Function composition and its generalizations are cool, but let's not blind ourselves and claim that they're the key ingredient for good programming.
We should be looking for more ideas that make programming better with minimal learning cost. Examples of such ideas that were wildly successful in the past: garbage collection, modules and imports, Unicode strings. Each of those was justified by practice rather than theory.
We should be looking for more ideas that make programming better with minimal learning cost. Examples of such ideas that were wildly successful in the past: garbage collection, modules and imports, Unicode strings. Each of those was justified by practice rather than theory.