A physics engine prototype developed in 1996 by an independent creator has been re-released as an interactive web experience, offering insight into the technical foundations of vehicle behavior in early Grand Theft Auto titles.

The simulation, originally coded in GFA BASIC for an Atari ST computer, was later rewritten in C and integrated into the game engine for what would become the first Grand Theft Auto. It employed a basic rigid-body dynamics model, a relatively uncommon approach in consumer games at the time, to manage vehicle movement and collision.

The recreation, now available through a browser-based platform called Motion Lab, allows users to interact with three distinct vehicle types: a standard car, a spaceship variant, and a simplified brick-shaped model. Each mode uses the same underlying physics core but applies forces differently — through pointer drag, thrust mechanics, or engine and steering inputs.

Collision boundaries can be toggled, and camera behavior adjusts dynamically based on speed and position, mimicking design choices from the original game. Controls are accessible via keyboard and touch interfaces.

The simulation’s tire model relies on simplified resistance calculations rather than realistic friction physics. Sideways motion is heavily restricted while forward movement experiences minimal damping, creating a stylized but functional sense of traction. Steering is achieved by adjusting the heading of front wheels, which alters force direction and rotates the vehicle.

The project’s creator has emphasized that the recreation is entirely independent of Rockstar Games and Take-Two Interactive, the current rights holders for the GTA franchise. A public disclaimer confirms no affiliation with the commercial series.

The release coincides with the 30th anniversary of the original prototype and serves as both a technical retrospective and an educational example of early game physics implementation. It illustrates how constrained computational models can produce recognizable gameplay behaviors through iterative design rather than precise physical accuracy.