Most Julia built-ins are defined so that the first argument is the (Smalltalk style) message receiver, but in(x,y) reverses the apparent standard, testing whether x is in y (the message receiver).
append(x,y) appends y to x (the message receiver); push(x,y) pushes y onto x (the message receiver); in(x,y) should test whether y is in x, not the reverse. IMO, defeating orthogonality is a mistake. What's the justification for 'in()' violating the usual message receiver semantics?
