@MartinEnder Let's say Lisp to keep things simple. The program is interpreted as a grid and padded with spaces like a normal 2D language. (I'm actually finding designed transition rules for this to be a fairly interesting problem, so I might make a challenge out of it)

s/designed/designing/

@Dennis you allude to the significance of multiple names for a single function in the docs, but don't address it again later? is it useful so you can a recursive function where one intermediate result grows, as long as it shrinks for every other call?

← 1 message moved from talk.tryitonline.net

The docs are still WIP. Even if F and F are declared together, F(1) can reference f(1), but not F(1).

hm, yeah you couldn't really make them mutually recursive with a single definition

you've also got a typo "grater" somewhere in there

Probably full of typos. ;)

@Dennis actually, in that part you're saying that your integers are all strictly greater than zero, but you're using zero in most of the examples afterwards

>_<

Dennis must have been very tired when he's written the readme ;p

btw, is the implementation memoised?

It is not. Without some tweaks, that would affect DOS.

DOS?

Divide Or Surrender.

oh okay

hm, I've got this Fibonacci program:

	f
f
	dot g
g
	f dip
        f dip dip

now I wanted to extract f dip into another function h (so that I can save bytes by getting rid of f)

but that gives zero

am I running into some DOS trouble there?

I'm not seeing how though, there should still always be at least one dip between any two calls to the same function

I assume you tried this.

	f
f
	dot g
g
	h
        h dip
h
	f dip

yep

The problem is that the Surrender exception only affects the calling function. During the same call to g, h may return normally but h dip abnormally.

sorry, not following

What will raise an exception is the f dip call. h either returns the result, or its own argument.

oh, I see

so I end up with the base case (0,0)

Precisely.

I guess that's optimal then? extracting the f or dip into the definition of f has similar problems

well f doesn't even make sense because there's no automatic mapping happening, but extracting dip doesn't work, because then g(1) won't fail

Almost optimal. You can sum once at the end.

	dot f
f
	f dip
        f dip dip

oh damn that's neat

It's also my first Dodos answer. :) I went through the exact same process last, from fg, to trying fgh and wondering why it fails, to golfing the g.

oh right, I didn't see you had already answered the Fibonacci challenge

I'd just assumed you'd started with HW :D

Fibonacci is way easier in Dodos. ;)

that's true

parity test: tio.run/##S8lPyS/@/58zJTFJIU0hnSuNizNNISWzgCudixPI/P//v5EpAA

first pass at a multiplication program: tio.run/##LYxRCoAwDEO/… haven't really golfed it yet, because it's not valid for our multiplication challenge anyway (which requires signed inputs)