Did that for simple polynomials back in the early 80s. The approximation of an arbitrary function is more interesting to me, precisely because I didn't do that

Did that for simple polynomials back in the early 80s. The approximation of an arbitrary function is more interesting to me, precisely because I didn't do that, I guess :-) Cal State (Stan) only had (some) Lisp available because I was interested in AI. We only had shallow treatment of it (GC scoffed at...), and I've only been discovering deeper features second hand by reading about it the last 10 or 15 years.