I know this might sound like a "get off my lawn, kid" comment. Reading through what Mr. Levin answers here, he makes a lot of good points that I have thought a lot about lately. We have a lot of "neat learning to code" sorts of kids environments (Scratch comes to mind). It's like we have decided that new programmers can't learn to program if we don't give them some graphical environment that requires a minimal amount of actual code and copious amounts of sprites and game tidbits. But a lot of us were taught to program (or just hacked away until we learned) on systems with half the functionality of Levinux. I meet a lot of new "professional" developers who wish they had learned on systems like Mr. Levin describes instead of "newer" IDEs where you drag and drop bits of things and they "just work". When asked why they didn't try, they always seem to say something about not knowing where to start. There is nothing wrong with newer IDEs and drag and drop development, but this is a great way to push all of that aside and learn the "muscle memory" of development that Levin describes so that developers can become "spontaneous natural coding problem solvers".
Hi, Mike Levin here. Sorry, can't help following my Referrers (HitTail habits die hard). The "knowing where to start" problem is exactly what I'm trying to address with Levinux. Something I didn't get into in the interview was how I actually started out on the Amiga computer, and had the carpet pulled out from under me, and then moved to Active Server Pages in Windows, and again had the carpet pulled out from under me. The short stack is a fool me twice, shame on me reaction to the state of dev-tech. If not your main dev platform, then at least a safety net.
Oh, and get off my lawn!