Show HN: Hallucinate – Massively Multiplayer Online Rave

Original Article ↗ Hacker News Discussion ↗

Project & Implementation

  • Web-based “massively multiplayer online rave” where users control low-poly dancers in shared rooms with YouTube DJ sets.
  • Source code is MIT-licensed on GitHub; initial repo lacked README/screenshots but they were added after feedback.
  • Built in TypeScript with many flat src/*.ts files; shaders written directly for performance after rejecting Three.js as too slow.
  • Animation system and many optimizations reportedly generated/refined with an advanced GPT model.

Code Quality & Portfolio Concerns

  • Some commenters see it as a strong creative portfolio piece.
  • Others flag “orange/red flags”: flat project structure, weak commit messages, pervasive magic numbers (especially in shaders).
  • Debate over whether such criticism should be public (helpful feedback vs “poisoning” the thread).

Performance, Controls & UX

  • Server hit scaling issues (“HN hug of death”); intermittent crashes and timeouts reported.
  • Developer targets hundreds of concurrent users using client-authoritative movement, dead-reckoning, and syncing only state changes.
  • Movement initially on IJKL; later WASD added as default with layout toggle. Jump (“b for bounce”) implemented after feedback.
  • Various UX critiques: desire for FPS-like strafing, different jump physics, reset-from-stuck, crosshair for motion sickness, different camera/pan behavior.
  • Mobile initially nonfunctional; later updated to support tap-to-move and clickable controls.

Sync, Music & MMO Feel

  • Currently each user manually starts YouTube video; rooms are not globally time-synced.
  • Some argue per-user start preserves full-set experience; others say lack of sync undermines shared-rave feel.
  • Plan mentioned to eventually sync room playback; also interest in accurate player counts and global chat.

Moderation & Safety

  • Multiple reports of racist slurs and toxicity; some say this “fouled” the experience.
  • Current mitigations: word filters (users evade), manual IP blocking; idea of admin tools and better moderation UI.
  • Suggestions include automatic moderation using ML models (e.g., LLM-based filters).
  • A few note such behavior is “inevitable” in anonymous public spaces.

Comparisons & Related Projects

  • Compared to earlier browser clubs and games (e.g., past online festivals, virtual worlds, and VRChat raves).
  • Users share their own similar or more advanced VR/Unity DJ projects and discuss latency challenges for remote back-to-back mixing.

Rave Culture & Social Context

  • Long sub-threads on real-world raves: secret warehouses vs today’s commercial EDM festivals, PLUR ethos, phone bans, ticket prices.
  • Discussion of MDMA/Molly use, harm reduction, drug-testing, and how substances affect sociability vs long-term personality.
  • Debate over whether social media and streaming reduced in-person party culture by eliminating boredom and adding FOMO.

Overall Reception & Ideas

  • Strong positive emotional response; many describe it as “fun,” “fire,” “good internet,” and say they spent unusually long in it.
  • Feature wishes: skin-tone cycling, usernames, psytrance/jungle rooms, VR version, platformer-style motion capture, decentralized/gossip-based networking, NPC vs player clarity, motion heatmaps.
  • Some users experience motion sickness or technical issues and say that prevents repeat use, but overall enthusiasm is high.