MIX again is cool. The more lower level the language is, the better you understand the algorithm. Scheme maybe not so much. DS&Algo is one of the important things that need to be taught in low level languages.
I don't think that is what my experience taught me. Writing in Scheme got to the heart of the algorithms part much faster than MIX. With MIX I spent much more time trying to get syntax correctness than understanding what I was doing.
Further doing these things in Scheme has had long lasting benefits when I need to work in a functional programming space. In my experience, going from functional to imperative was easy, my colleagues that had to go the other way seem to struggle more.
That said, some data structures & some algos need to be written at a very low level to understand why standard Big O algo analysis can occasionally lead you astray. But I don't think they all need to be.
I disagree: the higher level the language, the better one can focus on the algorithm itself rather than getting bogged down in irrelevant details or language-specific details.
For example, being able to swap by doing "a, b = b, a" in python allows one to focus on the rest of the algorithm rather than getting bogged down in: "temp = a; a = b; b = temp". Generally speaking, the more lines of code the harder it is to see what the code is doing.
