[PRL] the "success" of types

[Paul A. Steckler]

> Just in case it isn't clear: I am not arguing against stating 
> invariants in imperative programming. I am only arguing against buying 
> the "types did it all, and nobody else was around" story and dumping 
> the functional programming idea.

Here's the Haskell equivalent of Scheme's for-each:

 mapM_ :: Monad m => (a -> m b) -> [a] -> m ()

It's in fact more general than for-each, because 
it's parametric in the monad.

As a programmer, the type of this combinator 
gives me understanding beyond the understanding 
I have from reading the textual description of 
for-each.  Or at least, the type conveys that 
understanding much more concisely.

I'd say that types have definite added value, even 
where there's a functional core.

By the way, did you notice how Algol60 declined 
precipitously after the Kennedy assassination?

-- Paul