Update: March 2026
I've been making progress with lots of different kinds of puzzles this month. For the classics, I finally found the time recently to buy a puzzle game about a week and a half after its release; I've been spending so much time solving it that I've already unlocked part of the postgame. For other games, I've reached the two-thirds point in my latest pack and I've recently been invited to help design puzzles for a small project. For mysteries, I figured out on a surface level what happened to the Kudamono editor: I'm pretty sure it's just being migrated to a new domain, and I'm feeling optimistic that nothing is truly lost. For metapuzzles, I continued work on coding my puzzle solver to the extent that it can solve the example Norinori puzzle in a reasonable amount of time. Currently, my largest stumbling block is trying to figure out how to represent whether two specific cells are orthogonally connected as a Boolean expression of whether specific cells conduct or not; this would unlock loops and area counters beyond dominoes.
I set the following puzzles to use for testing my solver. Naturally, it hits a wall with average-sized puzzles, but I suspect I can add optimizations that pursue promising leads.
https://pzprxs.vercel.app/p?batten/5/4/021412i23i
https://pzprxs.vercel.app/p?tawa/4/5/3/b2b5f32h
Comments
Post a Comment