Writing

Technical articles, insights, and lessons learned from building software systems. I write about architecture decisions, engineering practices, and the challenges of delivering reliable software at scale.

139 articles

August 29, 2026 Reverse-Engineering omp.sh's Dithered Background How omp.sh draws its pink dithered planet with a 2D canvas and no WebGL, worked out from the minified bundle and rebuilt as five variations.

procedural-generationjavascriptfrontend

August 28, 2026 Self-hosting a Linear or Jira Replacement: Plane and Huly on 8 GB The two open-source issue trackers worth running in place of Linear or Jira. Huly's own README asks for 8 GB minimum, and the box had 3.6 GB free.

self-hostinglinearjiraproject-managementdockerarm64hetznerinfrastructure

Site BlockerAugust 28, 2026 Simple custom website blocker for Chrome A Chrome extension that blocks sites for eight hours, a day, a week, or permanently, with no way to lift a block early from the popup.

javascripttoolingclaude-code

August 27, 2026 Building with vgpu, Vercel's WebGPU Library I pulled four shader demos out of vgpu's example gallery and shipped them as live sites. One of them took five rounds of debugging to run smoothly.

webgpuwgslvgpuvercel

Token Efficiency: Where the Tokens GoAugust 27, 2026 Comparing graft and jCodeMunch Two MCP servers that index a repo with tree-sitter. I measured both on the same code: 886 tokens of tool schema against 27,526.

llmtoolingmcpclaude-codetoken-efficiencygraftjcodemunch

AI Coworkers: Not Just Chatbots or AgentsAugust 27, 2026 Headlong vs ai-coworkers: What Should an Agent Do When Nothing Is Happening? Laude's Headlong thinks continuously and costs $1-2 an hour at idle. My runtime returns before the model is called. One design question separates them.

llmagentsarchitecturetooling

August 26, 2026 Self-Hosting cal.diy on an arm64 Box The image its README advertises has zero tags. Building it took nine minutes and shipping it would have taken two hours, so the build moved to CI.

self-hostingdockerarm64hetznerinfrastructure

Token Efficiency: Where the Tokens GoAugust 25, 2026 Referencing a skill instead of installing it A paper argues installed skills spend a scarce prompt budget. I measured mine at 2,225 resident tokens, then ran the @skills reference client.

agentstoolingllm

August 24, 2026 Where Agent Behaviour Lives: dsh, Claude Code and pi DeepSeek's dsh keeps its hardcoded prompt to one sentence. Claude Code puts orchestration in prose. pi hands the choice to you. I read the source of all three.

llmagentstoolingdeepseekpi

ilo: A Programming Language for AI Agents, Not HumansAugust 23, 2026 Four Places to Catch a Model's Mistake BAML repairs the output, Hyperlambda gates the execution, Zero explains the failure, ilo shrinks the surface. Four languages for agents, four different moments to intervene.

language-designilollmagents