I like your project, I have been a fan of both Haskell and Factor, and happy to see somebody trying to unify their qualities. I like Haskell's type system and r

I like your project, I have been a fan of both Haskell and Factor, and happy to see somebody trying to unify their qualities. I like Haskell's type system and robust fundamentals, but I also like the effectively syntax-less nature of Factor, and the evaluation order being the same as read order. I have a question: Why not base this project on Factor, which has a pretty decent existing implementation and compiler? (But few people use it..) I think one could add better typing to Factor by creating a separate module for words (like docs and tests are now) which would describe the types of each word (and also other stipulations). Then these could be type checked separately. As an aside, I always thought that it would be nice if a programming language could specify "stipulations", kinda like assertions but instead of having them inside word definitions (programs), have them outside (so that you could refer to them). For example, if something is a monad instance, the stipulation could specify that it has to follow monad laws. It would be then possible to automatically verify these stipulations (like QuickCheck does it, for example) or just assume they are true and use them for automated program deriving (or during compilation for optimization). A word annotated as having a certain type is just a special case of stipulation. Factor already has some stipulations like that in the form of additional word annotations (e.g. the word is free of side effects).