I've been hermitting up in the mountains for too long. It's time to go talk to people and get excited about things again. I'm bored and I don't know what the world is up to.
I'm going to be in NYC all next week, and maybe SF some time in July. I'm open to talking to anyone about anything.
Come the fall I want to be working on hard on something. Maybe I'll get a job, maybe I'll do a lot of consulting, maybe something else. I'm very open to suggestions.
I've also been talking to a friend about working on reinforcement learning environments. I'd love to talk to anyone who knows anything about that.
I'll get a proper devlog out eventually, but some teasers:
- Continued ruminations along the veins of borrow-checking without type-checking. I've sketched out a fully statically-typed approach to staging instead, which would allow the borrow-checker to be richer. I think I might be able to recover ergonomics for live reloading, repls etc through having very low syntactic-overhead reflection on an
anytype. - I'm rewriting focus from scratch. Short-term goal is just architectural improvements (eg I have deterministic simulation testing now so I can do end-to-end fuzzing and snapshot tests). But I'm also trying to make the editing model rich enough to build tools like magit while still being a small sans-io rust library instead of a lispy operating system. Definitely inspired in part by abont.
- I've been working on a boardgame with a friend. I'm trying to bring the creativity and rule-interaction-heavy gameplay from games like Slay The Spire and Baba Is You into a multiplayer board game, while still keeping it simple enough to play (correctly) by hand. It's already a pretty fun game, even though there's still a lot of work to do. We also have some (novel?) ideas around using superhuman bots playing millions of games against each other and using the game stats to assess balance and counterplay. Kind of a CI for fun-ness.