Tailscale VPN: Connect Your Own Services on Your Own Infrastructure
AI . Network . Personal . Security . ToolingProxmox in the data center, a laptop in the home office, a Raspberry Pi at the edge of a field, and a DGX Spark for inference: your own hardware is rarely in one place. NAT, firewalls, and dynamic IPs separate the devices. Manual WireGuard means: exchanging keys, maintaining IPs, updating firewall rules. Tailscale builds on
Team Setups in Larger Projects: Delivering Across Team Boundaries
PersonalModern features no longer fit into a single team. That has become normal. An IoT device sends events. An MQTT service forwards them. A mobile app displays them. Add deployment on AWS, a database schema, push notifications, UX for edge cases, IT security for authentication and privacy, and operations for three environments. That is five
Token Governance for LLMs: Why Shared Budgets Fail (and How to Do It Better)
AI . PersonalYour team shares a GitHub Enterprise account. Copilot is enabled for everyone. A shared budget, no individual limits, no visibility into who consumes what: “it’s in the budget.” Then, in mid-June 2026, GitHub switches its billing from Premium Request Units to token-based billing. Suddenly the budget is no longer “unlimited”. The error message arrives faster
Running gokrazy on a Banana Pi BPI-R1
CLI . Go . Hacking . TestingMotivation I had a BPI-R1 sitting in a drawer. The board is from 2014 — only ancient OpenWrt and outdated Armbian images exist for it. Nobody maintains this platform anymore. The hardware is fine though: dual-core Allwinner A20, 1GB RAM, SATA, and a BCM53125 5-port Gigabit switch on the board. Not bad for a dedicated
T-shaped Skillsets – why the most valuable people don’t fit into a box
PersonalI never wanted to lock myself into just one area. Not because depth doesn’t matter to me – but because I’ve always been interested in topics beyond my core field. A while ago, I came across the term T-shaped. And I immediately felt: that describes it exactly. Most job postings still read like shopping lists: five
Those Who Ask Questions Drive Quality
PersonalI work with a colleague who finds something in almost every code review or text proofread. Almost every time. At first, that was exhausting – not because the remarks were wrong, but because they cost time. Time I hadn’t planned for. In hindsight, that was exactly the point. Many of the detailed improvements turned out
Democracy Requires Attention – Which We’ve Long Since Sold.
Personal8 seconds. That’s supposedly the average human attention span today – shorter than that of a goldfish. The number is largely made up. That the attention span has noticeably decreased, less so. This has direct consequences for democracy. Democracy is not a passive state – it requires that people weigh complex issues, evaluate sources, tolerate contradicting
Prompt Engineering Is Dead. Long Live Context Engineering.
AI . PersonalI stopped optimizing my prompts. Not because it doesn’t work – but because it’s the wrong question. The right question is: What’s in the context when the agent starts working? The Problem with Long Conversations Anyone who works regularly with LLMs knows the pattern: at the beginning of a chat, everything runs smoothly. After twenty,
Why Writing Still Matters – And Is Becoming More Important
Personal“I’ll just have an LLM generate that real quick.” I hear this sentence daily by now. And I get it – I use AI myself, have code generated, run my own [[02-Projekte/OpenClaw/Übersicht|OpenClaw]] node, have images generated. That’s not the problem. The problem is what gets lost in the process. Writing Is Thinking – Not Just
Philosophical Exchange with an AI – Part 1
PersonalIn this series, I want to test an approach to discussing philosophical and social topics with an AI. My goal is not only to reflect on big questions, but also to use AI to find facts that can either support or challenge my own ideas and observations. This post is part 1 of an ongoing