enclose.horse

Original Article ↗ Hacker News Discussion ↗

Gameplay & Overall Reception

  • Widely praised as a simple, clever, very fun puzzle; some call it “Wordle-like” (one daily puzzle, shared results) and compare it to Rodent’s Revenge, Chat Noir, JezzBall, ChuChu Rocket, Pathery, etc.
  • Core goal clarified for confused players: place limited walls to enclose the horse; score is area reachable by the horse (cherries add extra value).
  • Several people report the game is more challenging than it looks but satisfying to reason about and to reach “Perfect” solutions.

UI, Visuals, and Interaction

  • Frequent complaints that 3D walls visually overlap two tiles, hide cherries, and feel inconsistent with the rest of the 2D look; many request flat, one-tile fences.
  • Some find the constant “wiggle” animations distracting and ask for a way to disable them.
  • Mobile and small-screen users report mis-taps and want larger hitboxes; Firefox users note right-click issues, later learning left-click removes walls.
  • Feature requests: clearer wall-count font (1 vs I), better indication of “Best” and an easy restore button (which already exists but isn’t obvious).

Daily Format & Replayability

  • One-submit-per-day design sparks debate:
    • Supporters liken it to crosswords/Wordle: a bounded daily ritual that discourages bingeing.
    • Critics dislike artificial scarcity, calling it limiting or “condescending,” preferring unlimited levels.
  • Clarifications: you can freely experiment before submitting; can replay previous dailies and many user levels via the menu; some ask for a stronger warning about single submission and for synchronized global release times.

Comparing & Learning from Solutions

  • Many want an easy toggle or side‑by‑side view between their solution and the optimal one; currently “View Optimal” appears after submission, but returning to your own layout is clumsy.
  • Mixed views on showing max/target score or “distance from optimal”:
    • Pro: encourages deeper optimization.
    • Con: enables brute force and changes the character of the puzzle.

Algorithms, Solvers & Theory

  • Lively thread on how to compute optimal solutions for larger grids: mentions constraint programming, CP‑SAT, SAT/SMT, Answer Set Programming with clingo, ILP, graph cuts, NP-hardness, and flood‑fill constraints.
  • Several external solvers are built from screenshots using image analysis + MILP/constraint solving; others discuss LLM-based solvers and pitfalls like “horseless pockets.”

Analytics & Ethics

  • Some hope the author collects gameplay analytics to order levels by difficulty or build a commercial bundle.
  • Others strongly resist any hidden tracking, arguing for explicit opt‑in and distinguishing benign gameplay metrics from invasive surveillance.