Examples in chapter 10 (in Laws)

[rpominov]

Hey, @DrBoolean! I just read chapter 10, it's great as always! Noticed couple of minor issues in examples.

# 1

https://github.com/MostlyAdequate/mostly-adequate-guide/blob/master/ch10.md#laws

var tOfM = compose(Task.of, Maybe.of);
var assemble = _.compose(_.concat, _.concat);

liftA3(assemble, tOfM('Rainy Days'), tOfM(' and Mondays'), tOfM(' always get me down'));
// Task(Maybe(Rainy Days and Mondays always get me down))

I think assemble should be defined as:

var assemble = liftA3(_.concat);

# 2

https://github.com/MostlyAdequate/mostly-adequate-guide/blob/master/ch10.md#interchange

var v = Task.of(_.reverse);
var x = Task.of('Sparklehorse');

v.ap(Task.of(x)) == Task.of(function(f) { return f(x) }).ap(v)

Here var x = Task.of('Sparklehorse'); should be var x = 'Sparklehorse';.

[DrBoolean]

Thank you! For # 1, I think _.concat will get called prematurely since it partially applies during each ap. You're intuition is right and we could do it with liftA3(curry3(_.concat)), but it works as is. # 2 is totally right, I'll update now. Thanks!

[rpominov]

I still don't understand how it will work. _.compose(_.concat, _.concat) creates a curried function that operates on arrays or strings, but liftA3(assemble, tOfM...) will call it with Maybies as arguments. Is that right?

var assemble = liftA3(_.concat) also won't work though. Because _.concat is two arity. So the correct implementation is:

var assemble = liftA3(_.compose(_.concat, _.concat));

Btw, it was pretty tricky to grasp that _.compose(_.concat, _.concat) creates a 3-arity concat :)

[DrBoolean]

Yeah, that's a really bad example... I'll try to come up with something a little more digestible.

[rpominov]

I think I finally get it. I didn't know that some Maybe implementations include .concat() and when called call .concat() on the wrapped value. The example should work with a Maybe like that. Although, if I remember correctly, it was stated somewhere in previous chapters that the implementation from Folktale is used in the book. And it doesn't seem to include .concat() https://github.com/folktale/data.maybe/blob/master/lib/maybe.js

And yeah, the example in general seems a little too hard to grasp :)

[salper]

The current definition for #1 in chapter 10 is:

var tOfM = compose(Task.of, Maybe.of);

liftA2(_.concat, tOfM('Rainy Days and Mondays'), tOfM(' always get me down'));
// Task(Maybe(Rainy Days and Mondays always get me down))

But using the book definition of Maybe and data.task, I can make it work using:

const tOfM = compose(Task.of, Maybe.of);

const concatMap = curry((m1, m2) => liftA2(concat, m1, m2));

liftA2(concatMap, tOfM('Rainy Days and Mondays'), tOfM(' always get me down'));
// Task(Maybe(Rainy Days and Mondays always get me down))

Am I missing something ?

[DrBoolean]

Thanks @salper, yes I really need to get on fixing this. You're definition is a good fix, but I should update to a simpler example anyways.

[salper]

Arf, I could actually have made it point free:

const concatMap = liftA2(concat);

What do you mean by simpler ?

[DrBoolean]

Instead of liftA2(liftA2(concat)) I should do:

var CTM = Compose.of(Task.of(Maybe.of(x)))
where
Compose.prototype.map = function(f) { return this.map(map(f)) }

Well...on second thought. Maybe I'll just sneak in the extra liftA2