r216-forth

Forth also requires very minimal resources to implement which can be fun in constrained environments. Here it is on a TI-84+ calculator[0] or even a computer in The Powder Toy.[1] The runtime design is a little quirky but straightforward and you can extend the system as you execute (or even change interpreter semantics), or add optimizations such as some form of JIT quite easily. [0] https://github.com/siraben/ti84-forth [1] https://github.com/siraben/r216-forth

A Forth is so easy to implement that it can be done on a graphing calculator[0] and in-game computer[1] using very similar designs. [0] https://github.com/siraben/zkeme80 [1] https://github.com/siraben/r216-forth

There are some interesting electronics driven by arcane magic R216K2A by LBPHacker. A 16-bit computer https://powdertoy.co.uk/Browse/View.html?ID=2303519 Complete with instruction manual and assembler: https://trigraph.net/powdertoy/R216/manual.md Of which siraben has written a Forth implementation for: https://github.com/siraben/r216-forth

There's a Powder Toy subcommunity which focuses on building computers in-game. The state-of-the-art (subframe) involves exploiting particle evaluation order to create CPUs that can run at 1 instruction per frame. I even wrote a Forth implementation for such a 16-bit 8K RAM computer. Good times. [0] https://github.com/siraben/r216-forth

The R216 computer in powdertoy: https://trigraph.net/powdertoy/R216/manual.md A Forth for the R216 computer: https://github.com/siraben/r216-forth

Implementing a Forth is a great incremental process, all you need is a processor with stacks, registers, and RAM. Words are just lists of addresses. Once you have the base execution routines, e.g. DOCOL, NEXT etc., implementing the rest of the interpreter is easy. It can even be done on a calculator or in-game computer![1] [1] http://github.com/siraben/r216-forth