Strict arguments with foreign import

Hi all,

I'm just wondering -- for "primitive" C types (int, float, char, etc), 
why was the decision made in the FFI to have function declarations look 
like e.g.

  foreign import ccall "math.h sin" sin :: CDouble -> CDouble

rather than

  foreign import ccall "math.h sin" sin :: !CDouble -> !CDouble

I'm fairly sure any foreign imported functions must be strict in their 
arguments, and it would make things a bit clearer for Haskell 
programmer.  Am I missing something (non-)obvious in my reasoning?


-- 
% Andre Pang : trust.in.love.to.save  <http://www.algorithm.com.au/>