building blocks that enable the automation and optimization of complex processes. In the context of Yocto-based embedded systems, Agent Skills can be used, for example, to:\nGenerate Recipes: Automatic creation of BitBake recipes based on requirements or templates. Optimize Build Configurations: Analysis and adjustment of build configurations to increase efficiency. Detect and Fix Errors: Identification of potential errors in recipes or configurations and suggestions for fixes. Generate Documentation: Automatic creation of documentation for recipes, configurations, and build processes. Agent Skills are typically modular and can be combined or extended as needed. In other areas of non-technical IT, Agent Skills are developed collaboratively in open source.\nWhen and Why Do I Need Agent Skills? # Complex Build Processes: If your Yocto-based build processes are complex and require many manual steps, Agent Skills can help automate and simplify these processes.\nRecurring Tasks: If you frequently have recurring tasks such as creating recipes or configuring build environments, Agent Skills can take over these tasks and save time.\nError-Prone Processes: If certain processes are error-prone and require frequent manual checks, Agent Skills can help detect and fix errors early.\nKnowledge Management: If you want to share and document knowledge and best practices within your team, Agent Skills can serve as a knowledge database and provide this information.\nScaling: If you want to scale your development processes, Agent Skills can help increase efficiency and improve quality.\nReproducibility: Agent Skills can increase the reproducibility of agent-based development compared to pure prompting. By using Agent Skills, you can ensure that certain tasks are always executed in the same way, which improves the consistency and reliability of your development processes.\nAre There Open Source Agent Skill Libraries for the Yocto Project? # Yes.\nThere have been Agent Skills for the Yocto project since approximately early 2026 on GitHub (search: agent skills yocto).\nAre Open Source Agent Skill Libraries for the Yocto Project Universally Applicable? # No.\nWith Agent Skills, it is generally important to note that software-based products/projects are always based on multiple aspects:\nFor Yocto-based operating systems, the following applies:\nContain generic software components: These include, for example, software components like systemd, which are used in all or at least most Yocto operating systems. Integration of generic software components is often hardware-specific. System requirement-dependent, specific software components: A Yocto operating system that has real-time requirements will typically use different programming languages and middleware or frameworks for communication between components than one without real-time requirements. Domain-dependent, specific system architecture: Embedded systems typically differ more significantly from competing products than is the case in non-technical IT. The operating system of a company in the consumer electronics sector differs fundamentally from one in the energy industry. How Cost-Intensive and Time-Consuming Is It to Develop Your Own Agent Skills? # The development of Agent Skills is a sub-area of modern software development.\nAgent Skills should:\nBe reviewed. Be optimized. Be evaluated. Be reviewed/audited regarding IT security. Be maintained: Agent Skills often refer to specific versions of software components that should be consistently reviewed and updated with new updates. Are Agent Skills Used in Companies? # Yes.\nFrom my network, I know that Agent Skills are used for the development of Yocto-based operating systems. These are not publicly available for various reasons (IT security, protection of intellectual property, etc.).\nWhat Are the Implications of Using Agent Skills? # Agents depend on dynamic interaction with the BitBake build. This means that the most cost-effective efficiency improvements are possible within locally executed BitBake builds. Agent-based optimization of CI-based BitBake builds is possible but much more complex than optimizing local BitBake builds. Agents use models to respond to prompts. Technically, cloud providers that provide models (Anthropic, \u0026hellip;) can access the prompts and responses. Depending on company-internal requirements for IT security, intellectual property, and possibly data privacy, it may be necessary that either models (used by agents to respond to prompts) are either hosted locally or provided within the internal IT infrastructure. ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/ki-basierte-yocto-entwicklung/","section":"Blog Posts","summary":"Why Should I Use AI for the Development of Yocto-Based Embedded Systems? # The development of embedded systems with Yocto can be complex and time-consuming.\nThis is problematic:\nSignificant development and time effort flows into the BitBake-based build system instead of features, bug fixes, and security patches. Well-trained Yocto developers are rare. Internal Yocto training within the team is time-consuming. Requirements in the area of IT security (supply chain security) bind development and time resources: Security patches. Security requirement engineering/management. Threat modeling. Prioritization conflict between new features and IT security best practices. AI can:\n","title":"AI-Based Development of Yocto-Based Embedded Systems","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/artificial-intelligence/","section":"Categories","summary":"","title":"Artificial Intelligence","type":"categories"},{"content":"For efficient creation and maintenance of blog posts, I use artificial intelligence.\nAll english blog posts are generated from the german blog posts. Because I review german blog posts only I do not guarantee for the correctness of the english blog post versions.\nAll content is based on personal experience and is personally reviewed by me before publication.\nThank you for your understanding.\nHave fun reading and learning!\nRecent AI-Based Development of Yocto-Based Embedded Systems July 7, 2026\u0026middot;Updated: August 11, 2026\u0026middot;903 words\u0026middot;5 mins Embedded Systems AI Yocto ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/","section":"Blog Posts","summary":"For efficient creation and maintenance of blog posts, I use artificial intelligence.\nAll english blog posts are generated from the german blog posts. Because I review german blog posts only I do not guarantee for the correctness of the english blog post versions.\nAll content is based on personal experience and is personally reviewed by me before publication.\nThank you for your understanding.\nHave fun reading and learning!\nRecent AI-Based Development of Yocto-Based Embedded Systems July 7, 2026·Updated: August 11, 2026·903 words·5 mins Embedded Systems AI Yocto ","title":"Blog Posts","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/","section":"Categories","summary":"","title":"Categories","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/cloud/","section":"Tags","summary":"","title":"Cloud","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/container/","section":"Categories","summary":"","title":"Container","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/containers/","section":"Categories","summary":"","title":"Containers","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/data-privacy/","section":"Tags","summary":"","title":"Data Privacy","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/datenschutz/","section":"Tags","summary":"","title":"Datenschutz","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/deutschland/","section":"Tags","summary":"","title":"Deutschland","type":"tags"},{"content":"In this article, I describe the technology stack and development workflow I used for this website.\nHugo as Static Website Generator # I chose Hugo because it:\nIs blazing fast - pages are generated in milliseconds Offers simple Markdown syntax - perfect for blog content Markdown files are ideal for agentic development - can be easily read and edited by AI agents Has a theming system - allows easy customizations Themes as Git submodules - can be cleanly integrated and maintained as separate Git submodules Provides multilingual support - ideal for my German/English website Requires no database - everything is static files Has significantly less maintenance overhead than JavaScript-based websites - no constant updates of JS/TypeScript dependencies needed Blowfish as Hugo Theme # For the design, I use the Blowfish theme because it:\nOffers a modern, clean design Is responsive - works on all devices Contains many shortcodes for advanced features Supports dark mode Has good documentation Provides very useful shortcodes such as: {{\u0026lt; mermaid \u0026gt;}} for rendering Mermaid-diagrams {{\u0026lt; codeimporter \u0026gt;}} for flexible code snippet embedding {{\u0026lt; mdimporter \u0026gt;}} for flexible markdown snippet embedding {{\u0026lt; video \u0026gt;}} for flexible video embedding (local, URL) \u0026hellip; The theme is excellent\u0026hellip; many thanks to Nuno Coração from Portugal.\nZed as IDE for Agentic Development # For code editing, I use Zed, a modern IDE optimized for agentic workflows:\nExtremely fast - Based on Rust with native performance AI integration - Zed AI for intelligent code completion Agentic engineering - Support for parallel agents Multi-agent collaboration - Parallel working agents Markdown support - Built-in Markdown support for seamless Hugo integration CSS support - Built-in CSS support for theme customizations HTML support - Built-in HTML support for theme customizations YAML support - Built-in YAML support for GitHub CI/CD configurations TOML support - TOML support via extension for Hugo configurations Git - In addition to the excellent built-in Git support, Git Firefly provides syntax highlighting GitHub Actions - GitHub Actions Extension and GitHub Actions Snippets Extension make working with GitHub Actions workflows easier Mistral AI Models # An important part of my development workflow is using Mistral AI, particularly with the extremely cost-effective Mistral Small model:\nMistral AI Integration in Zed IDE Extremely cost-efficient - The entire AI usage for setting up and migrating the old Astro-based website cost less than €5 (without using parallel agents and Spec-Driven Development with OpenSpec). Mistral AI Costs June 2026 Mistral AI Costs July 2026 Markdown-optimized - Ideal for editing Hugo content Agentic development - Enables automated content creation and editing Multilingual support - Perfect for my bilingual or multilingual content. Writing posts in the native language, automated translation into foreign languages. Generative Engine Optimization (GEO) considers the topicality of the content. If content is not regularly updated or outdated content is not removed, the website\u0026rsquo;s ranking deteriorates. In this respect, AI Depending on the prompting, models generate commit messages like the following for traceability of generated code.\nMistral AI Git Commit Message Agent Prompts # As always in agentic workflows, the right methodology is crucial: For smaller and larger changes to the website, clever prompts like the following are usually absolutely sufficient:\nExemplary prompt for translating posts Agent Skill Libraries? # As always, prompts can be made reusable with Agent Skills and shared with the community, if desired, as a Skill Library.\nSo far, I have not found any or any suitable open source Agent Skills libraries for me. Therefore, I have written my own Agent Skills. All Agent Skills are located in the standardized project-specific directory path \u0026lt;Project\u0026gt;/.agents/skills/. This ensures that Agent Skills work with every agent (e.g. Hermes Agent) and every IDE (not only with the Zed IDE I use).\nHugo # Currently, there are no public Skill Libraries on GitHub for Hugo.\nMy Hugo-specific Agent Skills are located in \u0026lt;Project\u0026gt;/.agents/skills/hugo/.\nBlowfish Theme # Unfortunately, there is also (not yet) no skill library for the Blowfish Hugo theme.\nMy Blowfish-specific Agent Skills are located in \u0026lt;Project\u0026gt;/.agents/skills/hugo-blowfish/.\nWebsite Content # For the content aspects of the website, I have written Agent Skills for:\nGenerating new German posts -\u0026gt; \u0026lt;Project\u0026gt;/.agents/skills/post-create Translating from German to other languages (currently English) -\u0026gt; \u0026lt;Project\u0026gt;/.agents/skills/post-translate Updating non-German posts based on the current state of German posts -\u0026gt; \u0026lt;Project\u0026gt;/.agents/skills/posts-update Spec-Driven Development? # I have also experimented with the Spec-Driven Development framework OpenSpec. Research (YouTube, \u0026hellip;) shows that OpenSpec compares to other Spec-Driven Development frameworks like BMad Method (absolute overkill, compared to OpenSpec about factor 8 worse \u0026ldquo;performance\u0026rdquo; in terms of implementation time and costs) and Spec Kit (somewhat more complex and cumbersome than OpenSpec).\nUsing Spec-Driven Development can be interesting for larger websites. For my previous use cases, I found the overhead too large.\nTo ensure traceability of website changes, Spec-Driven Development is not required:\nI never use model-generated commit messages (or prevent the generation by appropriate instructions in the \u0026lt;Project\u0026gt;/AGENTS.md file). Instead, I review all agentic code changes and manually create Conventional Commits or \u0026ldquo;Conventional Merge Request Squash Commits\u0026rdquo; (summary of multiple features, fixes, etc. in a commit message summarized for a merge request).\nTo cover the technological aspects of a website with Conventional Commits, the common types feat (new/changed/removed feature) and fix (problem resolution) are sufficient.\nTo cover the content aspect of a website with Conventional Commits, the extension of Conventional Commits to include the types page (new/changed/removed page) and post (new/changed/removed blog post) is required.\nPerhaps I will still write an Agent Skill that prepares commit messages for me without automatically committing. So far, the manual effort has been reasonable.\nGitHub for Version Control # I use GitHub for code versioning.\nCode Versioning # The entire source code of this website is versioned on GitHub:\nFree - €0 per month for code versioning Complete history of all changes Branches for experimental features Continuous Integration (CI), Continuous Deployment (CD) and Hosting # AI-optimized websites must load as fast as possible. The majority of my website visits come from countries that are not in Germany and the EU.\nTherefore, I want to use a Content Delivery Network (CDN)-based solution for deployment.\nMy previous hosting with Hetzner benefited from the world\u0026rsquo;s largest international internet hub in Frankfurt. However, Hetzner does not offer a CDN-based solution.\nThere are very good European providers of CDN services (e.g. bunny.net). However, they (so far) do not offer me a solution for the integration of CI, CD, and hosting with reasonable overhead.\nTherefore, I have now decided to migrate CI, CD, and hosting to CloudFlare Pages.\nIn my case, this is absolutely uncritical from an IT security perspective, because no secrets are needed and I consider the GitHub integration to be good (minimal data access, minimal repository interaction rights). From a data privacy perspective, the use is justifiable because: Many other German website hosting providers (e.g. Jimbo) consciously host with CloudFlare for IT security reasons And the majority of European internet users accept cookies beyond what is technologically required without hesitation. Workflow # Code is pushed to git repository. No (slow) execution of a CI pipeline is needed. CloudFlare Pages automatically detects code changes via GitHub integration (regarding IT security ok). CloudFlare Pages builds the production page for commits in the branch defined for production deployments. Optionally, preview deployments can be configured and used. Benefits # Free - €0 per month for CI/CD Automatic - no manual deployment needed Fast - changes are online in seconds Reliable - fewer errors through automation Versioned - every deployment is related to a Git commit CDN-based hosting - free hosting, extremely fast website loading times worldwide How \u0026ldquo;Good\u0026rdquo; is the Website? # To check the optimization of websites, I currently use pagespeed.web.dev. Based on the hyperlinks on the website, the service appears to be based on Google Lighthouse.\nSince Google still holds the largest market share in the search engine market, this should not be a bad choice. In particular, the analysis of agentic browsing (optimization of the website for agent-based website access) is relevant to me.\nThe website has a very good score and is therefore almost complete.\nPageSpeed Insights Results - Mobile PageSpeed Insights Results - Computer The current score of the website can be found here.\nConclusion # An extremely cost-efficient, efficient, agentic web development setup\nThis setup allows me to focus entirely on content while the technical infrastructure runs cost-effectively, reliably, and with minimal maintenance in the background. The combination of open-source tools, AI support, and automated workflows represents a modern, forward-looking approach to web development that is ideal for both individual developers and small teams.\nTechnological Advantages # Hugo + Blowfish: Lightning-fast static page generation with professional design and Git submodule support. Minimal maintenance effort by avoiding JavaScript/TypeScript (few security updates needed) and static website (extremely small IT security attack surface). Zed IDE: Optimized development environment with native support for all relevant file formats (Markdown, HTML, CSS, YAML, TOML) Mistral AI: Cost-effective AI support for agentic workflows and content creation CloudFlare Pages: Good integration of CI, CD, and hosting. Extremely fast website loading times worldwide. Static website (extremely small IT security attack surface, no impact of security incidents at hoster CloudFlare on other IT infrastructure) and automated deployment. Data privacy technically justifiable. Financial Advantages # One-time costs:\nUnder €10 AI costs for the complete migration from Astro to Hugo Time effort of approx. 5 hours. Ongoing costs:\n€0/month for CI, CD, and hosting (GitHub and CloudFlare Pages are free) AI costs depending on mood and desire :) Time effort depending on mood and desire :) Time effort: Minimal effort by avoiding JavaScript/TypeScript (few security updates needed) Workflow Advantages # Agentic development: AI agents can work directly with Markdown files Speed: At least 5-10x faster than without agentic workflow Automation: CloudFlare-based CI and CD Scalability: Easy extension through shortcodes and theme customizations Future-proof: Modern technologies with minimal dependencies ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/entwicklung-dieser-webseite/","section":"Blog Posts","summary":"In this article, I describe the technology stack and development workflow I used for this website.\nHugo as Static Website Generator # I chose Hugo because it:\nIs blazing fast - pages are generated in milliseconds Offers simple Markdown syntax - perfect for blog content Markdown files are ideal for agentic development - can be easily read and edited by AI agents Has a theming system - allows easy customizations Themes as Git submodules - can be cleanly integrated and maintained as separate Git submodules Provides multilingual support - ideal for my German/English website Requires no database - everything is static files Has significantly less maintenance overhead than JavaScript-based websites - no constant updates of JS/TypeScript dependencies needed Blowfish as Hugo Theme # For the design, I use the Blowfish theme because it:\n","title":"Development of This Website: Technology Stack and Workflow","type":"posts"},{"content":" Was ist EuroStack? # EuroStack ist eine europäische Graswurzelbewegung mit über 400 Industrieunterzeichnern, Technologen und Wirtschaftswissenschaftlern, die radikales industrielles Handeln fordert, um die souveräne digitale Infrastruktur Europas aufzubauen. Sie behandelt den dringenden Bedarf an technologischer Unabhängigkeit auf allen Ebenen: von KI und Cloud Computing bis hin zu Chips und Konnektivität.\nDie Krise # Europa sieht sich mit kritischen Abhängigkeiten von nicht-europäischen Technologien konfrontiert, die Sicherheitsrisiken darstellen und die Souveränität beeinträchtigen. Aufgrund struktureller Fragmentierung ist Europa deutlich hinter den USA und China zurückgefallen. Ohne tiefgreifende Veränderungen wird die Abhängigkeit Europas von nicht-europäischen Technologien innerhalb von drei Jahren nahezu vollständig sein.\nWas EuroStack fordert # Schaffung von Nachfrage: Regierungsverpflichtungen zum \u0026ldquo;Kauf Europäisch\u0026rdquo;, um Investitionen in europäische Lösungen zu fördern.\nSchaffung eines lebensfähigen Angebots: Bündelung und föderale Zusammenführung bestehender europäischer Vermögenswerte, Unterstützung von Open-Source-Lösungen und Entwicklung harmonisierter Standards für souveräne Clouds.\nFinanzierungsstrategie: Ein gewidmeter Fonds für souveräne Infrastruktur für kapitalintensive Bereiche wie Chips und Quantencomputing, ergänzt durch die Initiative Digitales Jahrzehnt.\nWarum jetzt? # Aktuelle geopolitische Entwicklungen haben strategische Autonomie zu einem dringenden Imperativ gemacht. Die Industrie ist bereit zu investieren – Europa muss entscheidend handeln, um seine digitale Zukunft zu schützen.\nMehr erfahren # Hauptinitiative: eurostack.eu Weißbuch: Europäische strategische souveräne digitale Infrastrukturen Industriebrief: Offener Brief an die EU-Führung Katalogprojekt: Europäisches Katalogprojekt Industrie-Unterstützung: 400+ Unterzeichner Verwandte Initiativen: Ökosystem \u0026amp; Partnerschaften ","date":"07. July 2026","externalUrl":null,"permalink":"/posts/eurostack-initiative-erklaert/","section":"Blog Beiträge","summary":"Was ist EuroStack? # EuroStack ist eine europäische Graswurzelbewegung mit über 400 Industrieunterzeichnern, Technologen und Wirtschaftswissenschaftlern, die radikales industrielles Handeln fordert, um die souveräne digitale Infrastruktur Europas aufzubauen. Sie behandelt den dringenden Bedarf an technologischer Unabhängigkeit auf allen Ebenen: von KI und Cloud Computing bis hin zu Chips und Konnektivität.\nDie Krise # Europa sieht sich mit kritischen Abhängigkeiten von nicht-europäischen Technologien konfrontiert, die Sicherheitsrisiken darstellen und die Souveränität beeinträchtigen. Aufgrund struktureller Fragmentierung ist Europa deutlich hinter den USA und China zurückgefallen. Ohne tiefgreifende Veränderungen wird die Abhängigkeit Europas von nicht-europäischen Technologien innerhalb von drei Jahren nahezu vollständig sein.\n","title":"Die EuroStack Initiative erklärt","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/digital-sovereignty/","section":"Categories","summary":"","title":"Digital Sovereignty","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/digital-sovereignty/","section":"Tags","summary":"","title":"Digital Sovereignty","type":"tags"},{"content":"Two northern German states are serious about digital sovereignty: Schleswig-Holstein and Mecklenburg-Vorpommern are exiting the Microsoft Cloud and switching to open-source solutions instead. This decision marks a significant step towards technological independence and data privacy.\nBackground of the Decision # The two states have decided to reduce their dependence on proprietary software and switch to open-source alternatives instead. This decision is based on several key factors:\nData Privacy and Compliance: Stricter requirements for the protection of personal data Cost Efficiency: Long-term savings through reduced licensing costs Technological Sovereignty: Independence from individual vendors Interoperability: Better integration with other public IT systems Timeline and Implementation # The migration is taking place in several phases:\n2024: Pilot phase with selected authorities 2025: Complete migration of state administrations 2026: Integration into municipal systems Technological Alternatives # Instead of Microsoft 365, the following solutions are being used:\nNextcloud: For file storage and collaboration OnlyOffice: As an office alternative Jitsi: For video conferencing Matrix/Element: For secure communication Impact on Public Administration # Data Control: Full control over citizen data Innovation Promotion: Strengthening local IT service providers Transparency: Open standards enable better auditing Collaboration: Easier data exchange between authorities National Significance # The decision by the two northern German states could set a precedent:\nBavaria: Examining similar steps North Rhine-Westphalia: Pilot projects started Federal Government: Discussion about nationwide strategy Conclusion: A Step in the Right Direction # \u0026ldquo;Digital sovereignty is not a luxury, but a necessity for modern administrative work.\u0026rdquo; — IT expert from the state government\nThe exit from the Microsoft Cloud by Schleswig-Holstein and Mecklenburg-Vorpommern shows that public administration and digital sovereignty are not contradictory. This decision could mark the beginning of a nationwide movement towards greater technological independence.\nFurther Information # ZenDiS: Center for Digital Sovereignty of Public Administration EuroStack Initiative: European initiative to strengthen European digital sovereignty ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/digitale-souveraenitaet-oeffentliche-verwaltung/","section":"Blog Posts","summary":"Two northern German states are serious about digital sovereignty: Schleswig-Holstein and Mecklenburg-Vorpommern are exiting the Microsoft Cloud and switching to open-source solutions instead. This decision marks a significant step towards technological independence and data privacy.\nBackground of the Decision # The two states have decided to reduce their dependence on proprietary software and switch to open-source alternatives instead. This decision is based on several key factors:\n","title":"Digital Sovereignty in Public Administration","type":"posts"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/digitale-souver%C3%A4nit%C3%A4t/","section":"Categories","summary":"","title":"Digitale Souveränität","type":"categories"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/digitale-souver%C3%A4nit%C3%A4t/","section":"Tags","summary":"","title":"Digitale Souveränität","type":"tags"},{"content":"Zwei norddeutsche Bundesländer machen Ernst mit digitaler Souveränität: Schleswig-Holstein und Mecklenburg-Vorpommern steigen aus der Microsoft-Cloud aus und setzen stattdessen auf Open-Source-Lösungen. Diese Entscheidung markiert einen bedeutenden Schritt in Richtung technologischer Unabhängigkeit und Datenschutz.\nHintergrund der Entscheidung # Die beiden Bundesländer haben beschlossen, ihre Abhängigkeit von proprietärer Software zu verringern und stattdessen auf Open-Source-Alternativen zu setzen. Diese Entscheidung basiert auf mehreren Schlüsselfaktoren:\nDatenschutz und Compliance: Strengere Anforderungen an den Schutz personbezogener Daten Kosteneffizienz: Langfristige Einsparungen durch reduzierte Lizenzkosten Technologische Souveränität: Unabhängigkeit von einzelnen Anbietern Interoperabilität: Bessere Integration mit anderen öffentlichen IT-Systemen Zeitplan und Umsetzung # Die Migration läuft in mehreren Phasen ab:\n2024: Pilotphase mit ausgewählten Behörden 2025: Vollständige Migration der Landesverwaltungen 2026: Integration in kommunale Systeme Technologische Alternativen # Statt Microsoft 365 kommen folgende Lösungen zum Einsatz:\nNextcloud: Für Dateiablage und Kollaboration OnlyOffice: Als Office-Alternative Jitsi: Für Videokonferenzen Matrix/Element: Für sichere Kommunikation Auswirkungen auf die öffentliche Verwaltung # Datenhoheit: Volle Kontrolle über Bürgerdaten Innovationsförderung: Stärkung lokaler IT-Dienstleister Transparenz: Offene Standards ermöglichen bessere Prüfbarkeit Zusammenarbeit: Einfacherer Datenaustausch zwischen Behörden Bundesweite Bedeutung # Die Entscheidung der beiden norddeutschen Länder könnte Schule machen:\nBayern: Prüft ähnliche Schritte Nordrhein-Westfalen: Pilotprojekte gestartet Bund: Diskussion über bundesweite Strategie Fazit: Ein Schritt in die richtige Richtung # \u0026ldquo;Digitale Souveränität ist kein Luxus, sondern eine Notwendigkeit für moderne Verwaltungsarbeit.\u0026rdquo; — IT-Experte der Landesregierung\nDer Ausstieg aus der Microsoft-Cloud durch Schleswig-Holstein und Mecklenburg-Vorpommern zeigt, dass öffentliche Verwaltung und digitale Souveränität kein Widerspruch sein müssen. Diese Entscheidung könnte den Anfang einer bundesweiten Bewegung hin zu mehr technologischer Unabhängigkeit machen.\nWeiterführende Informationen # ZenDiS: Zentrum für Digitale Souveränität der Öffentlichen Verwaltung EuroStack Initiative: Europäische Initiative zur Stärkung europäischer, digitaler Souveränität ","date":"07. July 2026","externalUrl":null,"permalink":"/posts/digitale-souveraenitaet-oeffentliche-verwaltung/","section":"Blog Beiträge","summary":"Zwei norddeutsche Bundesländer machen Ernst mit digitaler Souveränität: Schleswig-Holstein und Mecklenburg-Vorpommern steigen aus der Microsoft-Cloud aus und setzen stattdessen auf Open-Source-Lösungen. Diese Entscheidung markiert einen bedeutenden Schritt in Richtung technologischer Unabhängigkeit und Datenschutz.\nHintergrund der Entscheidung # Die beiden Bundesländer haben beschlossen, ihre Abhängigkeit von proprietärer Software zu verringern und stattdessen auf Open-Source-Alternativen zu setzen. Diese Entscheidung basiert auf mehreren Schlüsselfaktoren:\n","title":"Digitale Souveränität in der öffentlichen Verwaltung","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/embedded-systems/","section":"Categories","summary":"","title":"Embedded Systems","type":"categories"},{"content":"In diesem Artikel beschreibe ich den Technologie-Stack und den Entwicklungsworkflow, den ich für diese Webseite verwendet habe.\nHugo als statischer Webseiten-Generator # Ich habe mich für Hugo entschieden, weil es:\nBlitzschnell ist - Seiten werden in Millisekunden generiert Einfache Markdown-Syntax bietet - perfekt für Blog-Inhalte Markdown-Dateien sind ideal für agentische Entwicklung - können leicht von KI-Agenten gelesen und bearbeitet werden Theming-System hat - ermöglicht einfache Anpassungen Themes als Git Submodules - können sauber als separate Git-Submodules eingebunden und gewartet werden Multilingual-Unterstützung bietet - ideal für meine deutsch/englische Webseite Keine Datenbank benötigt - alles sind statische Dateien Wesentlich weniger Wartungsaufwand als JavaScript-basierte Webseiten - keine ständigen Updates von JavaScript/TypeScript-Dependencies nötig Blowfish als Hugo-Theme # Für das Design verwende ich das Blowfish-Theme, weil es:\nModernes, sauberes Design bietet Responsive ist - funktioniert auf allen Geräten Viele Shortcodes enthält für erweiterte Funktionen Dark Mode unterstützt Gute Dokumentation hat Sehr nützliche Shortcodes anbietet: {{\u0026lt; mermaid \u0026gt;}} für das Rendering von Mermaid-Diagrammen {{\u0026lt; codeimporter \u0026gt;}} für das flexible Einbinden von Code-Snippets {{\u0026lt; mdimporter \u0026gt;}} für das flexible Einbinden von Markdown-Snippets {{\u0026lt; video \u0026gt;}} für das flexible Einbinden von Videos (lokal, URL) \u0026hellip; Das Theme ist großartig\u0026hellip; vielen Dank an Nuno Coração aus Portugal.\nZed als IDE für agentische Entwicklung # Für die Code-Bearbeitung nutze ich Zed, eine moderne IDE, die speziell für agentische Workflows optimiert ist:\nExtrem schnell - Basierend auf Rust mit nativer Performance KI-Integration - Zed AI für intelligente Code-Vervollständigung Agentische Engineering - Unterstützung für parallele Agents Multi-Agent Collaboration - Parallel arbeitende Agents Markdown-Unterstützung - Built-in Markdown-Unterstützung für nahtlose Hugo-Integration CSS-Unterstützung - Built-in CSS-Unterstützung für Theme-Anpassungen HTML-Unterstützung - Built-in HTML-Unterstützung für Theme-Anpassungen YAML-Unterstützung - Built-in YAML-Unterstützung für GitHub CI/CD-Konfigurationen TOML-Unterstützung - TOML-Unterstützung per Extension für Hugo-Konfigurationen Git-Unterstützung - Zusätzlich zur sehr guten Built-in Git-Unterstützung bietet Git Firefly Syntax-Highlighting Github Actions - GitHub Actions Extension und GitHub Actions Snippets Extension erleichtern die Arbeit mit GitHub Actions Workflows KI-Modelle von Mistral AI # Ein wichtiger Teil meines Entwicklungsworkflows ist die Nutzung von Mistral AI, insbesondere mit dem extrem kostengünstigen Mistral Small Modell:\nMistral AI Integration in Zed IDE Extrem kosteneffizient - Die gesamte KI-Nutzung für das Aufsetzen und die Migration der alten Astro-basierten Webseite hat nicht einmal 5 € gekostet (ohne Nutzung paralleler Agents und Spec-Driven Development mit OpenSpec). Mistral AI Kosten Juni 2026 Mistral AI Kosten Juli 2026 Markdown-optimiert - Ideal für die Bearbeitung von Hugo-Inhalten Agentische Entwicklung - Ermöglicht automatisierte Content-Erstellung und Content-Bearbeitung Mehrsprachige Unterstützung - Perfekt für meine bilingualen oder auch multi-lingualen Inhalte. Verfassen von Posts in der Muttersprache, automatisierte Übersetzung in Fremdsprachen. Generative Engine Optimization (GEO) berücksichtigt die Aktualität der Inhalte. Werden Inhalte nicht regelmäßig aktualisiert oder nicht mehr aktuelle Inhalte entfernt, verschlechtert sich das Ranking der Webseite. Insofern ist KI Je nach Prompting erzeugen Modelle Commit-Nachrichten wie die folgende Zwecks Nachvollziehbarkeit von erzeugtem Code.\nMistral AI Git Commit Message Agent Prompts # Wie immer in agentischen Workflows, ist die richtige Methodik entscheident: Bei kleineren und größeren Änderungen der Webseite sind klevere Prompts wie der folgende in der Regel absolut ausreichend:\nExemplarischer Prompt zur Übersetzung von Posts Agent Skill Bibliotheken? # Wie immer können Prompts mit Agent Skills wiederverwendbar und mit der Community, falls gewünscht, als Skill Library geteilt werden.\nBisher habe ich keine oder keine für mich passendend Open Source Agent-Skills- Bibliotheken gefunden. Daher habe ich mir eigene Agent-Skills geschrieben. Alle Agent-Skills liegen in dem standardisierten Projekt-spezifischen Verzeichnisfpad \u0026lt;Projekt\u0026gt;/.agents/skills/. Das stellt sicher, dass Agents-Skills mit jedem Agent (z.B. Hermes Agent) und jeder IDE (nicht nur mit der von mir verwendeten Zed-IDE) funktionieren.\nHugo # Für Hugo existieren derzeit keine öffentlichen Skill Libraries auf Github.\nMeine Hugo-spezifischen Agent-Skills liegen in \u0026lt;Projekt\u0026gt;/.agents/skills/hugo/.\nBlowfish-Theme # Für das Blowfish Hugo-Theme gibt es derzeit leider ebenfalls (noch) keine Skill-Bibliothek.\nMeine Blowfish-spezifischen Agent-Skills liegen in \u0026lt;Projekt\u0026gt;/.agents/skills/hugo-blowfish/.\nWebseiten-Inhalt # Für die inhaltlichen Aspekte der Website habe ich mir Agent-Skills geschrieben für:\ndie Erzeugung von neuen deutschen Posts -\u0026gt; \u0026lt;Projekt\u0026gt;/.agent/skills/post-create die Übersetzung von deutschen in andere Sprachen (derzeit Englisch) -\u0026gt; \u0026lt;Projekt\u0026gt;/.agent/skills/post-translate die Aktualisierung der nicht-deutschen Posts auf Basis des aktuellen Stands der deutschen Posts -\u0026gt; \u0026lt;Projekt\u0026gt;/.agent/skills/posts-update Spec-Driven Development? # Ich habe auch mit dem Spec-Driven Development Framework OpenSpec experimentiert. Recherchen (YouTube, \u0026hellip;) nach ist OpenSpec im Vergleich zu anderen Spec-Driven Development Frameworks wie BMad Method (absoluter Overkill, im Vergleich zu OpenSpec ca. Faktor 8 schlechtere \u0026ldquo;Performance\u0026rdquo; was Implementierungsdauer und Kosten angeht) und Spec Kit (etwas komplexer und umständlicher als OpenSpec). Die Nutzung von Spec-Driven Development kann für größere Webseiten interessant sein. Für meine bisherigen Anwendungsfälle fand ich den Overhead zu groß.\nUm die Nachvollziehbarkeit der Webseiten-Änderungen ist keine Spec-Driven Development erforderlich:\nIch nutze nie von Modellen generierte Commit-Nachrichten (bzw. unterbinde die Erzeugung bereits durch entsprechende Anweisungen in der \u0026lt;Projekt\u0026gt;/AGENTS.md-Datei). Stattdessen reviewe ich alle agentischen Code-Änderungen und erstelle manuell Conventional Commits oder \u0026ldquo;Conventional Merge Request Squash Commits\u0026rdquo; (Zusammenfassung von mehreren Features, Fixes, \u0026hellip; in einer für einen Merge Request zusammengefassten Commit-Nachricht).\nUmd die technologischen Aspekte einer Webseite mit Conventoinal Commits abdecken zu können sind die gängigen Typen feat (neues/geändertes/entferntes Feature) und fix (Problembehebung) ausreichend.\nUm den inhaltlichen Aspekt einer Webseite mit Conventional Commits abdecken zu können ist die Erweiterung von Conventional Commits um die Typen page (neue/geänderte/entfernte Seite) und post (neuer/geänderter/entfernter Blog-Beitrag) erforderlich.\nVielleicht werde ich noch einen Agent-Skill schreiben, der mir die Commit-Nachrichten vorbereitet ohne automatisch zu commiten. Bisher war der manuelle Mehraufwand vertretbar.\nGitHub für Versionierung # Ich nutze GitHub für die Code-Versionierung.\nDer gesamte Quellcode dieser Webseite wird auf GitHub versioniert:\nKostenlos - 0 € pro Monat für Code-Versionierung Vollständige Historie aller Änderungen Branches für experimentelle Features Continuous Integration (CI), Continous Deployment (CD) und Hosting # KI-optimierte Webseiten müssen so schnell laden wie möglich. Die Mehrzahl meiner Webseiten-Aufrufe findet aus Ländern statt, die nicht in Deutschland und der EU liegen.\nIch möchte daher eine Content Delivery Network (CDN) basierte Lösung für das Deployment nutzen.\nMein bisheriges Hosting über Hetzner hat von dem weltweit größten, internationalen Internetknotenpunkt in Frankfurt profitiert. Hetzner bietet aber keine CDN-basierte Lösung an.\nEs gibt sehr gute europäische Anbieter von CDN-Angeboten (z.b. bunny.net). Allerdings bieten Sie (bisher) für mich keine Lösung für die Integration von CI, CD und Hosting mit vertretbarem Mehraufwand.\nDaher habe ich mich jetzt doch dazu entschlossen CI, CD und Hosting zu CloudFlare Pages zu migrieren.\nIn meinem Fall ist das IT-Security-technisch absolut unkritisch, weil keine Secrets notwendig sind und ich die GitHub-Integration für gut halte (minimale Datenzugriffe, minimale Repository-Interaktions-Rechte). Datenschutz-technisch ist die Nutzung vertretbar, weil: viele andere deutsche Anbieter von Webseitenhosting (z.B. Jimbo) aus einer IT-Security-Motivation heraus ganz bewusst bei CloudFlare hosten und der Großteil der europäischen Internetnutzer ohnehin ohne zu Zögern über die technologisch erforderlichen hinausgehenden Cookies akzeptieren. Workflow # Code wird in git-Repository gepusht. Es ist keine (langsame) Ausführung einer CI-Pipeline nötig. CloudFlare Pages erkennt über die GitHub-Integration (bezüglich IT-Security ok) automatisch Code-Änderungen. CloudFlare Pages baut bei Commits in den Branch, der für Production-Deployments definiert ist die Production-Page. Optional können zusätzlich Preview-Deployments konfiguriert und genutzt werden. Vorteile # Kostenlos - 0 € pro Monat für CI/CD Automatisch - kein manuelles Deployment nötig Schnell - Änderungen sind in Sekunden online Zuverlässig - weniger Fehler durch Automatisierung Versioniert - jeder Deployment steht in Beziehung zu einem Git-Commit CDN-basiertes Hosting - kostenloses Hosting, weltweit extrem schnelle Ladezeiten der Webseite Wie \u0026ldquo;gut\u0026rdquo; ist die Webseite? # Zur Prüfung der Optimierung von Webseiten nutze ich derzeit pagespeed.web.dev. Den Hyperlinks auf der Webseite nach, scheint der Service auf Google Lighthouse zu basieren. Da Google im Suachmaschinen-Markt immer noch den größten Martanteil hält, sollte das keine schlechte Wahl sein. Insbesondere die Analyse zur agentischem Browsen (Optimierung der Webseite, für Agenten-basierte Webseitenzugriffe) ist für mich relevant.\nDie Webseite hat ein sehr gutes Scoring und ist damit fast abgeschlossen.\nPageSpeed Insights Ergebnisse - Mobile PageSpeed Insights Ergebnisse - Computer Das aktuelle Scoring der Webseite ist hier zu finden.\nFazit # Ein extrem kosteneffizientes, effizientes, agentisches Web-Entwicklungs-Setup\nDieses Setup ermöglicht mir, mich vollständig auf die Inhalte zu konzentrieren , während die technische Infrastruktur kostengünstig, zuverlässig und wartungsarm im Hintergrund läuft. Die Kombination aus Open-Source-Tools, KI-Unterstützung und automatisierten Workflows stellt einen modernen, zukunftsweisenden Ansatz für Webentwicklung dar, der sowohl für Einzelentwickler als auch für kleine Teams ideal geeignet ist.\nTechnologische Vorteile # Hugo + Blowfish: Blitzschnelle statische Seitengenerierung mit professionellem Design und Git-Submodule-Unterstützung. Minimaler Wartungsaufwand durch Vermeidung von JavaScript/TypeScript (kaum Security-Updates nötig) und statische Webseite (extrem schmale IT-Security-Angriffläche). Zed IDE: Optimierte Entwicklungsumgebung mit nativer Unterstützung für alle relevanten Dateiformate (Markdown, HTML, CSS, YAML, TOML) Mistral AI: Kostengünstige KI-Unterstützung für agentische Workflows und Content-Erstellung CloudFlare Pages: Gute Integration von CI, CD und Hosting. Weltweit extrem schnelle Webseiten-Ladezeiten. Statische Webseite (extrem schmale IT-Security-Angriffläche, keine Auswirkungen von Security-Vorfällen beim Hoster CloudFlare auf weitere IT-Infrastruktur) und automatisiertes Deployment. Datenschutz-technisch vertretbar. Finanzielle Vorteile # Einmalige Kosten:\nUnter 10 € KI-Kosten für die komplette Migration von Astro zu Hugo Zeitaufwand von ca. 5h. Laufende Kosten:\n0 €/Monat für CI, CD und Hosting (GitHub und CloudFlare Pages sind kostenlos) KI-Kosten je nach Lust und Laune :) Zeitaufwand je nach Lust und Laune :) Zeitaufwand: Minimaler Aufwand durch Vermeidung von JavaScript/TypeScript (kaum Security-Updates nötig), Workflow-Vorteile # Agentische Entwicklung: KI-Agenten können direkt mit Markdown-Dateien arbeiten. Geschwindigkeit: Mindestens 5-10x schneller als ohne agentischen Workflow Automatisierung: CloudFlare basierte CI und CD. Skalierbarkeit: Einfache Erweiterung durch Shortcodes und Theme-Anpassungen. Zukunftssicher: Moderne Technologien mit minimalen Abhängigkeiten. ","date":"07. July 2026","externalUrl":null,"permalink":"/posts/entwicklung-dieser-webseite/","section":"Blog Beiträge","summary":"In diesem Artikel beschreibe ich den Technologie-Stack und den Entwicklungsworkflow, den ich für diese Webseite verwendet habe.\nHugo als statischer Webseiten-Generator # Ich habe mich für Hugo entschieden, weil es:\nBlitzschnell ist - Seiten werden in Millisekunden generiert Einfache Markdown-Syntax bietet - perfekt für Blog-Inhalte Markdown-Dateien sind ideal für agentische Entwicklung - können leicht von KI-Agenten gelesen und bearbeitet werden Theming-System hat - ermöglicht einfache Anpassungen Themes als Git Submodules - können sauber als separate Git-Submodules eingebunden und gewartet werden Multilingual-Unterstützung bietet - ideal für meine deutsch/englische Webseite Keine Datenbank benötigt - alles sind statische Dateien Wesentlich weniger Wartungsaufwand als JavaScript-basierte Webseiten - keine ständigen Updates von JavaScript/TypeScript-Dependencies nötig Blowfish als Hugo-Theme # Für das Design verwende ich das Blowfish-Theme, weil es:\n","title":"Entwicklung dieser Webseite: Technologie-Stack und Workflow","type":"posts"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/erkl%C3%A4rt/","section":"Tags","summary":"","title":"Erklärt","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/europa/","section":"Tags","summary":"","title":"Europa","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/europe/","section":"Tags","summary":"","title":"Europe","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/eurostack/","section":"Tags","summary":"","title":"Eurostack","type":"tags"},{"content":" What is EuroStack? # EuroStack is a European grassroots movement with over 400 industry signatories, technologists, and economists demanding radical industrial action to build Europe\u0026rsquo;s sovereign digital infrastructure. It addresses the urgent need for technological independence at all levels: from AI and cloud computing to chips and connectivity.\nThe Crisis # Europe faces critical dependencies on non-European technologies that pose security risks and compromise sovereignty. Due to structural fragmentation, Europe has fallen significantly behind the US and China. Without profound changes, Europe\u0026rsquo;s dependency on non-European technologies will be nearly complete within three years.\nWhat EuroStack Demands # Creating Demand: Government commitments to \u0026ldquo;Buy European\u0026rdquo; to foster investments in European solutions.\nCreating a Viable Supply: Pooling and federating existing European assets, supporting open-source solutions, and developing harmonized standards for sovereign clouds.\nFunding Strategy: A dedicated fund for sovereign infrastructure for capital-intensive areas like chips and quantum computing, complemented by the Digital Decade initiative.\nWhy Now? # Current geopolitical developments have made strategic autonomy an urgent imperative. Industry is ready to invest - Europe must act decisively to protect its digital future.\nLearn More # Main Initiative: eurostack.eu White Paper: European Strategic Sovereign Digital Infrastructures Industry Letter: Open Letter to EU Leadership Catalog Project: European Catalog Project Industry Support: 400+ Signatories Related Initiatives: Ecosystem \u0026amp; Partnerships ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/eurostack-initiative-explained/","section":"Blog Posts","summary":"What is EuroStack? # EuroStack is a European grassroots movement with over 400 industry signatories, technologists, and economists demanding radical industrial action to build Europe’s sovereign digital infrastructure. It addresses the urgent need for technological independence at all levels: from AI and cloud computing to chips and connectivity.\nThe Crisis # Europe faces critical dependencies on non-European technologies that pose security risks and compromise sovereignty. Due to structural fragmentation, Europe has fallen significantly behind the US and China. Without profound changes, Europe’s dependency on non-European technologies will be nearly complete within three years.\n","title":"EuroStack Initiative Explained","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/explained/","section":"Tags","summary":"","title":"Explained","type":"tags"},{"content":" ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/","section":"Florian Kromer","summary":" ","title":"Florian Kromer","type":"page"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/future/","section":"Tags","summary":"","title":"Future","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/germany/","section":"Tags","summary":"","title":"Germany","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/gesundheit/","section":"Categories","summary":"","title":"Gesundheit","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/health/","section":"Categories","summary":"","title":"Health","type":"categories"},{"content":" What is Typst? # Typst\nis a markup language (description of document layout in code), which also supports styling, scripting and context a compiler that converts markup input into output formats A service (you can use it, you don\u0026rsquo;t have to use it), which offers functions for organizations. Why should I use Typst? # Typst is simple # So far, the only real professional alternative to Typst was the markup language LaTeX, which was a real option. Pandoc is a LaTeX compiler often used under the Linux operating system (converts LaTeX-based input data into output formats). Using the two compilers, the properties can be compared in a practical way. Typst is much easier to learn and use compared to LaTeX. I wrote my master\u0026rsquo;s thesis in 2012 with LaTeX. If I had had a simpler alternative back then, I would have saved a lot of time and nerves.\nProperty Typst Pandoc Philosophy Modern, fast and intuitive; replacement for LaTeX with simple syntax. Universal document converter; uses LaTeX for PDF output. Target audience Users who want LaTeX-like results without LaTeX complexity. Users who need flexibility (e.g. Markdown → PDF, Word, HTML). Markdown support Native Markdown support with Typst-specific extensions. Standard Markdown + Pandoc extensions (e.g. tables, quotes, math). Workflow Direct: typst compile input.md output.pdf Indirect: pandoc input.md -o output.pdf (uses LaTeX as backend). Dependencies Typst (no LaTeX required). Pandoc + LaTeX distribution (e.g. texlive). Output quality (standard) Excellent – modern typography and good presets. Good – depends on the LaTeX template (e.g. eisvogel, default). Customization options High – simple syntax for layouts, but less granular than LaTeX. Very high – full LaTeX power, but requires LaTeX knowledge. Fonts \u0026amp; Unicode Great – supports Unicode out of the box. Good – depends on the LaTeX engine (xelatex or lualatex for Unicode). Speed Very fast – compiles in milliseconds. Moderate – depends on LaTeX compilation. Compilation time Almost instant for small documents. Slower (LaTeX is slow for large documents). Best use Fast PDFs from Markdown, modern documents, simple syntax. Conversion between many formats, academic writing with citations. Disadvantages Smaller ecosystem (fewer templates/packages), less mature than LaTeX. Requires LaTeX (slow), LaTeX knowledge for advanced customizations. Example command typst compile input.md output.pdf pandoc input.md -o output.pdf --pdf-engine=xelatex Typical templates Simple, modern templates. LaTeX templates such as eisvogel, default, or custom templates. Learning curve Low – easier than LaTeX, but not as easy as pure Markdown. Medium – Markdown is easy, but LaTeX customizations require effort. Suitability for beginners Good if you have no LaTeX experience. Medium – Markdown is easy, but LaTeX requires familiarization. Suitability for professionals Good for simple to medium-complex documents. Ideal for complex documents with many conversion requirements. There are many free templates # In the Typst Universe many free\nDocument templates Office templates Presentation templates Flyer templates Visualization templates Book templates are offered.\nThere are many happy users # According to the website, there are satisfied users in over 3500 universities/labs and over 1000 companies.\nTypst is particularly popular among data scientists: Joseph Barbier explains in Typst in Production: Turning Data Into Deliverables | Typst Meetup why he uses Typst.\nWhat does a Typst file look like? # Typst expects markup file(s) as input data. The easiest way is to get a first impression using the example in the Typst Playground. The syntax of Typst is relatively simple. The example below shows the use of formulas.\nTypst Math Example Which output formats does Typst support? # The Typst compiler supports the following output formats:\nPDF HTML PNG SVG Bundles How can I customize the layout? # The page configuration guide describes how\nyou can build a page consisting of header, body, footer and page numbers and how you can customize columns. You can also customize the layout of tables.\nIt is also recommended to look at the accessibility guide: This describes how output files can be used and understood by as many people as possible.\nThe guide for LaTeX users is interesting for those who want to switch from LaTeX → Typst markup language.\nHow can I install Typst? # The installation is easiest via package manager:\nLinux: Ubuntu: sudo snap install typst MacOS: brew install typst Windows: winget install --id Typst.Typst ","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/typst/","section":"Blog Posts","summary":"What is Typst? # Typst\nis a markup language (description of document layout in code), which also supports styling, scripting and context a compiler that converts markup input into output formats A service (you can use it, you don’t have to use it), which offers functions for organizations. Why should I use Typst? # Typst is simple # So far, the only real professional alternative to Typst was the markup language LaTeX, which was a real option. Pandoc is a LaTeX compiler often used under the Linux operating system (converts LaTeX-based input data into output formats). Using the two compilers, the properties can be compared in a practical way. Typst is much easier to learn and use compared to LaTeX. I wrote my master’s thesis in 2012 with LaTeX. If I had had a simpler alternative back then, I would have saved a lot of time and nerves.\n","title":"How can I generate high-quality PDFs as simply as possible from Markdown?","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/innovation/","section":"Tags","summary":"","title":"Innovation","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/ki/","section":"Tags","summary":"","title":"KI","type":"tags"},{"content":" Warum sollte ich KI zur Entwicklung von Yocto-basierten Embedded Systems nutzen? # Die Entwicklung von Embedded Systems mit Yocto kann komplex und zeitaufwendig sein.\nDas ist problematisch:\nErheblicher Entwicklungs- und Zeitaufwand fließt in das BitBake-basierte Build-System anstatt in Features, Bugfixes und Security Patches. Gut ausgebildete Yocto-Entwickler sind rar. Das Team-interne Yocto-Ausbildung ist zeitaufwändig. Anforderungen im Bereich der IT Security (Supply Chain Security) binden Entwicklungs- und Zeitresourcen: Security-Patches. Security Requirement Engineering/Management. Threat-Modelling. Priorisierungs-Konflikt zwischen neuen Features und IT Security Best Practices. KI kann:\nDas Arbeiten mit dem BitBake Build-System erheblich vereinfachen und beschleunigen. Agent Skills haben mehrere Nutzen: Sie sind die Grundlage für reproduzierbare und skalierbare Beschleunigung der Entwicklung als auch FOrtbildungsmaterial für interne Yocto-Ausbildung. Wie kann KI die Entwicklung von Yocto-basierten Embedded Systems vereinfachen und beschleunigen? # Automatisierung von Routineaufgaben: KI kann repetitive Aufgaben wie das Erstellen von Rezepten, das Konfigurieren von Build-Umgebungen und das Testen von Images automatisieren. Dies spart Zeit und reduziert Fehler.\nOptimierung von Build-Prozessen: KI kann Build-Prozesse analysieren und optimieren, um die Build-Zeit zu verkürzen und die Effizienz zu steigern.\nFehlererkennung und -behebung: KI kann potenzielle Fehler in Rezepten oder Konfigurationen erkennen und Vorschläge zur Behebung machen.\nWissensmanagement: KI kann als Wissensdatenbank dienen, die Best Practices, häufige Probleme und Lösungen speichert und bereitstellt.\nWas sind Agent Skills? # Agent Skills sind spezielle Fähigkeiten oder Module, die KI-Agenten nutzen, um bestimmte Aufgaben zu erfüllen. Diese Skills können als vorgefertigte Bausteine betrachtet werden, die es ermöglichen, komplexe Prozesse zu automatisieren und zu optimieren. Im Kontext von Yocto-basierten Embedded Systems können Agent Skills beispielsweise verwendet werden, um:\nRezepte zu generieren: Automatische Erstellung von BitBake-Rezepten basierend auf Anforderungen oder Vorlagen. Build-Konfigurationen zu optimieren: Analyse und Anpassung von Build-Konfigurationen, um die Effizienz zu steigern. Fehler zu erkennen und zu beheben: Identifikation von potenziellen Fehlern in Rezepten oder Konfigurationen und Vorschläge zur Behebung. Dokumentation zu generieren: Automatische Erstellung von Dokumentation für Rezepte, Konfigurationen und Build-Prozesse. Agent Skills sind in der Regel modular aufgebaut und können je nach Bedarf kombiniert oder erweitert werden. In anderen Bereichen der nicht-technischen IT werden Agent Skills of Open Source gemeinschaftlich entwickelt.\nWann und warum benötige ich Agent Skills? # Komplexe Build-Prozesse: Wenn Ihre Yocto-basierten Build-Prozesse komplex sind und viele manuelle Schritte erfordern, können Agent Skills helfen, diese Prozesse zu automatisieren und zu vereinfachen.\nWiederkehrende Aufgaben: Wenn Sie häufig wiederkehrende Aufgaben wie das Erstellen von Rezepten oder das Konfigurieren von Build-Umgebungen haben, können Agent Skills diese Aufgaben übernehmen und so Zeit sparen.\nFehleranfällige Prozesse: Wenn bestimmte Prozesse fehleranfällig sind und häufig manuelle Überprüfungen erfordern, können Agent Skills helfen, Fehler frühzeitig zu erkennen und zu beheben.\nWissensmanagement: Wenn Sie Wissen und Best Practices in Ihrem Team teilen und dokumentieren möchten, können Agent Skills als Wissensdatenbank dienen und diese Informationen bereitstellen.\nSkalierung: Wenn Sie Ihre Entwicklungsprozesse skalieren möchten, können Agent Skills helfen, die Effizienz zu steigern und die Qualität zu verbessern.\nReproduzierbarkeit: Agent Skills können die Reproduzierbarkeit von agentenbasierter Entwicklung im Vergleich zu reinem Prompting erhöhen. Durch die Verwendung von Agent Skills können Sie sicherstellen, dass bestimmte Aufgaben immer auf die gleiche Weise ausgeführt werden, was die Konsistenz und Zuverlässigkeit Ihrer Entwicklungsprozesse verbessert.\nGibt es Open Source Agent Skill Bibliotheken für das Yocto-Projekt? # Ja.\nEs gibt seit ungefähr Anfang 2026 Agent Skills für das Yocto-Projekt GitHub (search: agent skills yocto) und BitBake GitHub (search: agent skills bitbake).\nSind Open Source Agent Skill Bibliothekn für das Yocto-Projekt universell einsetzbar? # Nein.\nBei Agent Skills ist im Allgemeinen zu beachten, dass sich SW-basierte Produkte/Projekte immer aus mehreren Aspekten basieren:\nBei Yocto-basierten Betriebsystemen gilt:\nEnthalten generische SW-Komponten: Dazu gehören Beispielsweise SW-Komponenten wie systemd die in allen oder zumindest den meisten Yocto-Betriebsystemen genutzt werden. Integration von generischen SW-Komponenten ist oft HW-spezifisch. System-Anforderungsabhängige, spezifische SW-Komponenten: Ein Yocto-Betriebsystem das Echtzeit-Anforderungen hat wird in der Regel andere Programmiersprachen und Middleware oder Frameworks für die Kommunikation zwischen Komponenten nutzen, als eines ohne Echtzeitanforderungen. Domänen-abhängige, spezifische Systemarchitektur: Embedded Systems unterscheiden sich in der Regel deutlicher von Konkurrenzprodukten, als das in der nicht-technischen IT der Fall ist. Das Betriebsystem eines Unternehmens im Bereich der Unterhaltungselektronik unterscheidet sich fundamental von einem im Bereich de Energiewirtschaft. Wie kostenintensiv und zeitaufwändig ist es eigene Agent Skills zu entwicklen? # Die Entwicklung von Agent Skills ist ein Teilbereich der modernen SW-Entwicklung.\nAgent Skills sollten:\nGereviewt werden. Optimiert werden. Evaluiert werden. Bezüglich IT Security gereviewt/auditiert werden. Gewartet werden: Agent Skills beziehen sich in vielen Fällen auf spezifische Versionen von SW-Komponenten, die konsistent zu neuen Updates gereviewt und ggf. aktualisiert werden sollten. Werden Agent Skills in Firmen eingesetzt? # Ja.\nAus meinem Netzwerk weiß ich, dass Agent Skills für die Entwicklung von Yocto-basierten Betriebsystemen eingesetzt werden. Diese stehen aus diversen Gründen (IT Security, Schutz von Intellectual Property, \u0026hellip;) nicht öffentlich zugänglich zur Verfügung.\nWelche Implikationen hat die Nutzung von Agent Skills? # Agents hängen von einer dynamischen Interaktion mit dem BitBake-Build ab. Das bedeutet, dass die kostengünstigsten Effizienzsteigerungen im Rahmen von lokal ausgeführten BitBake-Builds möglich sind. Agenten-basierte Optimierung von CI-basierten BitBake-Builds ist möglich, aber viel aufwändiger als die Optimierung von lokalen BitBake-Builds. Agenten nutzen Modelle, um Promts zu beantworten. Technisch gesehen können Cloud-Dienstleister, die Modelle bereitstellen (Anthropic, \u0026hellip;) Je nach Firmen-intenen Anfordungen an die IT Security, Intelectual Property und ggf. auch Data Privacy kann es notwendig sein, dass entweder Modelle (werden von Agenten genutzt, um Prompts zu beantworten) entweder lokal gehostet oder im Rahmen der internen IT Infrastruktur bereitgestellt werden. ","date":"07. July 2026","externalUrl":null,"permalink":"/posts/ki-basierte-yocto-entwicklung/","section":"Blog Beiträge","summary":"Warum sollte ich KI zur Entwicklung von Yocto-basierten Embedded Systems nutzen? # Die Entwicklung von Embedded Systems mit Yocto kann komplex und zeitaufwendig sein.\nDas ist problematisch:\nErheblicher Entwicklungs- und Zeitaufwand fließt in das BitBake-basierte Build-System anstatt in Features, Bugfixes und Security Patches. Gut ausgebildete Yocto-Entwickler sind rar. Das Team-interne Yocto-Ausbildung ist zeitaufwändig. Anforderungen im Bereich der IT Security (Supply Chain Security) binden Entwicklungs- und Zeitresourcen: Security-Patches. Security Requirement Engineering/Management. Threat-Modelling. Priorisierungs-Konflikt zwischen neuen Features und IT Security Best Practices. KI kann:\n","title":"KI-basierte Entwicklung von Yocto-basierten Embedded Systems","type":"posts"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/k%C3%BCnstliche-intelligenz/","section":"Categories","summary":"","title":"Künstliche Intelligenz","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/markdown/","section":"Tags","summary":"","title":"Markdown","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/microsoft/","section":"Tags","summary":"","title":"Microsoft","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/%C3%B6ffentliche-verwaltung/","section":"Categories","summary":"","title":"Öffentliche Verwaltung","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/open-source/","section":"Categories","summary":"","title":"Open Source","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/open-source/","section":"Tags","summary":"","title":"Open Source","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/pdf/","section":"Tags","summary":"","title":"PDF","type":"tags"},{"content":" What is Podman? # Podman is an open-source tool for managing containers and pods. It was developed to provide a secure and simple alternative to Docker. Podman is daemonless, which means it does not require any background processes to manage containers. This makes it more secure and resource-efficient.\nPodman is part of the Red Hat ecosystem and is supported by the Open Container Initiative (OCI). It is compatible with Docker and can use many Docker commands directly.\nWhat Use Cases Can I Use Podman For? # Podman is versatile and suitable for a wide range of use cases:\nDevelopment Environments: Podman can be used to create isolated development environments that are independent of the host system. This is particularly useful for developing applications that require specific dependencies or configurations. The open-source Development Containers (devcontainers) specification can be used for this, but it is not mandatory. For example, the Linux operating system Project Bluefin uses Podman and devcontainer.\nCI/CD Pipelines: Podman can be integrated into Continuous Integration and Continuous Deployment (CI/CD) pipelines. On the one hand, container images can be created. On the other hand, container-based systems can be run in CI pipelines. Podman is very popular among DevOps and system administrators. First impressions can be gained from the Red Hat Enterprise Linux YouTube Channel (search: podman).\nMicroservices: Podman is excellent for managing microservices. Each microservice can run in its own container, improving isolation and scalability.\nWeb Servers: On a physical server or virtual server (VPS), Podman can be used to run many web server containers in parallel.\nEmbedded Systems: Podman is suitable for edge computing scenarios where containers need to be run on devices with limited resources. Podman\u0026rsquo;s daemonless architecture makes it ideal for such environments. In many modern operating systems, software components are either directly integrated into the operating system as a \u0026ldquo;compound of containers\u0026rdquo; (Podman Pods, more on this later) or made dynamically updatable. This greatly simplifies the testability of complex features and software component interactions. With good test coverage, regular security patches are no longer a nightmare with unpredictable impacts (regressions).\nWhat Alternatives to Podman Are There? # There are several alternatives to Podman that can also be used for container management:\nDocker: The most well-known container platform, but it requires a daemon and is therefore less secure. More information can be found on the Docker website. LXC/LXD: Linux containers that provide a complete operating system environment. More information can be found on the LXC website. rkt: Another container runtime developed by CoreOS. More information can be found on the rkt website. CRI-O: A container runtime specifically developed for Kubernetes. More information can be found on the CRI-O website. Why Should I Use Podman and Not an Alternative? # Podman offers several advantages over other container management tools:\nDaemonless: Podman does not require a background process, making it more secure and resource-efficient. Rootless: Podman can run without root privileges, further enhancing security. Compatibility: Podman is compatible with Docker and can use many Docker commands directly. Integration: Podman integrates well with existing systems and workflows. Security: Podman provides a secure environment for running containers. Is Podman Compatible with Docker Compose Files? # Yes, Podman is compatible with Docker Compose files. Podman can use Docker Compose files directly to manage containers and pods. This makes it easy to migrate existing Docker Compose files to Podman.\nWhat Are the Differences Between Podman and Docker? # Podman and Docker are both tools for managing containers, but there are some important differences:\nArchitecture: Podman is daemonless, while Docker requires a daemon. This makes Podman more secure and resource-efficient. Root Privileges: Podman can run without root privileges, while Docker requires root privileges. Integration: Podman integrates better with systemd and other system services (Quadlets, see below), while Docker requires its own infrastructure. Compatibility: Podman is compatible with Docker and can use many Docker commands directly, but there are some differences in implementation. Security: Podman provides a secure environment for running containers, while Docker is more vulnerable to security issues due to the daemon. What Are Quadlets? # Quadlets are a new feature in Podman that improves integration with systemd. Quadlets allow containers to be managed as systemd services, simplifying the management and monitoring of containers.\nAdvantages of Quadlets:\nEasy Integration: Quadlets enable seamless integration of Podman containers with systemd, simplifying container management and monitoring. Automatic Start: Containers can be automatically started at system boot, improving service availability. Monitoring and Logging: Quadlets use systemd\u0026rsquo;s monitoring and logging functions, simplifying troubleshooting and diagnostics. Easy Configuration: Quadlet configuration is simple and intuitive, making container setup and management easier. What Are Podman Pods? # Podman Pods are a group of one or more containers that are managed together. They allow multiple containers to be treated as a single unit, simplifying the management and orchestration of applications. Pods are particularly useful when multiple containers need to work closely together, such as in microservice architectures.\nShared Management: Pods allow multiple containers to be managed as a single unit, simplifying application management and orchestration. Shared Network Environment: Containers in a pod share a common network environment, facilitating communication between containers. Shared Storage Environment: Containers in a pod can share a common storage environment, facilitating data exchange between containers. Easy Scaling: Pods can be easily scaled by adding more containers to the pod. How Do Podman Pods and Kubernetes Pods Differ? # Podman Pods and Kubernetes Pods are similar, but there are some important differences:\nManagement: Podman Pods are managed directly by Podman, while Kubernetes Pods are managed by a Kubernetes cluster. Scaling: Kubernetes offers advanced features for scaling and managing containers that Podman does not. Orchestration: Kubernetes provides a complete orchestration platform, while Podman is more suited for managing individual containers and pods. Complexity: Kubernetes is more complex and requires more setup effort, while Podman is simpler to use. Podman is an excellent choice for developers and system administrators looking for a secure and simple way to manage containers. It offers a good balance between simplicity and functionality and is a strong alternative to Docker and other container management tools.\nIs Podman Compatible with Kubernetes Manifest Files? # Yes.\nPodman is compatible with Kubernetes manifest files. Podman can use Kubernetes manifest files directly to manage containers and pods. This makes it easy to migrate existing Kubernetes manifest files to Podman.\nBut\u0026hellip;\nKubernetes manifest files generated with Podman do not consider many details that are relevant in a Kubernetes deployment.\nWhat is Podman Machine? # Podman Machine is a tool that allows Podman to be used on macOS and Windows by creating a virtual machine (VM) that runs Podman. This is particularly useful since Podman was originally developed for Linux and does not run natively on other operating systems.\nWhen using Podman on Linux, you can avoid dealing with Machine. This is the case for most use cases of Podman.\n","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/podman/","section":"Blog Posts","summary":"What is Podman? # Podman is an open-source tool for managing containers and pods. It was developed to provide a secure and simple alternative to Docker. Podman is daemonless, which means it does not require any background processes to manage containers. This makes it more secure and resource-efficient.\nPodman is part of the Red Hat ecosystem and is supported by the Open Container Initiative (OCI). It is compatible with Docker and can use many Docker commands directly.\n","title":"Podman","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/public-administration/","section":"Categories","summary":"","title":"Public Administration","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/software-development/","section":"Tags","summary":"","title":"Software Development","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/software-engineering/","section":"Categories","summary":"","title":"Software Engineering","type":"categories"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/softwareentwicklung/","section":"Tags","summary":"","title":"Softwareentwicklung","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/sport/","section":"Categories","summary":"","title":"Sport","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/sports/","section":"Categories","summary":"","title":"Sports","type":"categories"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/strategie/","section":"Tags","summary":"","title":"Strategie","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/strategy/","section":"Tags","summary":"","title":"Strategy","type":"tags"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/","section":"Tags","summary":"","title":"Tags","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/categories/technologie/","section":"Categories","summary":"","title":"Technologie","type":"categories"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/categories/technology/","section":"Categories","summary":"","title":"Technology","type":"categories"},{"content":" What Are Proteins? # Proteins (also known as \u0026ldquo;proteins\u0026rdquo; from the Greek \u0026ldquo;proteos\u0026rdquo;, meaning \u0026ldquo;first place\u0026rdquo;) are structural building blocks made of amino acids that are found in every cell from hair to muscles and the brain.\nWhy Should We Consume Proteins? # They are primarily important for maintaining muscle mass, muscle growth, and muscle repair processes. But not only that: they are the structures of our organs, such as the heart. They are also building blocks of enzymes. Some of our most important metabolic hormones (\u0026ldquo;peptide hormones\u0026rdquo;) are also made of amino acids, including hormones like insulin, glucagon, and the insulin-like growth factor IGF-1. Since the body can only produce non-essential amino acids, but we also need essential amino acids, we must obtain these through our diet. It is important that we eat proteins that contain all nine essential amino acids (histidine, phenylalanine, isoleucine, lysine, valine, methionine, leucine, threonine, tryptophan).\nWhere Are Essential Amino Acids Found? # Essential amino acids are primarily found in animal foods such as beef, poultry, eggs, milk, or fish. Plant-based sources include hemp seeds, soy products, legumes, pumpkin seeds, peanuts, or quinoa.\nWhat Is Biological Value? # It is an evaluation system for classifying proteins and assesses how effectively the body utilizes a dietary protein. The body can particularly well utilize a dietary protein if it has a high content of essential amino acids. The biological value of whole egg (egg yolk plus egg white) is set at 100, and all other foods are measured against this standard.\nAre Animal Proteins Higher Quality Than Plant Proteins? # Yes, since the biological value of plant proteins is slightly lower. However, by combining different plant proteins, the amino acid spectrum can be significantly enhanced. This makes it possible to achieve biological values of over 100 through the intake of plant-based proteins (e.g., peas with rice, beans with corn, or potatoes with eggs). Importantly, the combination of proteins does not necessarily have to occur within the same meal.\nDoes Volksshake Powder Contain All Nine Essential Amino Acids? # Yes. A mixture of up to ten plant proteins ensures that all nine essential amino acids are included.\nWhat Is the Biological Value of Volksshake Powder? # The biological value of Volksshake powder is between 90 and 100. For comparison: cheese, tuna, beef, or rye have a biological value of 83, so this value can be considered very high.\nHow Much Protein Should I Consume? # Your requirement depends on your age, health condition, activity level, diet, and other individual factors. The data on this is not entirely clear. International recommendations range from 0.8g/kg body weight to up to 3.1g/kg body weight in extreme strength sports. According to the German Nutrition Society (DGE), it should be at least 0.8g/kg body weight (minimum intake). People over 65 years are recommended 1.0g/kg body weight by the DGE, and athletes up to 1.6g/kg body weight. Pregnant and breastfeeding women are recommended 0.9-1.2g/kg body weight by the DGE. For active people who train regularly, a requirement of 1.5g/kg body weight can be assumed. For example, with a body weight of 70kg, you would have a daily requirement of just over 100g of protein.\nSource: Ernaehrungsradar.de - Proteine: Versprechen, Funktionen, Fakten\nWhat Do I Need to Eat in a Day to Get a Total of About 100g of Protein? # 200g tofu (24g protein) 200g oatmeal (28g protein) 100g salmon fillet (25g protein) 200g low-fat quark (22g protein) 150g yogurt (6g protein) Are Plant Proteins Healthier Than Animal Proteins? # In the very large-scale Nurses Health Study from the USA, nutrition experts at Harvard University found that the intake of plant proteins in middle age was associated with a lower occurrence of certain chronic diseases. Advantages of plant protein sources include a usually higher fiber content, lower calorie density, and less saturated fatty acids. For example, 100g of cooked red lentils provide 10.4g protein, 0.7g fat, 18g carbohydrates, and 7.6g fiber with 138 total calories. In comparison, 100g of cooked eggs provide 12.6g protein, 10.6g fat, and 1.1g carbohydrates with 155 total calories.\nWhat Happens If I Don\u0026rsquo;t Eat Enough Protein? Does It Affect My Performance? # A deficiency of amino acids inhibits the effect and production of proteins and also the peptide hormones involved in numerous metabolic processes. This affects the immune system, performance, joints, and muscle building. Possible symptoms include decreased performance, fatigue, sleep problems, mood swings, muscle weakness, thyroid problems, and blood sugar fluctuations. Your physical and mental performance can therefore be significantly impaired by insufficient protein intake.\nWhat Are the Benefits of Volksshake Protein Powder? # Volksshake protein powder is:\nHealthy - because it is purely plant-based, contains all essential amino acids, and has a high biological value. It comes without added sugar and colorings. The organic varieties do not even contain natural sweeteners. Delicious - because it is available in many flavors and can enhance the taste of your meals. Practical - because you can consume it in many forms and use it, for example, during travel or on vacation to meet your protein needs. Well-tolerated - compared to legumes, which are poorly tolerated by many people due to their high fiber content, Volksshake protein powder is well-tolerated. Soy-, gluten-, and lactose-free - and therefore also suitable for people with intolerances or allergies to these substances. Affordable compared to competing products and some other protein sources. A high-quality regional product - because it comes from a family business in Baden-Württemberg and is produced there under regular voluntary laboratory controls. Volksshake guarantees freedom from genetic engineering and pesticides. What Is Included in Volksshake Powder? # Depending on the variety, it is a mixture of the following ten proteins:\nPea protein Pumpkin protein Rice protein Hemp protein Almond protein Sunflower protein Quinoa protein Fava bean protein Chia seed protein Flaxseed protein In addition, there are natural flavors, Ceylon cinnamon bark powder, and the thickening agent guar gum.\nHow Expensive Is Volksshake Protein Powder Compared to Competing Products? # At 27.99 euros/kg, Volksshake is in the lowest price segment of protein powders. Most competing products, which often contain artificial additives and sugar or sugar substitutes, range from 30-60 euros/kg.\nHow Expensive Is Volksshake Protein Powder Compared to Other Protein Sources? # For this, we set the price of 100g of protein from Volksshake powder:\n1kg Volksshake powder contains 750g protein and costs just under 28 euros. 100g protein from Volksshake powder therefore costs 3.70 euros. For comparison, eggs:\n1 egg contains about 7g protein and costs 0.40 euros. 100g protein is contained in 14 eggs and costs 5.68 euros. For comparison, tofu:\n100g tofu contains 12g protein and costs one euro. 100g protein is contained in about 800g tofu and costs 8 euros. For comparison, low-fat quark:\n100g low-fat quark contains 12g protein, 500g low-fat quark costs about one euro. 100g protein is contained in about 800g low-fat quark and costs 1.67 euros. For comparison, beef:\n100g beef contains 18g protein. 1kg beef costs about 15 euros. 100g protein is contained in about 600g beef and costs 9 euros. For comparison, red lentils:\n100g red lentils (uncooked) contain 21g protein. 1 kg red lentils costs about 3 euros. 100g protein is contained in about 500g lentils (but there are some losses due to the cooking process) and costs 1.50 euros. How Can I Consume Volksshake? # Volksshake powder can be enjoyed as a delicious shake, or you can stir it into your morning muesli, mix the powder with yogurt or quark, or mix it into all kinds of sauces.\nWhat Healthy Recipes Are There with Volksshake? # Breakfast porridge: Pour hot water over 50g oatmeal, 20g flaxseeds, 20g wheat bran, and 15g Volksshake protein powder of your choice, let it soak briefly. Add nuts of your choice, turmeric, and cinnamon powder, and finish with a pinch of pepper. Fresh seasonal fruit goes well with this. Alternatively, you can prepare the porridge for a simple lunch or dinner: use the neutral organic protein powder and instead of nuts, fruit, and cinnamon, add vegetable broth and blanched vegetables, and season with your favorite herbs. Quick ice cream: Put 150ml milk, 15g Volksshake protein powder of your choice, 150g ice cubes, 150g frozen fruit into the blender and blend for 3 minutes at medium speed. Who Wrote This Article? # To be completely honest\u0026hellip; it wasn\u0026rsquo;t me: I am fortunate to live with a very competent partner. Thank you for all the tips you have already given me and for writing this article!!!\nWhere Can I Buy Volksshake? # You can purchase Volksshake through online retail at volksshake.de.\n","date":"July 7, 2026","externalUrl":null,"permalink":"/en/posts/volksshake/","section":"Blog Posts","summary":"What Are Proteins? # Proteins (also known as “proteins” from the Greek “proteos”, meaning “first place”) are structural building blocks made of amino acids that are found in every cell from hair to muscles and the brain.\nWhy Should We Consume Proteins? # They are primarily important for maintaining muscle mass, muscle growth, and muscle repair processes. But not only that: they are the structures of our organs, such as the heart. They are also building blocks of enzymes. Some of our most important metabolic hormones (“peptide hormones”) are also made of amino acids, including hormones like insulin, glucagon, and the insulin-like growth factor IGF-1. Since the body can only produce non-essential amino acids, but we also need essential amino acids, we must obtain these through our diet. It is important that we eat proteins that contain all nine essential amino acids (histidine, phenylalanine, isoleucine, lysine, valine, methionine, leucine, threonine, tryptophan).\n","title":"Volksshake","type":"posts"},{"content":"","date":"July 7, 2026","externalUrl":null,"permalink":"/en/tags/yocto/","section":"Tags","summary":"","title":"Yocto","type":"tags"},{"content":"","date":"07. July 2026","externalUrl":null,"permalink":"/tags/zukunft/","section":"Tags","summary":"","title":"Zukunft","type":"tags"},{"content":"\u0026ldquo;If you want to go fast, go alone. If you want to go far, go together.\u0026rdquo; — African proverb\nI have been an active club member in sports and outdoor organizations since my youth.\n2025–Present: SSC Jachenau e.V. # Active membership. Ski club with nordic and alpine skiing, and mountain biking division.\n2005–Present: Deutscher Alpenverein (DAV) - German Alpine Club # Active membership in various regional sections of the German Alpine Club.\n2008–2011: Enshin Dojo Freiburg, Mu-Shin e.V. # Active membership. Full-contact Enshin Karate and Muay Thai boxing.\n2006–2007: University Sports Shotokan Karate # Active membership. Semi-contact karate in the German Karate Federation.\n2005–2006: University Sports Taekwondo Division # Active membership. Semi-contact Taekwondo in WAKO BWKBV.\n2005–2006: University Sports Krav Maga # Active membership. Self-defense training.\n1998–2003: Football Club # Active membership in football club from C-youth to A-youth teams.\n","externalUrl":null,"permalink":"/en/vereinsmitgliedschaften/","section":"Florian Kromer","summary":"“If you want to go fast, go alone. If you want to go far, go together.” — African proverb\nI have been an active club member in sports and outdoor organizations since my youth.\n2025–Present: SSC Jachenau e.V. # Active membership. Ski club with nordic and alpine skiing, and mountain biking division.\n2005–Present: Deutscher Alpenverein (DAV) - German Alpine Club # Active membership in various regional sections of the German Alpine Club.\n","title":"Club Memberships","type":"page"},{"content":"Disclosure according to the German \u0026ldquo;Digitale-Dienst-Gesetz\u0026rdquo; (DDG) § 5\nName: Florian Kromer\nAddress: Laich 12 1/5, 83676 Jachenau\nContact: +49 1515 489 25 92, kromer-florian@protonmail.com\nSales tax identification number: According to paragraph 19 UStG no sales tax is charged.\n","externalUrl":null,"permalink":"/en/impressum/","section":"Florian Kromer","summary":"Disclosure according to the German “Digitale-Dienst-Gesetz” (DDG) § 5\nName: Florian Kromer\nAddress: Laich 12 1/5, 83676 Jachenau\nContact: +49 1515 489 25 92, kromer-florian@protonmail.com\nSales tax identification number: According to paragraph 19 UStG no sales tax is charged.\n","title":"Imprint","type":"page"},{"content":"This website does not process or store any personal data from you except what is technically necessary.\nHosting is provided by the European provider Hetzner on European infrastructure.\nThe website uses privacy-friendly web analytics provided by Hetzner.\nThis Webseite hyperlinks Steady. Data is processed and stored in compliance with European data privacy guidelines: steady.page/de/eu, steady.page/de/privacy.\n","externalUrl":null,"permalink":"/en/datenschutz/","section":"Florian Kromer","summary":"This website does not process or store any personal data from you except what is technically necessary.\nHosting is provided by the European provider Hetzner on European infrastructure.\nThe website uses privacy-friendly web analytics provided by Hetzner.\nThis Webseite hyperlinks Steady. Data is processed and stored in compliance with European data privacy guidelines: steady.page/de/eu, steady.page/de/privacy.\n","title":"Privacy Policy","type":"page"},{"content":" Professional Experience # I have worked in various types of companies (startups, medium-sized companies, corporations) in many roles and have technologically steered projects. This includes:\nBuilding data-oriented teams (defining role requirement profiles, participating in recruiting interviews) Building Embedded SW teams (defining role requirement profiles, participating in recruiting interviews) Technological steering of projects in different company sizes and cultures September 2012–present: SW Architecture and SW Development # In my professional career, I have worked in various roles:\nEmbedded SW Developer: Development of embedded systems and real-time applications Embedded SW Architect: Design and architecture of embedded systems Backend Developer: Development of server-side systems and APIs My experience includes working in:\nStartups: Agile development and rapid iterations Medium-sized companies: Practical solutions with limited resources Large corporations: Scalable systems and enterprise architectures Technological Leadership # Architecture decisions in complex projects Technology selection and evaluation Mentoring development teams Quality assurance and best practices Team Building and Recruiting # Defining role profiles for development teams Participating in job interviews and selection processes Building and leading Embedded SW teams Promoting data-oriented ways of working ","externalUrl":null,"permalink":"/en/beruflich/","section":"Florian Kromer","summary":"Professional Experience # I have worked in various types of companies (startups, medium-sized companies, corporations) in many roles and have technologically steered projects. This includes:\nBuilding data-oriented teams (defining role requirement profiles, participating in recruiting interviews) Building Embedded SW teams (defining role requirement profiles, participating in recruiting interviews) Technological steering of projects in different company sizes and cultures September 2012–present: SW Architecture and SW Development # In my professional career, I have worked in various roles:\n","title":"Professional Activity","type":"page"},{"content":"\u0026ldquo;Competence obligates. Whoever can do something, should do it.\u0026rdquo; — Quote of a German Politician\nSports Volunteering # 2026–present: SSC Jachenau e.V. # Mountain bike trainer for children and youth. Teaching basic techniques, safety, and nature experience in mountain biking.\n2010–2011: Enshin Dojo Freiburg, Mu-Shin e.V. # Karate trainer for children and adults. Leading training sessions in Enshin Karate and Muay Thai boxing for different age and skill groups.\nOpen Knowledge/Source # 2014–present: Open Source Contributions (GitHub) # Contributing to open source projects at github.com/fkromer, primarily focused on software architecture and IT security topics.\n2026: EuroStack.eu Initiative # Volunteering with the EuroStack Initiative Foundation e.V. (non-profit organization) researching and suggesting open source software and European services to the European Catalogue Project.\n2026: Center for Digital Sovereignty in Public Administration (ZenDiS) # Voluntary participation in forum discussions and public reviews on software architecture topics.\n","externalUrl":null,"permalink":"/en/ehrenamt/","section":"Florian Kromer","summary":"“Competence obligates. Whoever can do something, should do it.” — Quote of a German Politician\nSports Volunteering # 2026–present: SSC Jachenau e.V. # Mountain bike trainer for children and youth. Teaching basic techniques, safety, and nature experience in mountain biking.\n2010–2011: Enshin Dojo Freiburg, Mu-Shin e.V. # Karate trainer for children and adults. Leading training sessions in Enshin Karate and Muay Thai boxing for different age and skill groups.\n","title":"Volunteer Work","type":"page"}]