Get the latest tech news
Implementing Brainfuck in APL
s a pretty great language for quickly iterating on program architecture. In particular, here is a quick comparison of a couple brainfuck implementations.
This operator runs single-instruction executors recursively over the full instruction input, which lets us plug and play different implementations of the execution unit. Squinting our eyes a bit, this architecture is spiritially in the same family as an LR parser, since it fundamentally is limited to a small (1 token) parse window. Apart from it’s parsimony, what stands out to me with the v1 architecture is that it doesn’t bake in the “single instruction at a time” constraint, making data-parallelism more manifestly accessible.
Or read this on Hacker News