I let this site go quiet. Not dramatically — just the usual “I’ll write when I have something polished,” which is how blogs die.
So this is me restarting it on purpose.
Why revive it now#
A few things stacked up:
- I still make stuff — Godot experiments, half-finished game ideas, tooling setups that eat weekends. If it isn’t written down, it evaporates into chat logs and private notes.
- Agent tooling got interesting — OpenCode-style agents, model CLIs, setup adventures that deserve field notes rather than hype threads.
- The stack still works — Hugo + Blowfish on GitLab Pages with jbenda.de is fine. I didn’t need a rewrite. I needed content and a clearer structure.
I bumped the site stack, cleaned leftover scaffold junk, and added an About page plus a Projects section so this isn’t only a reverse-chronological post list with a profile picture.
What this site is (and isn’t)#
Is: a public lab notebook. Games, tools, process, occasional life-adjacent tech.
Isn’t: a magazine, a personal brand machine, or a place that only publishes when every sentence is perfect.
Older posts stay. They’re history — Space Trader, engine rambling, the Hugo pluralization rabbit hole. New posts will mix gamedev journals with tooling field notes.
What I plan to write next#
Rough queue, not a contract:
- OpenCode / agent setup adventure (environment hell, day 1 vs day 7)
- Notes on other CLIs and agent frameworks as I actually try them
- Traffic ticket game: player fantasy, catharsis, and how far “punish bad drivers” should go
- MiningBoy-style clone: vertical slice goals before scope explosion
Project stubs already live under Projects so status can change without waiting for a full essay.
Cadence#
Target: about one post every week or two. If I only have twenty minutes, a short note still counts. Drafts can be imperfect; shipping beats silence.
If you’re reading this because you found an old post — hi, the site is alive again. If you’re me from three months from now: don’t wait for polish.
