Re: Haskell, anyone?

[Hadrien Grasland <hadrien.grasland@gmail.com>]

Le mardi 17 novembre 2015 19:04:53 UTC+1, Paul Rubin a écrit :
> > I would say that recursion is the goto statement of functional
> > programming languages : necessary in order to achieve
> > Turing-completeness, and occasionally useful, but makes control flow
> > messy and programs very difficult to reason about,
> 
> No that's continuations ;-).  Recursion works fine and is simpler
> than loops to reason about and simple to use.

Consider these two C code snippets:

int iterate_recursion(int n) {
...if (n < 2) {
......return 1;
...} else {
......return iterate_recursion(n-1);
...}
}

int iterate_goto(int n) {
...beginning:
...if (n < 2) {
......return 1;
...} else {
......n = n-1;
......goto beginning;
...}
}

You have to admit that in the end, the basic approach is almost exactly the same, and people will need to reason about it in exactly the same way ("okay, so at the start of the iteration, the input parameter is indeed n, but after that, we'll decrement it and go back to the beginning of the function, this time with the input parameter equal to n-1, and then...").

Hence my claim that recursion is, indeed, the ultimate goto, including when it comes to ease of understanding. Guy Steele will be most pleased.