Reflecting on a recently posted article of mine, Runtime type checking, past, present, and future (hopefully), got me curious about how others view (and might attempt to solve) the need for optionally typed subroutine signatures.