I've encountered this idea that k's terseness somehow improves instruction cache use before. Can you explain further? It seems nonsensical, since instruction ca

I've encountered this idea that k's terseness somehow improves instruction cache use before. Can you explain further? It seems nonsensical, since instruction caching is about machine code, not source code. Why should it use the instruction cache better than any other JIT? Or is it interpreted, in which case "the terseness of the language improves cache use" might seem more of an admission than a boast... :-)