I have a variation here:

function [] = ...
function [x:xs] = ...
 
Another of mine is the classical:

print function arguments -- or variations

but I finally learned to use '$' here .

 

I think that part of the problem  is that haskell syntax is powerful but terse, and therefore is not easy for a compiler to guess the intensions in the programmers.
For instance, in any other language the expression 'map f'  would give the error "missing second parameter",  while in haskell could be a partial function application.

However, some of ghc messages could probably  be made more 'beginners-friendly'.