@isaacg what should we do with the non-functional parts of Pyth? If, break, else, etc

@isaacg by the way we can generalize augmented assignment, such that =<op><var>... just becomes <var>=<op><var>... in general, not just binary ops

so to negate Z you could just do =_Z

@orlp I would make I function, that is implicit ). I'd keep B as is and remove E entirely.

@orlp That's how it currently works

=hZZ prints 1.

@isaacg ah k

Catching up the discussion about mJJ2 from a week ago.

Shouldn't we just change the parameter order of m?

So mJ2J gives [2, 2], because it maps each element of urange(assign(J,2)) to J.

I can't see any disadvantages (beside that the notation is unfamiliar).

And also follow through this notation on all other lambda-funcitons, like order-by, enumerated-map, reduce, ...

@Jakube because it doesn't make any sense - you're effectively time traveling

and it's impossible in Pyth

I explained it before, if you scroll up

it's because we're a fixed arity prefix language

you can't have an arity depend on what comes AFTER that thing

If we change the parameter order of map, then it is no time traveling.

In the expression map(a, b), b the iterable is evaluated first, and then mapped using the lambda-expression a.

So if we change the order to map(b, a), the parsing order and the evaluation order is the same.

So the first J assigns, all other J only evaluate.

@orlp

that's a possibility

oh now I see what you mean

It is be really easy to implement. And can save 2 chars on various challenges.

For instance on this one: goo.gl/BbJOx1

we can't swap the order of L though

actually

that generalizes a bit

you can't swap the order for things where the argument is optional

not sure if there are any other functions for which the argument is optional

What do you mean by swapping the order of L?

But your right, optional arguments are difficult.

like, since calling L is optional the lambda has to be first

(in Pyth5 L is quite different)

Oh, I see

L no longer replaces y - you call it using L

Would there be any other conflicts?

but when you define L you instantly call it

Yes, I know.

But I think it could work on everything else, like map, enumerated-map, filter, order-by, group-by, reduce.