I'm reworking the (already pretty old) concept of literate programming, basing my research on the implementation written originally by Dr Ross Williams in 70's called funnelweb. Given new hardware of modern times all the ugly hacks and compromises of the C implementation (including ugly delimiters, unnecessary terse syntax, not using a recursive descent parser but relying on byte values while parsing text) a new take on the subject might bring it to today's tool chain.