
Funny, a 7 year old post about “forgotten” techniques. I was actually just researching Fibonacci caches as one of the impactful, early, optimisations of unification (the inference technique that underlies prolog). I was on one of my “everything old is new, still relevant, and all models of computing are actually the same” kicks researching AI for my game. Turns out that complex state machines and hierarchical task networks look the same when you squint, and there’s about a million ways to dynamically evaluate preconditions depending on what model you decide you need.
I used to work in a computer lab, open plan, where we all had CRTs. I sat across from the main DB admin, who had TWO monitors for all the work he was doing (wild stuff to have dual CRTs back in those days.) Due to the layout, my monitor sat in-between his, facing the opposite way of course. I loved degaussing my monitor because: