Vyxal

lyxal

00:47

@AaronMiller so 3 3λ2|+; should print 6?

@hyper-neutrino @user ^

@UnrelatedString ^^

Unrelated String

01:00

think so yeah

lyxal

I was going to push the change to democracy, but merging and stuff screwed me over so I pushed straight to master

everyone ignore PRs 102, 103 and 104 please

lyxal

01:15

@user just because I overwrote vyxal.py on the e branch doesn't mean I won't be using your functions

here's the point where the functions are

lyxal

01:40

def t_lambda_b079a1a7b0a180bab100272e08a21f84(parameters, arity=1, self=None):
    global context_level, context_values, input_level, input_values, retain_items, printed, register
    context_level += 1
    input_level += 1
    this_function = t_lambda_b079a1a7b0a180bab100272e08a21f84
    stored = False
    if 'stored_arity' in dir(self): stored = self.stored_arity;
    if arity != 1 and arity >= 0: parameters = pop(parameter_stack, arity); stack = parameters[::]
    elif stored: parameters = pop(parameter_stack, stored); stack = parameters[::]

that's how =ς is transpiled

=ς being inner-product by equality

hyper-neutrino

03:07

that's... quite long lol

@lyxal uh if i'm reading that right, yes i think so

lyxal

Good

@hyper-neutrino and probably not how I'll even stick with transpiling transformers

hyper-neutrino

lol

wasif

05:23

Vyxal interpreter was visited 24813 times last month

very nice

lyxal

05:39

Me when relevant gist: gist.github.com/Lyxal/1151ae11b96e0f37b19d0125a4ff640a

hyper-neutrino

05:56

in The Nineteenth Byte, 15 secs ago, by A username

I prefer stack based, although something that visualised the stack would be useful.

step-by-step debugger when

/hj

lyxal

06:07

@hyper-neutrino hmm yes let's step by step debug a transpiled program which has absolutely no traces of the original program.

\s

hyper-neutrino

:thonk:

Razetime

@hyper-neutrino come to stax

we have step by step debugging AND mid program breakpoints

lyxal

Yeah but how often do you see stax win?

Why would you downgrade for one feature?

*two features

Razetime

lol stax does win, and even in very recent challenges

although I don't really check who's the shortest it is still very competitive

@lyxal mind you, stax may not be in hyperactive dev, but it is probably the only good golflang other than vyxal with a dev who still updates the language

A username

06:39

I pushed a button!

(Dictionary compression actuallyworks)

lyxal

07:23

@Ausername lambdas are now called at end of execution if they are on the top of the stack

Well, more generally, if they are printed

AviFS

08:43

Anyone want to save my brave Vyxal answer?

lyxal

I added a serious attempt as a comment

AviFS

And no upvote :p

Here's the reply comment I added:

> You should post that separately, it's all yours! This was a protest answer, to be honest. Against dictating how to solve something rather than what to solve. And it seems to me a slippery slope in enforcing the how, so I'm testing the extreme end of it.

A username

09:11

What happened to Lyxzfeed?

lyxal

13:01

heh I've done it

I've figured out how to transpile transformers

so far, only monadic transformers (they only group one element) have commands

and then, those aren't actually implemented

as in, it generates valid python, but the functions it references aren't written yet

but still

this is good progress

-ƇƬ turns into

def _lambda_b7862c3cee257609281116dd2e33f584(parameter_stack, arity=-1, self=None):
    global context_level, context_values, input_level, input_values, retain_items, printed, register
    context_level += 1
    input_level += 1
    this_function = _lambda_b7862c3cee257609281116dd2e33f584
    stored = False
    if 'stored_arity' in dir(self): stored = self.stored_arity;
    if arity != 1 and arity >= 0: parameters = pop(parameter_stack, arity); stack = parameters[::]
    elif stored: parameters = pop(parameter_stack, stored); stack = parameters[::]

oh, and

24 hours ago, by lyxal

@hyper-neutrino I'm going to flex on you by getting a good night's sleep lol :p

kekw

o/

user

13:35

@lyxal Ye

user

13:48

Idea: Digraph to tell if a number exists in an ordered generator. The generator may be in ascending or descending order, but you won't have to evaluate an infinite generator to tell if a value exists in it or not.

user

14:51

@lyxal :)

Idea: "Tags" for each command that don't show up in the descriptions, but make it easier to search for each command based on what it looks like or what it does

pxeger

@user assuming it's only countably infinite

user

idk what that means :/

pxeger

well, that wouldn't work for a generator of strictly ascending/descending arbitrary-precision fractional/real numbers. E.g. if you want to test if 1/1 exists in x = 1; y = 0; while 1: x <<= 1; y += 1/x; yield y it would never return

user

In that specific case, it wouldn't, but if you're looking for, say, 0.3, and your generator is 4, 2, 1, 0.5, 0.25, ..., that'd work

pxeger

15:06

yes

it's still worth having as a builtin I think, I'm just pointing out that it's not necessarily true that it can be evaluated in finite time

user

Oh yeah, definitely

@pxeger In that case, I'd just use it anyway and then say that Python's just too slow

pxeger

lol

lyxal

23:24

@user odd, it should already work for sequences like that

Look at line 182

user

But Vyxal doesn't always know if a sequence is monotonic or not, does it? Sometimes you'll have to tell it explicitly

Gimme a minute, I'll try to come up with an example

lyxal

@user not always, but it will recognise infinite lists as so

user

But not all infinite lists are monotonic

lyxal

23:44

How would I be able to tell if an infinite list is monotonic?

user

@lyxal Try running 3.5 0λ1;λ›;ẋc, it's a monotonic infinite list, but obviously Vyxal doesn't know that

@lyxal You wouldn't, but the user writing a program would

Husk has £/oelem for this, but making it a single byte is probably a waste of space

lyxal

23:59

I see

Transcript for