NAME
wada -- Software Engineer
SYNOPSIS
DESCRIPTION
I try out ideas by actually building them, and write up how it went — including the ones that did not work.
TOPICS
- personal-projects — Things I built on a whim, such as taskdog, houritsu and ttymap
- local-llm — Models on an RTX 5090, compared on real work
- home-server — A NixOS home server, and what I found when it broke
- home-assistant — Wiring appliances and sensors into home automation
RECENT
- 2026-11-07 Reading Japanese Law Amendments in houritsu Without Knowing Git
- 2026-11-04 I Had Claude Code Check My Waybar Fix: AT-SPI Could See the Button but Not Press It
- 2026-11-01 Why My ESF-551 Scale Kept Disconnecting from Home Assistant: Japan's Measurement Act
- 2026-11-01 I Turned Every Japanese Law into Git History
- 2026-10-03 What Separates AI Users Is How Fast the Feedback Loop Turns ── I Was Burning 30% of My Company's Tokens, So Here Is What I Do
- 2026-10-01 Letting tev1, a 4B Decision Model, Play a Game
- 2026-08-22 I Built Four AI Automations and Deleted Them All in 7 Hours, 6 Days, 11 Days and 12 Days — the Survivor Is the One That Makes No Decisions
- 2026-08-15 Qwen3.8 27B Measured on an RTX 5090 — and Why vLLM Didn't Work
- 2026-08-15 Only the Design of a Local LLM Eval Harness Survives — Rebuilding It for Qwen3.8 27B
- 2026-08-04 What Should We Call the Inverse of a Meat Proxy?