I agree that the consistency of esac isn't very nice. I prefer } to esac/fi to end, actually. "end" is horrible. You can consecutively }}}} or )))) in a way

I agree that the consistency of esac isn't very nice. I prefer } to esac/fi to end, actually. "end" is horrible. You can consecutively }}}} or )))) in a way that reduces some of the whitespace-significance jealousy end end end end But I don't think the syntax is bad. In fact, "fi" and "esac" give some hints to the reader in deeply nested structures (I can hear it already: then don't deeply nest) The two line = { is only for compound functions; a single expression with a "where" seems fine. I do agree that code layout conventions that reduce the usable vertical space are questionable; scrolling sucks. I didn't especially like SML (or Ocaml) syntax, but they're both solid languages. http://www.smlnj.org/dist/working/index.html http://www.smlnj.org/dist/working/index.html seems to contradict the story by Mythryl's author that SML/NJ isn't maintained. His fork doesn't seem to do 64-bit, while SML/NJ does. I'd be interested in hearing more about the reason this project is not just a set of libraries for SML/NJ. I guess the author has strong opinions about syntax and didn't like ML's.