That’s great, thanks! I see there is a mini Haskell there too. But yes, a smallest possible ML is what I had in mind.