Scott Manley used a ZX Spectrum to pilot a lunar-lander simulation by interfacing the vintage computer with kRPC and a Python program [1].
The experiment tests the boundaries between legacy hardware and modern software. By attempting to control a complex simulation with a first-generation home computer, Manley explored whether the primary bottlenecks in computing are raw processing power or the ability to move data.
Manley connected the ZX Spectrum to the simulation to determine if the CPU and memory of the old machine could handle the necessary calculations for a spacecraft landing [1]. The setup required a bridge between the 1980s hardware and current software environments to facilitate communication.
Despite the age of the hardware, the project revealed that the processor was not the primary obstacle. The limitations of the experiment were found in the input and output capabilities of the machine, and the effort required to write software for a defunct platform [1].
"It turns out that the limitations in this are not related to the CPU or memory, but to the I/O and my patience for developing old software," Manley said [1].
The project demonstrates that while modern software requires immense power for graphics and physics, the basic logic of control can often be handled by much simpler systems, provided the data can actually be transmitted [1].
“The limitations in this are not related to the CPU or memory, but to the I/O”
This experiment highlights a fundamental distinction in computer science between computational capacity and connectivity. While the ZX Spectrum's CPU is negligible by modern standards, the fact that it could still execute the logic for a lunar landing suggests that the 'bloat' in modern software is often unrelated to the core logic of the tasks being performed.
