1. 20 Sep, 2020 1 commit
  2. 19 Sep, 2020 3 commits
  3. 14 Sep, 2020 3 commits
  4. 13 Sep, 2020 1 commit
  5. 10 Sep, 2020 2 commits
    • selfReferentialName's avatar
      now to get it to build · f164355b
      selfReferentialName authored
      f164355b
    • selfReferentialName's avatar
      Get the type inference machinery to a mostly done state · 812ac935
      selfReferentialName authored
      Type inference has taken me a lot of time to implement. I think I have
      an algorithm that should work with most usual cases. I am, however,
      aware of some cases that will cause unify to loop infinitely. The first
      and simplest is when a constraint is simplified to itself. In this case,
      it should be made a Skolem. Speaking of Skolem, the second problem is
      that we never actually substitute Skolems. We should do this when we
      know that the Skolem has a type (and this should be an error outside of
      a relevant Forall). The final case I know of is when type variables
      have circular dependencies. I think that this should only happen in a
      malformed program (and possibly not at all), but this requires more
      looking into. It doesn't seem to have any easy answers.
      812ac935
  6. 08 Sep, 2020 1 commit
  7. 06 Sep, 2020 2 commits
  8. 05 Sep, 2020 3 commits
  9. 04 Sep, 2020 1 commit
  10. 03 Sep, 2020 2 commits
  11. 02 Sep, 2020 16 commits
  12. 30 Aug, 2020 1 commit
  13. 26 Aug, 2020 1 commit
  14. 25 Aug, 2020 1 commit
  15. 24 Aug, 2020 2 commits