fx2d-physics-engine

Fx2D Roadmap

This file tracks the next feature and robustness targets for Fx2D.

Priority Targets

1. Add more collision shapes. Start with a few high-value shapes that unlock better level geometry and character collision:
   • capsules
   • line segments / edges
   • chain or polyline colliders
   • rounded boxes or other rounded convex variants
2. Add higher-level query and event systems. Build public scene/world APIs for:
   • ray casts
   • overlap queries
   • shape queries
   • sensors / trigger-only fixtures
   • buffered contact and sensor events exposed after each step
3. Add continuous collision support. Reduce tunneling for fast movers with:
   • time-of-impact style sweeps
   • speculative contacts
   • fast-body or bullet-style handling for selected entities
4. Harden broad-phase and collision robustness. Continue the registry-backed dynamic AABB tree cleanup and tighten collision stability for:
   • stacking
   • restitution edge cases
   • friction stability
   • sleeping / waking interactions
5. Expand test coverage and regression coverage. Add more automated checks for:
   • YAML scene loading
   • joints and motor control
   • collision manifolds and solver regressions
   • broad-phase updates and removal paths
   • fast-moving body edge cases once CCD lands
6. Add more examples and docs around newer features. Prioritize:
   • joint control examples for position, velocity, and effort
   • scene YAML examples that include joints
   • query/event examples once those APIs exist

Why These Matter

• More shapes improve practical scene authoring and reduce the need for awkward polygon approximations.
• Query and event APIs make Fx2D more usable as an engine subsystem, not just a step-and-render loop.
• Continuous collision is one of the biggest gaps for fast-moving bodies.
• Better tests and solver regression coverage are key to making the engine more trustworthy as features grow.

This site is open source. Improve this page.