i believe the point is more that bootstrapping a higher-level programming environment up from machine code shouldn't be the sole domain of so-called genius. a forth basically writes itself once you know how the execution model works, but these days we don't focus on teaching the end to end skills required to take arbitrary hardware and target it with a new, simple language.

This; it's not rocket science. I am no university educated (maybe something compared to a community college) and yet understanding how a Forth and/or a Lisp bootstrap themselves from small cores it's something every programmer should experience at least once, even from books. If SICP for Scheme and the ones for Common Lisp are too complex, the author from https://t3x.org has two or three for scheme, one of them for almost a micro Common Lisp (less than Elisp actually), a Forth and several more. Oh, and the code runs on 8086 PC's, Unix and some of them even under CP/M 2.2. It's crazy, and eye-opening too.

With a Forth you can literally see how floats are built from 'integer' blocks in RAM. Heck, you can even see how floats are done by using integer and string related words (printing them) . And you can see the 'odd' integer based stack in a live way with once your entered a double or a float. Binary representation in the spot.

I'm familiar with SICP and t3x. Sorry if my statements made it sound like only a genius could figure this out. I think most coders I know are just trying to solve a concrete problem with the existing building blocks. Really smart (or interested) people tend to go a bit broader and deeper. I don't think HN folks are very representative of the overall coding industry - not even close.