advent-of-code

[fleet fox]

man

i have to remove my examples as well

noooo!!!

[low condor]

They are from your input?

[fleet fox]

no but they are from the puzzle text

the example they give

im putting it in my code lol

[hazy ermine]

thats fine ig

or is it

[low condor]

Hmm 🤔 none of my unit tests would be acceptable if we couldn’t do that

[fleet fox]

is it fine

🤔

[hazy ermine]

i believe so since its the same for everyone

and its publicly available on the website anyway

but not sure

[fleet fox]

hmmmm

[vestal token]

so far, what's been the hardest day?

[hazy ermine]

day5 probably

[low condor]

Deffo Day 5 P2 but Day 1 P2 and Day 3 get honourable mentions

Day 1 P2 annoyed me the most lol

[vestal token]

i feel that there hasn't been an increase in difficulty from day 1 to day 9

[magic oxide]

day 8 p2 annoyed me the most, day 1 p2 took me the longest relative to how difficult it was, day 5 p2 took me the longest period

[hidden musk]

it hasn't really been linear

[low condor]

Eric deliberately doesn’t make it exactly linear, but even so this year has been a rollercoaster

[hidden musk]

i genuinely gave up on doing day 5 part 2 efficiently

[magic oxide]

tbf u were doing it in like what

prolog?

[hidden musk]

yeah

[hazy ermine]

weekend puzzle tomorrow

[low condor]

The language challenge has been fun - not because I’ve done it but because it’s been entertaining to watch from the sidelines

[magic oxide]

you know what would be an added challenge for the roulette?

no docs

or stackoverflow

or anything

u have to figure out how the language works just by running it

[hazy ermine]

come on that would mean we wont be learning anything

[fleet fox]

i would lose it

[magic oxide]

i would think you'd learn a lot simply because you're forced to to solve the problem

[hazy ermine]

true but there should also be enough time

since doing it like that takes a lot longer

[royal gull]

Day 8 p2 I solved accidentally quickly with my reading incomprehension.

[hazy ermine]

😂 😂 😂

man i honestly expected little more hints for d8p2

[royal gull]

What happened was: ||On first read, I assumed it looped back to the start after reaching the finish, making the LCM solution obvious||

[lime fulcrum]

today is kind of hard

[hidden musk]

that's messed up

that's actually wouldn't be fun though

if you just go in without knowing anything then you just have to guess at everything

[woven solar]

same, half an hour later I was like "wait, that shouldn't have worked"

[magic oxide]

yeah on second thought i agree

but

[woven solar]

but the input just happens to be such that you don't need the fully correct solution

[magic oxide]

if you had some prior experience with the language

[hidden musk]

you would need a lot of experience to do it without docs

like, python or rust, i could probably do. not a random language

[magic oxide]

i guess i haven't done the roulette so idk how difficult it is

[hidden musk]

well i guess it depends how much "prior" is

like, i had used vb, ocaml, and haskell

but i was still dying 😩

[minor cave]

The repo stats are gonna look real cute though!

[hidden musk]

yeah it's quite colorful

https://github.com/jonathan-d-zhang/aoc2023-roulette

[magic oxide]

dang that purple one

[hidden musk]

vb

the card day

[royal gull]

Waiting for my sql day.

[hidden musk]

👎

[flat sequoia]

I don’t do this fwiw

I have code to programmatically retrieve the input but that’s basically it

And a thin wrapper that just opens the right files

And I only started auto fetching input halfway thru last year

[simple raven]

I saw you got first today right? congratz

[flat sequoia]

Yeah that was crazy!!

[simple raven]

just had a look at your solution, amazing you managed to write that in 1:28 😄

[flat sequoia]

I think it’s a combination of problem solving experience (quick thinking and adapting existing patterns you’ve seen) and knowing the language super well

[simple raven]

not at all surprised that p2 took you 20s 😄

[flat sequoia]

Yeah I read p2 and was like that’s it?

Almost didn’t believe my reading comprehension

Idk personally having a library doesn’t seem that helpful to me, though maybe some parsing helpers would be beneficial

But I feel like the times when parsing is annoying is when the format is really weird and idk if your library would work anyway

Also wow I’m like barely ahead of this Kevin guy

Sold too hard on day 8 I guess

[pine tiger]

oh god turns out i've been doing aoc all wrong 😭

[pine tiger]

dw as soon as problems become nontrivial i'll drop to like 9th or something

[flat sequoia]

Honestly like same

I always troll on those like do a MAX instead of MIN ahem

I need to slow down a bit sometimes

[pine tiger]

my brother in christ you've been #1 every goddamned time

[jovial jay]

Looks like I'll have to remove the inputs file. 😦

but, does this mean I can import from URL instead?

[peak dock]

i just gitignored the file

[jovial jay]

ok, looks like i can use URL instead

[hidden musk]

at least you'll learn about the reflog now 📈

[pine tiger]

hm

i take it can't even copy the example input?

[hidden musk]

i think that's fine. you can see those without logging in anyway

and i haven't seen anything saying otherwise

[simple raven]

well technically we are not allowed to copy any text at all from the assignments, but I wouldn't expect there to be much issues with copying the tests

[hidden musk]

o rly?

[simple raven]

https://www.reddit.com/r/adventofcode/wiki/faqs/copyright/inputs/ https://adventofcode.com/about#legal

[pine tiger]

well

now i've seriously no cue

*clue

as to what to do

[jovial jay]

that's why i'm changing my code to load from URL instead.

[pine tiger]

i mean then you have to do the login and everything i presume
which i have no idea how to do lmao

[jovial jay]

no you don't. importing from URL works without login

[hidden musk]

don't you need to be logged in to access inputs though

[jovial jay]

you can copy URL when logged in, and then paste it into your code to load from there

[hidden musk]

you need your session token

[simple raven]

you could just scavenge the request from aoc-lube

        response = requests.get(
            url=URL.format(year=year, day=day) + "/input",
            headers=HEADERS,
            cookies=TOKEN,
        )

[pine tiger]

what’s headers and token

[minor cave]

token is your session token. If you log into AoC and check your browser storage/cookies, you should see something like session. A long string of alphanumeric. You need that

[simple raven]

HTTP headers, I don't know how website stuff works https://requests.readthedocs.io/en/latest/api/

and token your user token

[minor cave]

for our leaderboard feature in Sir Robin, it looks something like this:

url = "https://adventofcode.com/{year}/leaderboard/private/view/{leaderboard_id}.json"
AOC_REQUEST_HEADER = {"user-agent": "PythonDiscord AoC Event Bot"}
cookies = ... # we read this in from our envars, but it's the AoC session cookie for my account
async with aiohttp.request("GET", url, headers=AOC_REQUEST_HEADER, cookies=cookies) as resp:
    ...  # do whatever we need with the leaderboard results

[wet yoke]

That feeling when you finish day 9 part 1 right before heading out for a social function and I just can't physically do part 2 yet. 😮‍💨

[untold briar]

At least you finished

Imagine being really close but not able to finish part 1 before going out

🙃

[ripe meteor]

Saturdays do be like that

[sleek cave]

programming problems like aoc have actually rewired my brain
my first thought upon grabbing this box of milk was "oh hey, that's a tiling, i wonder how many ways you can tile an m by n rectangle with 1 x 3 and 3 x 1 tiles"

and now i'm actually intrigued

because ik tilings of 2 by n using 1 x 2 and 2 x 1 can be solved using dp

but that's a wayyyy simpler problem

[ripe meteor]

Tile bad, wood floorboards is where its at

[dense osprey]

how did you guys write your cycle detection algorithm in yesterdays part 2?

due to the graph not having to be a functional once, my goto choice wont work

and a "subcycle" could be repeated multiple times in once cycle

[simple raven]

I don't know how big of a spoiler you want but ||the input is special making the task much easier than you (and originally I) think||

[dense osprey]

oh

oh

||input has the same amount of strings ending on Z as strings ending on A||

[simple raven]

yup, that's the first hint

[dense osprey]

||does the solution involve the chinese remainder theorem||

[simple raven]

afaik you should be able to solve it with that, want another hint?

[dense osprey]

no 1sec

i think i can do it

[dense osprey]

the nearest Z nodes is different for every A node

but im not sure if that matters

im trying to think of an input where the optimal result wont have the nearest z node for each a node

but if it doesn tmatter

[dense osprey]

found the solution

idk why

||lcm of nearest distances||

idk how id prove it

found it randomly

i dont understand my solution

[simple raven]

oh boi you're in for a rollercoaster of comments if you go to either the solutions channel or reddit 😄

[dense osprey]

oh what do you mean?

are there a lot of people who dont understand it

[simple raven]

the input ||was specially designed to allow for your solution||

||you need to make like 5 or 6 assumptions for it to work||

[dense osprey]

oh tbh i dont like that

[dense osprey]

not a bad problem

but

i dontr like problems like that

they couldve said that in the description of the problem

[simple raven]

this is why it works #1182550042091991180 message

[mellow mesa]

Am I wierd for not liking yesterday for giving too few hints / and not liking today for basically giving us the solution.

[dense osprey]

cause when i try solving problems like this, i try to design some algorithm such that every input complying wit hsome set of constraints would work on it

[woven sable]

honestly i would've liked today and day 5 to be flipped

[simple raven]

making puzzles is hard 🤷‍♂️

[woven sable]

today felt awfully easy for day 9

but i'm not complaining these puzzles r sm fun

[mellow mesa]

Yeye ofc, I love aoc either way, just when I think about it

[dense osprey]

but if you ignore that not telling us input constraints part, then it was one of the best problems so far this december

[mellow mesa]

This is also a closed off discussion forum, wouldn't really voice this on reddit or something

[dense osprey]

why?

[mellow mesa]

Just doesn't see it being necessary, especially when this is something we do for fun and the author has the heart in the right place

Just would have been nice to find the extrapolation part for today on my own (if I would have gotten to it)

[dense osprey]

yea i just sometimes would like to know if the input numbers are all <2^31

[rancid creek]

so I've put together a difficulty chart for advent of code based on recent years, and aligned it by weekday. There doesn't appear to be a "weekends are harder" trend as much as a "later in the month" harder trend.

[simple raven]

in my mind 1 is medium and 7 and 4 are switched, but that doesn't change much

oml 2019 was brutal 😄

[rancid creek]

for difficulty I used "time to complete part 2 by the last person on the top 100 leaderboard. It's not perfect, but it gives some idea of scale.

[simple raven]

so we are about to enter the main medium territory

ah okay

[rancid creek]

wait til you see 2018...

[simple raven]

I'm glad they are making AoC more mainstream over the years 😄

[rancid creek]

I think the skill of participant also increased over time, so that might be reflected in there as well, and why I'm not going back to the first 3 years.

some of those extremes were 2 hour + solves by the top 100 peeps.

[cyan sleet]

normalizing hardness per year might make sense

less people competed back then

[ancient surge]

whats advent of code?

[cyan sleet]

like, divide all gold times by the longest gold time that year

[low condor]

I remember 2020. Only one "extreme" puzzle and it wasn't even that difficult, it just involved a lot of "legwork"

[rancid creek]

probably a good idea.
but I'm done for now.

[simple raven]

coding challenge https://adventofcode.com/2023

[cyan sleet]

2020 day 20?

[low condor]

yup - the seamonster one

[ancient surge]

alr probably too hard for me then rn

[lusty rover]

u wot

[dense osprey]

average time on leaderboard or time for first solution?

[rancid creek]

https://adventofcode.com/2020/day/20

[rancid creek]

position #100 for part 2

[dense osprey]

ah

[cyan sleet]

I don't recall it being that hard

[low condor]

yup - agreed. Not difficult, but it takes legwork

[rancid creek]

first person to finish was at 25 minutes, the leaderboard filled at 1 hour 15 minutes

[rancid creek]

fun with google sheets

[lusty rover]

[lusty rover]

Why not red for insnae?

[rancid creek]

I added in insane as an after thought, and it only really applies to the older years.

chart inspired by : https://www.maurits.vdschee.nl/scatterplot/

[cyan sleet]

was 2019 2nd insane space cards?

[dense osprey]

but isnt that due to less people participating?

or not mainly due to the problems difficulty

[rancid creek]

yeah https://adventofcode.com/2019/day/22

[cyan sleet]

wrong reply, but thanks

I liked that problem

[royal gull]

Love this

[cyan sleet]

(as I've said before)

[dense osprey]

10007 is a prime number

idk if it matters

why are there so many numbers of the form 100.0007 which are primes

1e9 + 7 is also one

17 is one

[cyan sleet]

it kinda does

[dense osprey]

107 is one

[cyan sleet]

||17||

[dense osprey]

oh nvm 1007 isnt and 100007 also isnt

[lusty rover]

Do you think the problems got eaiser or the people got better?

[dense osprey]

havent read the whole problem yet

was at this part

[cyan sleet]

mostly just more people

and same size leaderboard

[rancid creek]

Probably a little of both.

[simple raven]

omg how do you even approach writing code for that?

I assume the problem essentially asks you to create some sort of hyper-optimised look-up table?

[cyan sleet]

it comes down to some number theory, operations modulo a prime

basically, after you figure out what one shuffle does you can extrapolate quickly

[simple raven]

is it similar to the ranges question couple days ago? ||instead of simulating individual cards you suffle around ranges of numbers?||

[cyan sleet]

do you know the matrix power approach for computing fibonacci numbers?

[cyan sleet]

not quite

[simple raven]

nah, my background is in physics and even then I struggled with math 😦

[cyan sleet]

basically when you have one shuffle figured out you can solve the thing in roughly log(that big number) time

[simple raven]

oh that's pretty good wow

[cyan sleet]

so the fib matrix thing boils down to this matrix on the left describing one step of fibonacci

[0 1] [a] = [ b ]
[1 1] [b]   [a+b]

[simple raven]

makes sense

[cyan sleet]

in general if you mult by that matrix n times you take n steps

[dense osprey]

lin alg

i hate it

i dont want to know the context

[cyan sleet]

but how quickly can we do that?

[cyan sleet]

this is a cool thing though

[dense osprey]

stuff involving lin alg cant be cool

[cyan sleet]

computing the nth fibonnacci number in O(log n) time

[simple raven]

I like lin alg, it's like watching a magic trick I don't understand but it's cool to see how it's done

[cyan sleet]

(ignoring the raw cost of large integer operations...)

[cyan sleet]

so the trick is this, say we wanted to do n=85

[dense osprey]

its so imporatant

important

[simple raven]

left side matrix to the nth power?

[cyan sleet]

expressed in powers of two: 85 = 64 + 16 + 4 + 1

we can compute powers of two pf a matrix easily by just squaring the matrix

[cyan sleet]

yeah

[sharp tapir]

hey could someone possibly give me a pointer on day7?

[simple raven]

do you then do

[0 1]^n [a] = [ Fn ]
[1 1]   [b]   [ Fn+1]
``` ?

[sharp tapir]

hand_rank={'A':14, 'K':13, 'Q':12, 'J':11, 'T':10, '9':9, '8':8, '7':7, '6':6, '5':5, '4':4, '3':3, '2':2}

dict={'Five of a Kind':[],
      'Four of a Kind':[],
      'Full House':[],
      'Three of a Kind':[],
      'Two Pair':[],
      'One Pair':[],
      'High Card':[]}


with open('Day 7\input.txt','r') as file:
    lines=file.readlines()
    for i in lines: 
        hand,bid=i.strip().split()
        
        dict_temp={}
        for c in hand:
            if c not in dict_temp:
                dict_temp[c]=1
            else:
                dict_temp[c]+=1
        
        # classify the hand
        if 5 in dict_temp.values():
            dict['Five of a Kind'].append(hand)
        elif 4 in dict_temp.values():
            dict['Four of a Kind'].append(hand)
        elif 3 in dict_temp.values() and 2 in dict_temp.values():
            dict['Full House'].append(hand)
        elif  3 in dict_temp.values() :
            dict['Three of a Kind'].append(hand)
        elif list(dict_temp.values()).count(2)==2:
            dict['Two Pair'].append(hand)
        elif 2 in dict_temp.values() : 
            dict['One Pair'].append(hand)
        else:
            dict['High Card'].append(hand)
        
        print(dict)
    for i in dict.items```
im kinda stuck here because i now need to literally sort through the hands with the strongest first. Is this the way for a noob/intermediate to go about it ?

(sorry for the massive wall)

if you want me to delete i can

[cyan sleet]

let a=b=1 or whatever your starting values are

and yeah

[simple raven]

this is pretty much how I coded it, then to sort the hands I did sorted(hand, key=self.translate) where self.translate is a method that turns the hand's cards into its values use a dict similar to your hand_rank

[simple raven]

[0 1]^n [1] = [ Fn ]
[1 1]   [1]   [ Fn+1]
``` so this is how I get fibonnaci numbers super quick? how does that relate to the solution of day22

[sharp tapir]

so if its AAAAA, its value is [14,14,14,14,14]

[simple raven]

yeah

[sharp tapir]

fk

did you generate the category list of hands first ye?

[cyan sleet]

you can express the effects of a shuffle on the card you're tracking as a matrix

[simple raven]

yeah, in a dict just like you, then sorted the lists within each category

[sharp tapir]

like this

ok

ok let me see

thanks boss

[simple raven]

O.o huh

[dense osprey]

you made me think of my solution

[simple raven]

so instead of tracking the whole deck you just track where the one card 2019 goes?

[dense osprey]

and i noticed that i compute the type of your hand way too often

up to n times for each hand

cause i didnt save the type and just called a function to compute it in my custom comparator function

wait

are we talking about this years blackjack problem

[simple raven]

poker but yeah 😄 Camel Cards

[dense osprey]

poker

yea

poker

what was i thinking 😭

blacjack doesnt have hands

like these

[cyan sleet]

basically yeah

[sharp tapir]

huh

[sharp tapir]

question

when you iterate through your dictionary of all your hands

are you doing outside of the main loop ?

[low condor]

I did this too - gotta fix this in my code :)

[sharp tapir]

main loop being the one where each line / hand is being evaluated and classified

it doesnt make sense to do it there cuz much easier to sort at the end right

when all the ahnds are categorised

[simple raven]

I'm looking at a solution in python and it's incredible that people figured this out. In this solution https://github.com/mcpower/adventofcode/blob/501b66084b0060e0375fc3d78460fb549bc7dfab/2019/22/a-improved.py what is repeats? I assume cards is the new input but I don't have access to the problem assignment

[dense osprey]

i wont

[cyan sleet]

ah, let me share the relevant part

[simple raven]

the input size is tiny so we can afford to loop several times and do everything properly, so I have one for loop that sorts hands into ranks, second that translates each hand into numbers as we talked about earlier, third that firstly sorts the hands within-rank and then puts them into one final list in the correct order

[simple raven]

oh lol

[cyan sleet]

it's log of that second number

[simple raven]

I would love to see a reddit post "so I'm going to let my bruteforce algorithm run, see you at the heat-death of the universe"

[cyan sleet]

that first number you just need to be clever about, you can't just simulate the full deck like in part 1

[low condor]

makes sense - only bring a small performance benefit anyway

[cyan sleet]

but you can still compute the effects of the shuffle

[low condor]

my code drops from 0.2 to 0.02 secs

[dense osprey]

that is a big difference

[cyan sleet]

I liked my approach to the poker problem

[rancid creek]

I don't even remember what I was doing back then, but looking at my code I saw this part:

        o *= inv(1-i); i = pow(i, n, c)
        return (p*i + (1-i)*o) % c

which managed to get the right answer.

[cyan sleet]

compute 5 counts, and then stuff that and info about the hand into an integer

and sort by that integer

[low condor]

yeah but not in absolute terms

[simple raven]

I do a bunch of sorting and appending to lists etc and it runs in 4ms

[rancid creek]

2019/22 ?

[simple raven]

day7 part2 2023 😛

[rancid creek]

ah !

[simple raven]

I couldn't even sort a list in python in 2019 😄

[cyan sleet]

I guess in python you can just do a tuple/list, i.e. 5T552 would be [3, 1, 1, 0, 0, 5, 10, 5, 5, 2]

[rancid creek]

if you're curious this was my finished solution for 2019/22:

def solve(c, n, p, o=0, i=1):
    inv = lambda x: pow(x, c-2, c)
    for s in [row.split() for row in orders.splitlines() if row]:
        if s[0] == 'cut':  o += i * int(s[-1])
        if s[1] == 'with': i *= inv(int(s[-1]))
        if s[1] == 'into': o -= i; i *= -1
    o *= inv(1-i); i = pow(i, n, c)
    return (p*i + (1-i)*o) % c

for x in [(10007,10005,2019), (119315717514047,101741582076661,2020)]:
    print(solve(*x))

my initial part 1 was more brute force.

[simple raven]

I now kinda see what you're doing but honestly even if I wrote that code and it gave me the right answer I wouldn't trust it 😄 looks too esoteric

but gj on getting it done 🙂

[cyan sleet]

yeah, you need to realize the shuffle operations end up being simple operations when modulo a prime

[rancid creek]

number of lines ~10. time to complete listed at 15 hours rank #1200

[dense osprey]

that last for loop was unnecessary

imo

but my code isnt good either

[simple raven]

the for loop is just part 1 and part 2, why unnecessary?

[dense osprey]

print(solve(*tuple1)); print(solve(*tuple2))```

[rancid creek]

which last for loop? the one over the input file, or the one that was part1/part2 ?

[cyan sleet]

I think I was in the US on a business trip end of 2019, do I couldn't have done it quickly even if I wanted to 😅

[rancid creek]

part2 took me an hour, and part 1 took 14 hours, makes me think I went to bed and did it the next day.

[lime fulcrum]

I actually think my part 7 is okay ish. one is clean enough. two might need some comments

[cyan sleet]

2018 was the one year I did kinda do leaderboards, despite timezones

56

[rancid creek]

I think that was my first year using python so it was pretty sketchy code. 🙂

[sharp tapir]

how long has everyone been coding?

i started in june

[cyan sleet]

speaking of sketchy code

this year is quite sketchy code with the language roulette 😛

[rancid creek]

I used the same code for parts 1 and 2, the only change I did was to the input file for part 2 I substituted the Jacks for Wilds.
open(aocinput).read().replace('J','W')

[cyan sleet]

err...long ago

[sharp tapir]

leetcode

[rancid creek]

nice!

[sharp tapir]

but nov/dec dont count because i was doing a sql bootcamp in person

it should be much more active hah

[lime fulcrum]

that does seem to be more reasonable

[cyan sleet]

I did project euler stuff back in like...2009?

[sharp tapir]

jeez

ur ancient

one thing about dev community

is how supportive everyone is

[rancid creek]

2009 = ancient. 😐

[lime fulcrum]

    if "J" in hand:
        res = []
        jlist = [[] for _ in range(7)]
        for i in range(12):
            handtest = line[0].replace("J",cards[i+1])
            jlist[handvalue(handtest)].append([cardsortvalue(handtest,value),handtest])
        for categories in jlist:
            res.extend(sorted(categories))
        handtest = res[-1][1]
        group = handvalue(handtest)``` just not that clean

[noble skiff]

I did my first programming competition in like 2001 probably

I guess I'm pre ancient

Primordial?

[cyan sleet]

xelf was one of the first prototypes, he's kept up fine with the years with only minor maintenance

[hidden musk]

wow you're old

[simple raven]

I first met python when I did physics master's around 5 years ago, but the course was so dense I didn't learn anything, but I really started learning python during covid and been coding since, just not CS stuff as much as other people here

[sharp tapir]

im big in crypto

and

getting better there means knowing how to code

so here i am

[tepid cairn]

I started around Covid as well 👀

[sharp tapir]

i started in jun-23, dabbled start of the year w/ web3 but im not gonna count that

really really late

[noble skiff]

I got a kid as old as advent of code and t shirts older than 80 percent of participants 😛

[sharp tapir]

bit sad

[simple raven]

I got to level 60 in wow for the first time and learned python, one has been a little more useful than the other 😄

[zenith elbow]

damn! yall giving imposter syndrome to me

[woven sable]

2021

covid basically

[tepid cairn]

hah

[woven sable]

i feel like i peaked in covid coding wise tho

[sharp tapir]

covid rly got a bunch of ppl in

[woven sable]

i think i know a lot more langs now but idt i'll ever be as good as python as i was in 2021

[simple raven]

I had to learn to code during covid for my phd

[woven sable]

i j don't understand internals like i used to

[hidden musk]

sheesh

[sharp tapir]

ok yall distracting me need to finish day 7 aaaa

[lime fulcrum]

really coding? august. except if we count sps programming

[tepid cairn]

my dad would hoard old computers and stuff, I dug one out and got it running, and that was my first real introduction to computer related stuff

[noble skiff]

I built my first computer with my dad 1991

[simple raven]

tbf a lot of what I learned since then is not necessarily being better at python, but just understanding how computers work and being generally more comfortable with new languages so I can solve basic problems like change bits for a website or use a simplified version of javascript to do something, which I wouldn't have been able back when all I knew was python

[noble skiff]

It was his computer to be clear 😛

I was mostly watching

[woven sable]

i have like another really big side hobby and i got into programming by figuring out that i could do a lot of cool stuff with programming so i started coding and then i got really into it

i had like a new project every week

[simple raven]

my country didn't exist the way it does today in 1991 😄

[woven sable]

yeahh ig that's true

[noble skiff]

Montnegro?

[woven sable]

cuz i think i'm def a better problem solver now

and i can pick up langs easier now

[simple raven]

Slovakia

[noble skiff]

Ah I was relatively close

JeremyJoe is surely the most American name a Slovakian ever had

[simple raven]

I think that's the most valuable

[woven sable]

that's true

like i was horrible at learning new languages like one or two years ago and a few motnhs ago i j found kotlin and picked it up pretty fast

[simple raven]

almost as if my 12y/o self picked it as alias 😄

[noble skiff]

Kotlin is sweet

[woven sable]

i love it

only thing i don't like is the stdlib is pretty lacking

the json parsing is a pain

and there's no || lcm function so i had to search up an lcm calculator online for day 7 part 2 because i didn't want to put in the effort in creating one myself ||

that's a pretty useless gripe tho

overall it's amazing, esp the predicate functions

i was doing aoc in kotlin that day

[cyan sleet]

oh right, czechoslovakia

[rancid creek]

ah, I missed that part!

[noble skiff]

Why would json parsing be a pain

[cyan sleet]

that's...just outside my lifetime

[lime fulcrum]

they are pretty ease. implement euclidean algorithm

[cyan sleet]

if you can compute gcd you can compute lcm

[simple raven]

same

[zenith elbow]

alive before javascript is dope. do you remember gopher quicknir? if the web wasnt a thing advent of code would still work, we can have plaintext problems in gopher and live leaderboards through telnet or ssh

[woven sable]

idk there's not any stdlib support

which kind of annoys me

[noble skiff]

I do not remember gopher

I remember pre Internet days though

Like I remember when people started talking about it but all you could do was like log into certain libraries

No world wide web yet maybe

[sleek cave]

starting day 9!

first thoughts give me ||coefficient solving|| but i don't wanna figure out ||solving a system of linear equations|| (i think it has something to do with ||linear algebra|| but i don't know ||linalg|| at all 😂)

so i might just ||naively simulate for part one||

[zenith elbow]

ahh thats cool

[cyan sleet]

back when internet and www was pretty distinct things

[atomic cairn]

you dont need equations but it might be faster or look better

[zenith elbow]

they still are technically, just that most nontechnical people interchange the two due to the web being so mainstream

[sleek cave]

lmao nevermind today was a breather 😂

now i really want to do a ||linear algebra|| solution for fun

[sharp tapir]

could anyone assist me with day 7?

i have an answer but its saying its too low

for some reason..

happy to share code

it works on the test case

[low condor]

Will ping in #aoc-solution-hints

[rancid creek]

did you try the any of the alternate tests?

[sharp tapir]

I solved in #aoc-solution-hints

But still not sure why original solution doesn’t work

I thought my approach of getting the (length of the lists from inferior categories + index of hand +1 )should work but it weirdly doesn’t

[rancid creek]

I'm not even sure what that is. 🙂 But congratulations on getting it done!

[zinc kindle]

Going through old problems and wow, 2018 day 23 part2 was a nightmare to solve. 3d geometry is a pain, tried a solution based on ||overlapping irregular octahedra||, tried to write a 2d solution before extending it to 3d, couldn't. Gave up.
Tried a simple algorithm, gave wrong answer on sample inputs (or maybe I implemented it wrong).
Eventually I just went and ||wrote an algorithm to find a local maxima given a point, used rng to find a bunch of possible start points, and kept running until it seemed like it'd found the best solution. And that finally worked||

[real quartz]

Holy hell. I'm still stuck on Day 1. It's making me think that I need to get more experience in older AoC before I even attempt 2023's version

Whoever told me that the difficulty was all over the place wasn't kidding

[magic oxide]

day 1 is unusually hard this year yeah

[rancid creek]

day 1 was harder than other day 1s, but in the scheme of things it wasn't a harder aoc puzzle,.

[earnest forge]

I'll be honest I don't remember how I originally solved that puzzle but apparently ||Z3 has an Optimize class that can solve it fairly easily||

[sleek cave]

||overlapping irregular octahedra||
what in the actual world

[real quartz]

hmm, I'll keep trying

[earnest forge]

They aren't ||irregular||, idk why Matt said they are (it's ||3d Manhattan distance radius||)

[rancid creek]

The best thing is to just keep at it, if you get stuck ask for help and keep moving forward while you're learning.

[rancid creek]

only stop once you feel like you're getting nothing out of it.

[minor cave]

Also it's fine to skip days!

[earnest forge]

^^ maybe start with like day 6 or something

[rancid creek]

day 1 wasn't so hard that it's a skip day. day 5 could be a skip day.

[earnest forge]

(yeah, but the easiest way to get someone to feel like they're having fun is to start them with an easy one :P)

[real quartz]

Might as well, lol. Thanks guys.

[jovial jay]

I'm proud of solving Day 5 Part 2 and with a optimized solution too. So happy.

[real quartz]

Noice!

[slate berry]

What's the optimisation?

[earnest forge]

Presumably it's the ||interval arithmetic|| one (but maybe there's another better optimisation I hadn't thought of)

[zenith elbow]

wrap it with spoilers please! #1181459209422917683 message

[slate berry]

😔 I was reading

[jovial jay]

i moved it

[zinc kindle]

The initial shapes are regular, but the overlaps would be irregular I think. In retrospect, those could be split into regular shapes maybe I guess?

[earnest forge]

Ohh, I see what you mean although it isn't what you said

You meant ||irregular overlaps of octahedra||

[covert arrow]

I am actually ready today

GLHF

[earnest forge]

GLHF everyone

[minor cave]

Language Roulette: Day 10
The language is ... Fortran

[sleek cave]

I JUST REALIZED
AOC 23 LORE IS A CALL STACK

[shut trellis]

<@&518565788744024082> Good morning! Day 10 is ready to be attempted. View it online now at https://adventofcode.com/2023/day/10. Good luck!

[magic oxide]

oh right

I keep forgetting

doesn't look too bad

[minor cave]

midnight keeps sneaking up on me and then I see BillyBobby talk about it in pygen and scramble

[hollow wharf]

[magic oxide]

lmao

hopefully p2 isn't too hard

don't have much time today

[sleek cave]

p1 seems all right

[hollow wharf]

i dont even know how to parse it

https://tenor.com/view/white-dog-shaking-scared-nervous-dog-scared-and-shaking-gif-22874221

[sleek cave]

time to sleep on it and dream of grayscale wastelands where my sole company scurries away in forsaken labyrinths of black and rust

[fleet fox]

💀

[jovial jay]

Day 10 looks pretty bad

[earnest forge]

WTF is part 2

[fleet fox]

what u already solved part1?

damn

[jovial jay]

here, i feel like i still need to grasp day 10 before attempting.

it took me a while to grasp day 5

[cyan sleet]

now we're talking

[shy trench]

part 1 honestly wasn't that bad

but i'm completely stumped for part 2

[fleet fox]

how do u do part 1

i have no idea where to start

[minor cave]

let's spoiler solutions in this channel please

[shy trench]

mb

[tidal mist]

no solutions thread 😔

[minor cave]

9 more to go!

I think that's you on the pt2 lb though 👀

[covert arrow]

my p2 works for example but not actual 😭

[pine tiger]

LMAO SO CLOSE

[tidal mist]

i almost gave up because i thought i was too slow. but still lb. 🥳

[minor cave]

#1183281135791386705

[flat sequoia]

Woooo finally a fun problem

I think I would have done pretty well if I didn't have to wait 5 minutes on part 1

But I mixed up | and - by accident

😒

[pine tiger]

💀

[pine ravine]

looks like we got a hard one today aye?

[earnest forge]

38m was enough for 124th so I'mma go with yes

[pine ravine]

mmm i think i can do this but its going to be wildly innefficient

[flat sequoia]

I swear the hardest part for me was figuring out which direction the S went
I should have just looked at the input and hard coded LMAO

[woven sable]

i'm still stuck on part 1 wtf

[timber glacier]

help i keep getting an answer too high on part 2

i keep getting the same answer every time i make it too though

[royal gull]

In this day and age, why don’t we have ascii art?

[earnest forge]

...we do?

[royal gull]

I mean, for aoc problems. F 7 J? My eyes hurt

[earnest forge]

Oh you mean you wanted unicode box drawing characters?

[royal gull]

Yah

[earnest forge]

that's not ascii

[royal gull]

. Its 2am and my eyes are blurry 🙂

[earnest forge]

((and I think Eric wants to keep inputs as valid ascii which seems fair so as to not unnecessarily preclude the use of C))

[pine ravine]

owie my brain hurts

[drifting gazelle]

How to install discord components?

[flat sequoia]

idk I think it's cute

also i made a gif but idk why things are yellow

[pine ravine]

urs is cooler

[royal gull]

Maybe put in #1180191057498083418 ?

I love this

[untold briar]

Part 2 sure is fun

Though it's a bit hard haha

I have no idea how to check if a pixel is within a contour and had to consult google though

[pine ravine]

lets gooo that one was so fun

i did it all with color coding

[atomic cairn]

Today looks hard

But i wont be able to code today

[austere basin]

hey guys, i coded a solution for today part 1 which works for test input but it reaches the maximum recursion limit (around 1182 i think?). is there any way to get around this or do i just have to stop using functions??

[untold briar]

Use while loop and flags instead

[rocky spire]

use sys.setrecursionlimit(...) to change the max depth
The max you can set this limit to is a 32bit int, though it's hopefully enough

[untold briar]

Python is not built to withstand big recursive loops

Or what purply suggested, yeah

[austere basin]

right... it's fine guys it's not difficult at all to modify it to stop being recursive

||also... just curious from looking at the animations, is part 2 the number of squares contained within the loop?||

cuz if so that will be fun :)

[untold briar]

||Yes it is||

Also, remember to add spoiler

[austere basin]

OH MB SRY

wait how do i do that again?

[untold briar]

Select the text and click the eye button

[rocky spire]

|| spoiler ||

[untold briar]

Or add || at both ends

[austere basin]

ok done

roughly what ballpark was your answer in?

cuz idk if i need to optimise

ah nvm got it!

sry guys another quick question - does the maze have any touching corners that would shut off enclosed sections to each other?

basically what i'm asking is - are all the enclosed squares touching each other?

because if so i can just bfs from a square i know is definitely within the maze

[woven solar]

||very much no, IIRC||

[hidden musk]

throwback 😌. maybe you can judge our code today 🥺

[proud cosmos]

today was fun

[untold briar]

You could see from the examples that this is clearly not the case

But well, I couldnt think of an algorithm either and had to google haha

Algorithm hint: ||Lets say you are trying to determine if a certain point A is within the loop or not. If you draw a line from point A to the edge of the map, it may cut the boundaries once right?||
Hint (continue): ||It also may cut the boundaries more than once, but what are those numbers? Do they have something in common?||

[austere basin]

Haven't looked at the spoilers but I'm thinking of just bfsing from each point

[hazy ermine]

is the pipe structure symmetrical distance-wise?

[simple raven]

day10 I have finally met an unsurmountable challenge 😄 I've spent 2 hours buildings classes to solve part1 but I have no idea how to approach part2. Maybe later.

[untold briar]

Or you can click on my spoilers

I found that algorithm online too, not that I thought of it myself

We are dumb so we need to stand on the shoulders of giants

[simple raven]

I will have to completely rewrite my solution 😦 in hints someone talked about ||the boundary rules|| so I might try to implement that, I'm just sad I cannot use anything from part1

[untold briar]

Yeah there should be many ways to solve this problem

Do not limit yourself to anything

[simple raven]

my part1 was just ||1. make Maze class, 2. make Animal class, 3. let animal run through the maze||

[covert arrow]

So far I've seen 3 approaches for part 2:

• ||go line by line and track how many times you've crossed the path to see if gaps are inside the loop||
• ||floodfill with a scaled up version of the grid so there are gaps between walls for the fill to go through||
• ||floodfill from all points on one side of the loop (relative to direction of travel) since the inside will always be on one side if you traverse in 1 direction||

[lime fulcrum]

part one should be doable?

[untold briar]

Part 1 is just if/else

Also a while loop, or recursion

[lime fulcrum]

my idea ||make a grid. get form of s. follow orders by checking which one i was last and not going there. until back at the start.||

[simple raven]

this is not hints/spoilers channel 😉

[untold briar]

Opps

[lime fulcrum]

i mean if that is a spoiler then you have bigger problems

[untold briar]

But then what are we supposed to discuss here

[simple raven]

here, grab some of these |

[simple raven]

you can discuss just put solution/hints into spoilers

[simple raven]

that's sort-of what I did

[covert arrow]

you can't spoiler some puzzle inputs actually, since discord doesn't like || inside codeblocks

[simple raven]

but you can put it outside the codeblock

[covert arrow]

this is what happens

[simple raven]

for the first option: ||I thought we would have to check both row and column starting from each dirt block, is that not the case? how come we can just do line by line?||

[simple raven]

oh that makes sense for today

[lime fulcrum]

solution: replace all | with I

[covert arrow]

||At the start of the line, you'll be outside of the loop. You can still determine how many times you cross the loop with just that line|| (|||, FJ and L7 will result in crossing once||)

[untold briar]

Well that's what I did. Also, ||the 'lines' aren't required to be horizontal||

[untold briar]

||Though things get complicated if you go through an edge||

[covert arrow]

I haven't tried implementing it myself though, I used the third option

[untold briar]

For example, like this

Those edges coud well, really mess things up

So I did a little twist on the lines

Btw did anyone got 2 loops like this? I think probably not, but just want to ask.

[covert arrow]

nope

[lime fulcrum]

wouldn't that break the starting condition?

[simple raven]

oooooohhh hooold on, ||Can I just use my part1 to run the animal around the loop, turn all dots on, let's say the right side of the animal into O, then once the animal is done running, scan for all O's and turn all neighbouring dots into O's until I run out of O's?||

[lime fulcrum]

because you can't determine what s is?

[simple raven]

yeah it would

[magic oxide]

..|...
--J...
...F--
...|..

is the top right region connected to the bottom left region?

[simple raven]

I believe it is

[covert arrow]

Yeah that would work

[covert arrow]

Yes

[magic oxide]

..|...
--J...
..F---
..|...

how about this?

[covert arrow]

also yes

[magic oxide]

k good

thx

[simple raven]

damn I don't think I have time to solve this today :/

[magic oxide]

not sure if I have time either tbh

need to pack for tmr

[covert arrow]

both are just an extension of this case: ```
..........
.S------7.
.|F----7|.
.||OOOO||.
.||OOOO||.
.|L-7F-J|.
.|II||II|.
.L--JL--J.
..........

[simple raven]

same

[magic oxide]

oh wait I think I misunderstood the problem

what's the inside bit of

F------7
|......|
|.F--7.|
|.|..|.|
|.L--J.|
|......|
L------J
```?

ah wait nvm that can't happen

I see now

good ol Jordan curve theorem

[hidden musk]

fortran is dumb as hell ngl

[magic oxide]

I feel like there's a slightly clever solution here ||using pick's theorem||, but I'm too tired to figure that out

[hidden musk]

yesterday might have been the last roulette day 🥴

[magic oxide]

🥴

you're really gonna leave Fenix alone?

[hidden musk]

stop guilt tripping me 😌 ✊

have been trying to figure out how to read all the input into an array for the past 2 hours

[simple raven]

lol I made a mistake and accidentally filled in all of the correct I's 😄

[trim rain]

i had to take a break for celebrations so i'm gonna continue from day 7 today or tommorow 😦 gotta catch up

[jovial jay]

Same. Day 5...

[low condor]

Today is the most I've written for a solution in quite a while

whole solution is on the Day 10 spoilers channel if anyone cares to look :)

[dense osprey]

i dont want to solve this

but i have to

[low condor]

why do you have to?

[dense osprey]

if other people can solve this, i can

im not mentally challenged

and i want 50 stars this year

[magic oxide]

[low condor]

Just so you know, it's ok not to be able to, or not to want to, solve any particular puzzle

I'm a programmer for a living and some of the puzzles in previous years have really thrown me for a loop

[dense osprey]

oh its a 140x140 matrix

||what if i just floodfill for each tile if i can reach a border without crossing a pipe belonging to the main circle||

would prolly be a naive solution

but a solution

||obviously not floddfilling each tile||

[modern terrace]

how did i not discover this earlier.

[dense osprey]

||if a tile can reach a border, then every tile which can be reached from it can also reach the border||

[low condor]

I have a reply to this but will do so in the hints channel

[untold briar]

||You can totally floodfill, but you need to add a little twist to it.||

Lets continue in aoc-solution-hints channel

[jovial jay]

how do you folks do the spoiler things?

[minor cave]

||spoiler text goes here||

[jovial jay]

||I learned how to do the spoiler thing||

thanks

[hollow wharf]

||spoiler||

||got ya||

[royal gull]

Could you delete this or edit to spoiler ? It’s just a bit of a hint.

[dense osprey]

deleted it sorry

[untold briar]

Also you can try googling. Knowing which key word to put into google is also a good skill for a programmer I think.

I always try to google things myself first before asking for help.

[minor cave]

We don't have that kind of requirement and it's fine to ask that kind of thing for small stuff. Especially because it's relevant in this channel

[untold briar]

Ah I'm not repremanding him or anything, it's just a friendly advice

Since there will be times when you wont be able to ask for help

[jovial jay]

I know about Google things before asking. Just that I didn't felt like opening a browser, then searching at the time

[hollow wharf]

so somehow I started on 2015 AOC.. and completed 3 days lol

that "weather machine" link was confusing and took me back in time lol

time to start over on 2023 Day 1

[jovial jay]

you'll start regretting it on day 5, guaranteed

[hollow wharf]

lol is it more difficult than 2015?

[jovial jay]

never tried AOC before. I only started this month because I can grasp it now

[hollow wharf]

ah okay, yeah some of them seem like they would take me more than a day to do, but I did 3 days last night in 2015, so the beginning ones aren't so hard

[atomic cairn]

Tbh day 5 part 2 has been the hardest challenge so far

Imo atleast

I havent done today yet

[bright haven]

I liked day 10 a lot

[atomic cairn]

It looks fun and quite a challenge i wonder what part 2 is but im not gonna spoil myself

Il maybe do a bit of part 10 before i go to bed tonight otherwise i will try to do day 10 and 11 tommorow

I still need to find a day where i have alot of time for day 5 part 2

[untold briar]

Many here agreed that the difficulty is all over the place this year

[atomic cairn]

It really is

[untold briar]

So you could skip a day and turn back to it later

[atomic cairn]

Day 2 was way essier then day 1

Yesterday was easier then day 5

[hollow wharf]

I did not know you could skip, thats good. This is my first year doing aoc, mostly my first week interacting with this server

[untold briar]

Because day 5 is not necessarily easier than day 6, for example

😏

[untold briar]

You could do it in any order you want

[atomic cairn]

Day 5 broke my streak

[untold briar]

You can also wait until the last day and do it starting from day 25

[hollow wharf]

ahh, the reverse tree trimmer. classic

I heard you get hired automatically at any job you want if you do that

(in santa's workshop)

you could be santa

[untold briar]

Day 26 challenge: write a program that can solve all previous 25 challenges

[hollow wharf]

lul

[cyan sleet]

just a bunch of if statements to determine the day

[untold briar]

Btw I think you should go to spoiler channel or search for answers online

I find AoC to be a very good event for learning

[untold briar]

And lots of copy pasta

Or maybe some import commands if you were hyperly-organized the days before

[atomic cairn]

Oh ye im not coding rn but when im going to im indeed hanging out in there

[honest nacelle]

Don't even know where to start with part 2....

[low condor]

Want to discuss more in #aoc-solution-hints ?

[honest nacelle]

Yeah, I will in just a minute

Actually might just take a break, feeling a bit defeated atm

[low condor]

:( Ok let me know if/when you want to discuss about it

[hazy ermine]

@low condor you made a non-naive solution if i read the chat correctly... can you give some hints regarding that?

[honest nacelle]

god I have a headache

[low condor]

Because of AOC or just in general?

[honest nacelle]

Because of AoC, just having trouble wrapping my head around things today and a lil discouraged because I'm falling behind on my workplace's private leaderboard. I think I need to accept that I won't solve it right now, though, just banging my head on a wall at this point

[low condor]

Remember it’s ok not to solve a problem. 25 straight days of programming puzzles can take their toll

[peak willow]

I have been surprised to see how many different approaches there are to some of these problems.

Obviously some approaches are better or worse than others. What I've been wondering is whether there is a... not so much a "right" answer or approach but perhaps an intended/ideal one?

[bright haven]

that is one of the very interesting aspects of AoC puzzles it that they can be approached different ways.

[honest nacelle]

people have been brute forcing using GPU cores, it's wild

[peak willow]

I mean, pedagogically, that's probably the point, right? That you learn how to match the right tools to the right problems.

[peak willow]

Hahaha WHAT, I haven't seen that yet. That is bonkers and kinda cool actually

[honest nacelle]

My brain knows this, my heart can't help comparing myself to others. The human experience

[bright haven]

i was proud last year that I only earned 39 stars, because it meant i didn't drive myself nuts getting to 50.

[peak willow]

if it helps you feel better I just found this yesterday so I'm still stuck on part 2 of day 1. Although I think while I was lying down under my cat a little while ago I figured out an approach that will probably work well so I'm gonna eat food and then try it out in a bit 😄

[honest nacelle]

I feel like every day I learn that rest is the best thing for problem solving but I never believe it

[jovial jay]

now doing day 7, it's actually fun

[peak willow]

I looked at the hints and saw ONE idea that hadn't occurred to me and then suddenly I had a bunch more ideas. This is really fun! I love Python ❤️

[simple raven]

Gave it a good try but today's been too much for me. See you tomorrow

[earnest forge]

Nice

[low condor]

How do you learn it? Just by reading or by experience?

[atomic cairn]

well uhm very intersting case here

test case 1 works 2 doesnt but my solution is correct

[hidden musk]

@cyan sleet how is the fortran

[lusty rover]

Have you gotten it yet?

[hidden musk]

i gave up

[hidden musk]

i think i'm gonna have to finish my finals before doing more ngl

previous days have eaten up multiple hours per day 🥴

[sleek cave]

part 1 doesn't even seem too hard to me after i solved it

but somehow my code was absolutely riddled with errors during the solving process

spent two hours tearing my hair out while thinking it's supposed to be the easy start

and now part 2 is crazy

the festivities of aoc have truly begun

[flat sequoia]

This is crazy lol should I go back and do every aoc

I haven't done anything before 2020 😧

[jade atlas]

very nice.. I have only 105, exactly 1/4th 😅

[modern terrace]

8 minutes

[magic oxide]

today is going to be interesting

I landed in Japan at exactly 2pm 6 days ago, and my sim card is only for 6 days

so I might not be able to leaderboard because my sim card runs out and I have to change it

well that and also I'm bad at writing code fast

[jade atlas]

you could use wifi somewhere

[magic oxide]

I could but I don't really trust the free wifi

[earnest forge]

Oh that's fun

[pine tiger]

i mean idt aoc takes too much data

[earnest forge]

Their sim expires exactly at unlock lmao

[pine tiger]

oh.

man, that's nasty

[earnest forge]

I very nearly asked what time zone before realising they said japan so I could just google it lmao

[magic oxide]

lmao

[earnest forge]

Anyway GLHF everyone apart from stickie ig

[magic oxide]

yeah I landed right as the problem released

glhf

[modern terrace]

i thought japan was in the same time zone as ph but i realized beijing is too

[shut trellis]

<@&518565788744024082> Good morning! Day 11 is ready to be attempted. View it online now at https://adventofcode.com/2023/day/11. Good luck!

[modern terrace]

and that's far from japan

[minor cave]

Language Roulette: Day 11
The language is ... Perl

[tepid cairn]

oh hey, it's not midnight for me anymore

[flat sequoia]

Done

[hexed axle]

Sub 10 minutes today, wow

[earnest forge]

221/108

I spent too long on a part 1 solution that was completely useless for p2

[minor cave]

#1183636879917195274

[shrewd thicket]

time

[flat sequoia]

NOOOO

😭😭😭

I almost fell into the trap

[pine tiger]

today sucked absolute balls

[modern terrace]

YES

[shut gust]

just discovered about this

[modern terrace]

damn i sucked

[shut gust]

can't work on it until Sunday because of finals though ;_;

[pine tiger]

no p1 just took way, way too long
p2 i had to remember how dijkstra's worked as well

[woven solar]

||what's the dijkstra for||?

||ooh, I see, you used pathfinding||.

[flat sequoia]

Dijkstra 😧

it's an empty grid

[woven solar]

(||pathfinding wasn't necessary, because the distances are just Manhattan||)

[modern terrace]

closest i've ever been to the leaderboards in p1

i seriously had to count the distances across the test input manually to verify if my solution would actually work

that cost a lot of time i think

[flat sequoia]

I ||wrote 1000000 instead of 999999 in p2 which cost me a good 30s||

[covert arrow]

||I was using 100 - 1 for example input and then forgot to change it||

[flat sequoia]

NOOO

[flat sequoia]

Nick finished more than 30s before me tho so I'd only have gained 1 point

[shrewd thicket]

i spend a lot more time trying to figure out if my solution makes sense than just testing the solution

but ig i also use that time to think the solution through

[woven solar]

my template doesn't actually allow a test with different constants to the true input, so I had to skip the part2 test

it's annoying that the day doesn't provide the example answer for the right constant

[pine tiger]

truly a moment of all time

[flat sequoia]

for real

I had to write the right constant, change it to test, and change it back

[jade atlas]

I couldn't figure out why ||(scale * n_prev)|| did not work for me. Then I just tried ||((scale - 1) * n_prev)|| and the results matched. I spent a good amount of time thinking this will not work because the formula did not hold for ||scale=1||. Then I realized part-1 was actually ||scale=2||

[jovial jay]

Days are going too fast for me to catch up

[zenith elbow]

I had a fun time ||scaling up the density of empty space|| before calculating distances

[mossy basin]

scaling up density? you mean lowering the density

[zenith elbow]

ah yup

that's even cooler since empty space is already empty space

[magic oxide]

my method was more akin to ||scaling everything up and then scaling the galaxy space back down||

[cyan sleet]

I didn't have time for coding AoC yesterday 😔

[hollow wharf]

Do i still get 2 stars for doing day 10 today?

[mossy basin]

yep

[hollow wharf]

[shut trellis]

Wrong Channel

You are not allowed to use that command here. Please use the #aoc-bot-commands, #sir-lancebot-playground, #bot-commands channel(s) instead.

[hidden musk]

😦

[royal gull]

Too many holiday parties this weekend. Just finishing 10 p2 now.

Sql was not a good choice for that part.

[ripe meteor]

wasnt yesterday's roulette scala?

wait not that was day 9

[hidden musk]

Fortran day 10

[pine tiger]

god i have no idea why i'm so salty about day 11

this whole event is just a funny lil coding thing

[ripe meteor]

i cant believe i just got timed out by AoC

i've been submitting the sample input

im actually an idjiot 💀

[hollow wharf]

1

2

3

4

[ripe meteor]

youre not getting past the voice gate this way

stop before you get voice banned

[untold briar]

3 2 1 boom

😏

[hollow wharf]

Ok if it’s like this then its ok

How are you doing today?

You’ve ate well enough?

Tell me about yourself

Something interesting

[warm berry]

This is off topic for this channel

[hollow wharf]

Bruv

[warm berry]

And if you try to spam messages you will be voice muted for two weeks

[hollow wharf]

I havent spamed i was being nice 😊

[ripe meteor]

be productive member of the server
be allowed to use voice channels
now go to offtopics please

[warm berry]

⚠️ DO NOT SPAM to meet these requirements, or you will be banned from verifying! ⚠️

Any messages that are sent only to get your message count up are considered spam. To get your message count up without spamming, participate in conversations in ⁠python-discussion, or use the text chats for the voice channels while you listen.

[hollow wharf]

Ok cool

I dont know shit about python tho

Can i say shit? Opps

[slate berry]

How will they fit other problems

[ripe meteor]

we'll be coming down the islands

(i imagine)

[slate berry]

that's why the spaces lol

[peak dock]

lol

[wispy glacier]

that's great haha

[hidden musk]

not even first smh

[wispy glacier]

what's the difference between a sponsor and AoC++?

sponsor is company-backed?

[pine tiger]

even funnier is when you look at their github

[lime fulcrum]

looks like someone is very active

[peak dock]

smh

[compact wasp]

internship hell

[atomic cairn]

You are better off asking that in #python-discussion

[frail pelican]

oh sorry xd

[atomic cairn]

Np

[austere basin]

yo guys, anyone know of a good way to test code for deay 10 part 2? mine works with all the examples but gets the wrong answer on my input

[rancid creek]

is it close?

[austere basin]

how would i know?

[rancid creek]

might be an off by one on the replacement?

are you correct for the*10 and *100 numbers?

[austere basin]

i don't know....

i got 312

[rancid creek]

For the example data.

[austere basin]

ohh okay no the example data is all correct

it gets all of that right

[rancid creek]

ah wait, you're asking about day 10. sorry I was thinking day 11

[austere basin]

ah dwdw

[rancid creek]

which method did you use for part2 ?

and 312 is probably close, there are many different inputs, my andswer for part 2 was in the 300s as well so you're pbably close.

[austere basin]

hi sry for taking a while

the method i used is in #1035199133436354600 :)

[hollow wharf]

cap

[austere basin]

Not cap, it just got shut down ☠️

I can copy paste it and open a new one??

[rancid creek]

Which method is it? That link wasn't to your thread. Have you looked at any of the posted solutions?

[austere basin]

#1183875081919791134

[cyan sleet]

copilot pls

[modern terrace]

wait what

i didn't notice that

oh

[cyan sleet]

we're probably going back down

it's not the first year where days are out of order in the display

[hidden musk]

you did it?

[cyan sleet]

I did

[hidden musk]

incredible

[cyan sleet]

it's not...terrible

but I miss so much stuff

[hidden musk]

🤨

[cyan sleet]

it's very basic, feels very C-like in that way

I very much miss generics

[hidden musk]

yeah. i mean, it is older by like 20 years

[cyan sleet]

I did end up making some macro terribleness to shorten some code

[hazy ermine]

man I wish I could play along

[iron jewel]

guys i need help

i mean a question

[rancid creek]

Why can't you ?

[iron jewel]

what is the language use to make this interface?

[minor cave]

this is not the right channel. Try #ot2-never-nester’s-nightmare

[honest nacelle]

okay, taking a crack at today, wish me luck

I've given up on our private leaderboard for the moment, think it's healthier that way

[low condor]

best of luck!

[zenith elbow]

finally finished day 10 p2

completely missed that ||we're not just looking for . but also for pipes that aren't part of the path||

[low condor]

Advent of Reading Comprehension ;)

[magic oxide]

honestly it's a blessing that ||those pipes not part of the path also counted as spaces to check||

||otherwise pick wouldn't have worked||

[edgy dirge]

??

[shadow musk]

hola

[hazy ermine]

exams 😔

[lusty rover]

hola

[rancid creek]

hullo

[lusty rover]

como ce va?

or however you spell it

[rancid creek]

trouvé et toi?

ढूंढें और आप?
找到了，你呢？

fun with google

[hexed axle]

how is it the 12th day already?

[earnest forge]

GL everyone

[covert arrow]

GLHF

[zenith elbow]

24 hours is pretty fast :)

[peak dock]

comment ca va

[royal gull]

May you all get 12 lords a leaping

Or 12 drummers drumming

[shut trellis]

<@&518565788744024082> Good morning! Day 12 is ready to be attempted. View it online now at https://adventofcode.com/2023/day/12. Good luck!

[magic oxide]

bruh

I'm halfway up a mountain

[minor cave]

Language Roulette: Day 12
The language is ... PHP

[magic oxide]

lmao

[tepid cairn]

😩

[graceful dawn]

I really need to get on these challenges, finals have been kicking my butt

[covert arrow]

||p1 is bruteforcable||

[lapis prism]

ping

[jovial jay]

I have no hope to complete it this month either, but that's ok. I enjoyed what I can do

[hexed axle]

I've only gotten one star, and that was day 1. When I get time I'll sit down and worth through some of them. I think I prefer it as a list of fun puzzles over a timed event personally

solely based on the leaderboard, todays part 2 looks like a doozy

[stable shale]

Meanwhile I'm still trying to figure out whether to ||bruteforce|| P1 or try ||math|| (not really a spoiler, but just in case)

[timber glacier]

this is a confusing one what

I legit don't understand how to even get started

[stable shale]

Join the group of waiting for the leaderboard to fill so we can check Reddit

[timber glacier]

I just want hints not spoilers

[stable shale]

Yeah, read the hints some comments have, or pick a really confusing solution that you can't understand, and don't look too deep at it

[hexed axle]

Not the longest, but nearly 23 minutes to fill the leaderboard is pretty good

[stable shale]

P2 is gonna be horrifying, isn't it

[minor cave]

#1184002671103975506

[royal gull]

Ok, this is a "I'm not solving this tonight, I'm going to bed" day.

[stable shale]

Smart

[magic oxide]

I was gonna do aoc at the top of the mountain

and then my Sim card ran out

[covert arrow]

ayy I'm back on the global lb

[pine tiger]

massive W

[magic oxide]

aren't onsens also hot springs

ohh I see

lmao

[flat sequoia]

Bro I'm trolling

idk why my code doesn't work

[pine tiger]

i thought you fullsolved already 💀

[flat sequoia]

nop

I tried to do something clever

but it was wrong

so I wrote the dp

but it's not working

sobbing

i'm getting something ridiculously large 😧

[earnest forge]

Finally got something that runs in reasonable time and works for the sample, but it doesn't work for my input

._.

Ok nvm I wrote a bad optimisation

573/983 oof