Or all the other fevers we've suffered through. Haskell seems like an attractive language but I wish boosters wouldn't advocate their languages as the fix to the problems of large-scale software engineering issues when there's little or no evidence for their claims.