advent-of-code

[candid berry]

so, if the goal is to add 2 binary numbers using only 3 types of gates, you're essentially wiring up a bunch of full adders

there's 8 wires that need swapping, and ||3 of them are obviously wrong because the wrong type of gate is directly connected to the output||

you can also get pretty far by figuring out ||what the least significant bit with a wrong wire connected to it is||

if you set x to ||all 1 bits|| and toggle y between ||all 0s|| and ||all 0s except a 1 as the lsb|| then it's easy to see if addition is working for the N least significant bits

also it's helpful to know that ||there aren't any extra gates||

[jagged widget]

what is this even about

[hollow wharf]

that @jagged widget

[jade cedar]

|| 😭 ||

[dapper pawn]

.

[ripe meteor]

this is not on topic for the channel

[dapper pawn]

ok sorry

[hollow wharf]

Hey sup

[sacred oriole]

😀

[waxen minnow]

Hi

[paper folio]

@minor cave Is it too late to join the leaderboard for 2024?

[lime hill]

hi

[paper folio]

I'm not able to get the join code with the /aoc join command. I already completed the challenges, and wanted to add my self to the leaderboard for the role

[solar tinsel]

Tf

[minor cave]

I needed to update the cookies which is why the command errord. Can you try again?
I don't remember if we kill the leaderboard join in December of January though

[paper folio]

Ty! I think it worked!

[paper folio]

woot woot got the flair!

[hot scarab]

What does mean socket.AF_INET? I mean AF on front..

What is AF?

[rapid wolf]

Address Family, however, this is not the right channel, please read #❓｜how-to-get-help

@hot scarab^

[long bear]

How do I unlock the strings for my discord bot?

I need to unlock firewall as well, I use python on desktop

My firewall is corrupted

[tender swallow]

@long bear is this related to Advent of Code? if not, please ask in the general #python-discussion, or perhaps #discord-bots

[fathom kiln]

Hello guys

Nice to meet you

[rancid creek]

Hullo! Channel is a lot more active in December during the event.

[rancid creek]

But feel free to ask about any of the puzzles anyway. 🙂

[fathom kiln]

Sorry I'm new to discord I'm learning a little python for programming

[rancid creek]

Cool. Doing adventofcode puzzles for fun/learning?

[fathom kiln]

Yes

[rancid creek]

Some of them get impossibly hard for new coders, don't let that discourage you, it's ok to google how, or ask questions so that you can keep moving forward. Or just focus on the easier puzzles.

[fathom kiln]

Tank you

[solar tinsel]

[pulsar parrot]

hi im completely new to this and joind this server to find some help

idk if im in the right chat room for this message ether btw

[distant igloo]

#❓｜how-to-get-help

[tepid knot]

how can you make your message appear like this in discord?

[hybrid gull]

Use triple backticks (same key as ~) in either side of the pasted code and it will allow code formatting for multiple lines of text.

[vast schooner]

Is there a channel for GSoC 2025 contributer in python?

Does anyone contribute or interested in contributing to python this year?

[lusty rover]

What is that?

[lusty rover]

I mean, everyone in this server is already doing Python code

[distant igloo]

Google summer of code I presume

[uncut cypress]

I think they mean contributing to actually updating the language itself in newer versions

[lusty rover]

<@&831776746206265384>

[minor cave]

!cban 728561488536535102 Not fit for our community

[marsh current]

:incoming_envelope: :ok_hand: applied ban to @zinc bane permanently.

[hardy relic]

I'm just looking into AOC. Do you do all the workings in your own IDE or on the website

[ripe meteor]

Theres no way of solving the puzzles on the website itself

[hardy relic]

OK just the answer perfect thanks

And does the puzzle input always just strings ?

Like aoc2024-1

[tender swallow]

yes, it's always ASCII text.

[hardy relic]

Hmmm ok

[hardy relic]

Struggling with this first one haha

I guess they ain't easy

[normal loom]

no, some of them require a lot of thought

[hardy relic]

Nice

[hardy relic]

Dam I'm.so close i think lol It's not working though

[hardy relic]

what am i doing wrong here: for AOC2024-1

    answer: int
    ab = input.replace("   ","")
    ac = ",".join([ab[i:i+5] for i in range(0, len(ab), 5)])
    x = ac.rsplit(",")
    l1 = len(x)
    l2 = int(l1 / 2)
    ab1 = 0
    ab2 = 1

    for i in range(l2):

        if int(x[ab1]) > int(x[ab2]):
            higher = int(x[ab1])
            lower = int(x[ab2])
        
        else:
            higher = int(x[ab2])
            lower = int(x[ab1])
        
        z = higher-lower
        answer =+ z
        ab1 += 2
        ab2 += 2

    print(answer)

a = input()
calc(a)

[solar tinsel]

answer =+ z

Type of answer is defined but not the default value

[hardy relic]

hhmm

[solar tinsel]

Are you giving the input by pasting it in?

[hardy relic]

yeh im pasting in the massive string as 1 line,

this also doesnt work - give the incorrect numner:

    answer = 0
    ab = input.replace("   ","")
    ac = ",".join([ab[i:i+5] for i in range(0, len(ab), 5)])
    x = ac.rsplit(",")
    l1 = len(x)
    l2 = int(l1 / 2)
    ab1 = 0
    ab2 = 1

    for i in range(l2):

        if int(x[ab1]) > int(x[ab2]):
            higher = int(x[ab1])
            lower = int(x[ab2])
        
        else:
            higher = int(x[ab2])
            lower = int(x[ab1])
        
        z = int(higher-lower)
        answer += z
        ab1 += 2
        ab2 += 2

    print(answer)

a = input()
calc(a)

its working as far as, i can get the numbers seperated into a list, then i must be messing up somewhere calculation the difference of adding all the differeneces

[hardy relic]

holy shtit, im not doing it lowest to hightest

[hardy relic]

now i realised that my list is not all of the input... why would that be?

    answer = 0
    left = []
    right = []
    
    # Remove new lines
    ab = input.replace("\n", "")
    # Remove the spaces
    ab = ab.replace("   ", "")
    
    ac = ",".join([ab[i:i+5] for i in range(0, len(ab), 5)]) # add a comma between each number is the string
    x = ac.split(",") #create new lsit with each number as seperate
    #list = x.sort() # sorted list

    print("Length of full list:", len(x))
    print("First few elements:", x[:10])
    print("Last few elements:", x[-10:])

    ab1 = 0
    for i in range(len(x)):
        if ab1 % 2 != 0:
            left.append(x[ab1])
        else: 
            right.append(x[ab1])
        ab1 += 1

    print("Left length:", len(left))
    print("Right length:", len(right))

    return left, right

a = input()
calc(a)

[hardy relic]

ok and now in 5 hours i and realise that input has a max limit that im goingbeyond. wow

these are the fun mistakes that you will never do again

[cyan sleet]

input doesn't have a max limit

[ripe meteor]

Just use a file

Pasting a 1000char string into a terminal is insane

[hardy relic]

Yeh I'm guan look into that but I'm sure that will require a total restart on the code ha daym

[vestal mural]

Can anyone here help me with my code? I'm running into a problem where I don't know what to do, It's with lists and inputs and such and I can't seem to get it to work.

[marsh current]

Please react with ✅ to upload your file(s) to our paste bin, which is more accessible for some users.

[hardy relic]

Will a .txt file read like I pasted it?

[tender swallow]

this doesn't seem to be related to AoC. please see #❓｜how-to-get-help

[mortal bronze]

!paste

[hardy relic]

@solar tinsel i did it! day 1 complete. after using the file and a few tweaks! onto the next!

[solar tinsel]

Gl hf

[hardy relic]

Oossh number 2 is making me loopy

[glacial anvil]

number 2 for days

[sullen adder]

does advent of code have a non-oauth option?

[lusty rover]

I don’t think so

[sullen adder]

unfortunate, github suspended my account 😮‍💨

[dense gale]

&aoc link zak799

[shut trellis]

You have linked your Discord ID to zak799.

[minor cave]

Reddit accounts are generally easy to make

[sullen adder]

thanks

[strange zinc]

Hallo again people 🙂

[clear sundial]

Hey everyone! I’m Caleb. I’ll be doing Advent of Code in Python again this year (and maybe some TypeScript on the side).

As part of my own workflow, I put together a small CLI called "elf" that handles fetching inputs, caching them, submitting answers safely, and checking private leaderboards. Happy to share it in case it helps anyone.

GitHub: https://github.com/cak/elf
PyPI: https://pypi.org/project/elf/

Looking forward to hanging out here, comparing approaches, and seeing how everyone tackles the puzzles this year.

Happy coding and good luck on Day 1!

[lusty rover]

I'm super suprised that name wasn't already taken or reserved

[visual ridge]

What name mine?

I'm new here.

[lusty rover]

elf on PyPI

https://en.wikipedia.org/wiki/Executable_and_Linkable_Format

[visual ridge]

Oh, I'm new to python (and coming from AutoHotKey v2 after mastering it)

[terse comet]

it was

(that 0.1dev version looks like a project initialized from a template with zero changes though)

[lusty rover]

Yeah

http://pylonshq.com

Seems like the template company still exists

[clear sundial]

Yeah, @lusty rover and @terse comet it was! I had reached out the owner since it wasn’t being used since registered in 2014. I’m glad to put it use, Advent of Code is much more fun than file formats, imho. 😉

So thanks to Solos for the great package name!

[waxen wyvern]

hiii

Can every problem can by solved using python only?

[normal loom]

yes

[waxen wyvern]

:D** **

How hard is solving the problemas of the lats days?

I'm new in python, it's my first time in an advent-of-code, and I'm happy for finding it :)** **

[normal loom]

they do get fairly tricky. Don't worry about whether or not you will be able to complete all of them, just see how far you can get

[zinc kindle]

And also it's rare that a problem on one day is a prerequisite for a different day. Most of the time if you get stuck on e.g. day 7, you can still do days 8, 9 etc.

[lyric sigil]

i will miss being able to do aoc as soon as it releases..

[bright spindle]

I don't think you generally need to be a genius to solve all of them, but I think it'll be tough if you're not pretty experienced.

But yeah, try your best anyway

See how much you can solve

[waxen wyvern]

🤠👍

[tender swallow]

last year I managed to get up to day 20 solving everything
after that I couldn't do the part 2s anymore

[bright spindle]

I completed two previous years, and I don't feel like any problems were super hard. They don't require super advanced or obscure algorithms. Stuff you'll learn in an intro to DSA at most.

And most aren't even on that level.

[tender swallow]

a good chunk is just knowing when to apply memoization/caching

[bright spindle]

Yeah

[ripe meteor]

Last year one of them needed you to look at the input

With your eyes

[normal loom]

🎄

[terse comet]

in some years there were a few pretty bad ones

but yeah, give it a try and skip or stop if you must
(that post I linked is very old too)

[bright spindle]

Hm, whether they are bad or not seems to be debatable.

Some people in that thread complained about puzzles that take a lot of time, but I feel like it's kinda traditional to have one or two days like that.

[bright spindle]

Some people also don't seem to like it when you get a big advantage from manually looking at or preprocessing the input, which I feel is a little arbitrary.

[zinc kindle]

I don't think there were any last year, but usually there's a day where part1 requires writing an emulator for a toy programming language (where the input is code in that language). part2 then tends to require studying the given code by hand and figuring out the algorithm. This gave me a lot of trouble my first few years.

[cyan sleet]

beverage bandits was implementation pain

[runic drift]

I see there are only 12 puzzles this year, is it just the first 12 days of December?

[minor cave]

Yup

[open elbow]

I've never tried advent of code seems fun

[late portal]

Hi

My 2-3 advent of code, never completed it though
I hope to complete it this year 🙂

[worthy horizon]

I will be boring and do it all in no-std Rust this time

[open elbow]

im gonna do it in python because I don't know any other language 😎

[zinc kindle]

I plan to do it in Haskell this year, but I reserve the right to retreat to Python if I'm having trouble and redo it in haskell later (probably. I remember one time I spent ages trying to parse an input in Rust, before giving up and just using Python's ast.literal_eval)

[limpid grail]

omg AoC today

[limpid grail]

im doing rust too but not no-std

i will use this as an opportunity to further my knowledge in rust :ferrisCat:

maybbe i could try no-std after i do the std solution

i will be competing with my friends afterall
||this was just a placeholder challenge to print double the value you read from a file||

so excited

[remote mauve]

hello

python...

I will do it like 6 hours or 16 hours after it releases

[open elbow]

advent of code in an hr 21m

[rancid creek]

Hullooooooo all.

[open elbow]

1 hour remains

[rancid creek]

fun thing to experiment with then is seeing which new libraries will be helpful, if you don't know a lot of numpy aoc is a good place to practice.

[open elbow]

got it

[silent nymph]

Same plan

[open elbow]

aoc in 15 mintues

10 mins now

glhf guys

[covert arrow]

glgl

[open elbow]

hehe im the first one to do it here

1 MIN

Excel????

[magic oxide]

oh god

[open elbow]

😭

yeah imma just stick to python for now lol

[lusty rover]

I HAVE FOUND MY PEOPLE

[covert arrow]

im gonna speedsolve in python first

[open elbow]

less than a min

30 secs

[lusty rover]

Doing them in Excel is fun
https://github.com/shenanigansd/scratchpad/blob/main/events/advent_of_code/2015/01/excel/excel_day01.xlsx

[magic oxide]

welp i guess i get to procrastinate setting up my repo another day

[open elbow]

10 secs

[shut trellis]

<@&518565788744024082> Good morning! Day 1 is ready to be attempted. View it online now at https://adventofcode.com/2025/day/1. Good luck!

[open elbow]

oh hod

oh god

[topaz summit]

<@&518565788744024082>

[magic oxide]

how does one parse in microsoft excel

[sullen adder]

i misread the question already...

[potent pumice]

on it

[plush peak]

oh em gee

can someone who needs to freshen up on python attempt this

[oak marten]

im so cooked

[lusty rover]

Hmm
This is not a thing I know how to do well

[ivory marsh]

Part 1 looks straightforwards, so I'm scared for the part 2 twist

[open elbow]

aghghg I already got it wrong once

[sullen adder]

i did so bad on day 1 but it was fun ;;

[open elbow]

twice

i've got the gist of it, i'm stuck on how to get the proper answer tho

AHH I SEE

[gentle cape]

Wait aoc day 1 is out?

[ivory marsh]

Yep

[lusty rover]

yes

[old badge]

i really should sleep but this looks easy enough to try

[gentle cape]

Oh shi

[old badge]

it's harder than it looks i assume

[gentle cape]

Time to actually do it then I guess

[lusty rover]

God that was annoying

I'm gonna end up tapping out really early this year

[open elbow]

yeah so am i 😔

[ivory marsh]

It's always the worst when you get the correct answer from part one, but seem to be doing it completely wrong because the example makes 0 sense

[magic oxide]

doing this in excel isn't that bad actually

[lusty rover]

Oh you're already doing Excel lol

[magic oxide]

yeah i'm not doing python sols if i can help it

just roulette lang only

[lusty rover]

oh?

[open elbow]

I'm getting so close to replicating the example

[pseudo fulcrum]

what is advent of code ?

[potent jacinth]

Day 1 is dope. I can't find the document: I'm on phone. But just 6 hours ago I finished making a Python-like document parser in C that advances line by line, so if the document has multiple lines I just need to read the first byte and from there it's easy.

Time to figure out how to access the document on mobile

[potent jacinth]

Fun.

[lusty rover]

https://adventofcode.com/2025/about

lol the sponsor I got is my new vendor at work

[open elbow]

ok for some reason it's giving me the correct answer for the test data but not the actual data?

[lusty rover]

Edge cases!

[open elbow]

yeah that's what im tryna figure out

wow only 4 ppl so far

i'm beginning to give up this is actually pretty difficult

[old badge]

i got part 1 but part 2 is uuuogh

[open elbow]

i'm stuck on part one due to the edge cases

[magic oxide]

there are edge cases on p1?

[open elbow]

or im just stupid

goddamn it

my code works on the example, but not on the actual data

I NEED A HINTTTT

[tender swallow]

you can post your solution (with spoiler tags) in #aoc-solution-hints and ask for hints from there

[solid escarp]

Hey, question, doesn't going from 0 to 95(L5)
gives us a point?
like the lock does point to zero since it is circular, right?

[loud slate]

you're starting at 0 and turning left 5. You don't count 0 twice

[solid escarp]

aaah

that makes sense

me realising my solution was low, now handling the edge case, it is even lower

[half swan]

same

idk how I'm getting my ass handed to me so hard checked addition shouldn't be this hard 😭

[worthy horizon]

going back to my fix error roots

[naive wharf]

[day 1 part 2]||```py
n = 50
zeros = 0
for i in ndata:
for j in range(abs(i)):
if i < 0:
k = -1
else:
k = 1
n += k
n %= 100
if n == 0:
zeros += 1

we don't need efficiency where we're going!||

[half swan]

#1444916008325349478

[bright spindle]

||Yeah, I ended up resorting to that after struggling with modular arithmetics for way too long.||

||I had a solution that used modulo that ended up being off from the correct answer by only a little, so I assume I somehow failed to account for some obscure special case.||

[pulsar drum]

|| i also just did the "stupid" way and looped one by one, brain wasnt braining for figuring out the actual math ||

[tender swallow]

||did the manual simulation to get the star, then figured out the modular stuff at leisure||

[mossy basin]

new spoiler channel

[strange hawk]

true

[worthy horizon]

update: I had a single off by one error. Thankfully valgrind exists

[craggy cairn]

Hello everyone

[quaint comet]

I got it right on the first try 😄

[waxen wyvern]

First day, first part, and it's already not easy 😂

If it's 50, and them goes "R50", it stops in 0, right?

[torn sapphire]

yep

[half swan]

it might help drawing the dial if you can't visualize it

[lime fulcrum]

I felt like part 1 was somewhat easy. Part 2 I am still thinking about right now

[tender swallow]

?

[late portal]

day 1 puzzle was so fun

i was dumb, i was using 0 as start position

[vast jetty]

bro doesnt have the repo public

so sad

[bright spindle]

I do think that this year's day 1 is a bit harder than in previous years. Maybe it's because there are fewer days, so each one is a bit more involved.

[vast jetty]

its exactly because of that

[vast jetty]

if you wanna, post the solution!

[late portal]

where?

[vast jetty]

in spoilers ig

[late portal]

kk

[vast jetty]

altho really just make a public repo?

[late portal]

can we post it there from other langs than python?

[vast jetty]

ok fine i didnt make a public repo either

but its cuz its easier to just commit everything in my workspace which includes inputs and some throwaway files

[vast jetty]

doubt it

do the repo thing then

[late portal]

kk

[torn sapphire]

you can post stuff in languages other than python in our AoC channels

[vast jetty]

ah ok

[torn sapphire]

as long as you're not spoiling people outside the spoiler channels, of course

[late portal]

ill just make a repo

my code a bit big

i am bad at code golfing

[pulsar drum]

just rememnber to not include the question, or input

[shy tulip]

I'm trying it in haskell and have absolutly no idea what i'm doing 😅

[ripe meteor]

Average Haskell moment

[shy tulip]

Yeah i'm getting type errors but idk why

I had the acc and the input the wrong way round 😩

[bright spindle]

Why?

[bright spindle]

Going through an AoC in a new language is a good way to warm up

[tender swallow]

Eric Wastl requests people to not reupload puzzle text or full inputs

[bright spindle]

Ah

[naive wharf]

AoC in prolog...

[worthy horizon]

I decided to do it all in no_std Rust by just interfacing with libc. It turned out to be just enough new stuff and just enough old stuff

turns out the rust compiler is very fast if you don't make it compile millions of lines of code

I got to play with errno and segmentation faults and then add enough rustisms on top of that such that my solution code doesn't have any unsafe

[dusk wraith]

Does someone wants to Group up for Advent of code?

[ripe meteor]

AoC is not a group activity

[dusk wraith]

But it can be to brain storm

[thorn inlet]

crust

[worthy horizon]

nah, I'm not doing crust

I'm just building safe abstractions myself

(as the language was intended to be used)

[potent zealot]

part 2 man
checking for double counting while not having any visible way to track my mistakes on the 4k line test cases is rly sth
im doing this in c++ too

[viral grotto]

Am i the only 1 who created 2 func rotate left and right and incremented it by 1😭

[viral grotto]

How did u manage to write 4k lines it's like 7

[potent zealot]

not write, i mean the input.txt is 4k lines long

[viral grotto]

Oh kk

[pulsar drum]

if its a public repo, the puzzle inputs or texts shud not be included

but for private repos it shudnt be a problem

[white arrow]

how do i apply for the advent of code and see the challenge?

[tender swallow]

click [ Log In ] and log in using one of the providers

[white arrow]

i have done that

that’s all i’ve done

[tender swallow]

go to https://adventofcode.com/ and click on the first row

[white arrow]

awesome thank you

[void forge]

Completed Day 1

[white arrow]

im kinda confused

ive read the puzzle

ah ive just seen the document.

[lusty rover]

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

[solid escarp]

that glowing star on your screen is motivating me to complete day 1

[cyan sleet]

I have a few

[solid escarp]

🤩

This year i will complete my tree for sure

[deft terrace]

WOO!!

[white arrow]

How do i share my solutions to problems etc?

I've signed up

[tender swallow]

have you solved the problem, got the star?

[white arrow]

no i have no clue lol

just wondering how i do

[tender swallow]

you take your input, write code according to the problem statement, feed your input into your code, then type the answer into the website. which will tell you if you got it right or not.

[white arrow]

ah is that literally it

could i just make my own github repo to share my solutions or is that not allowed?

[void siren]

Yep, you can. Just make that clear and don't share stuff like the puzzle text or the inputs since you can't redistribute that.

[white arrow]

By puzzle text, you mean like the description of the task?

And my inputs are like, the one that the document contains

[void siren]

Yes.

[white arrow]

okay got it thank you

[minor cave]

Also, for folks interested in Language Roulette, because there are much fewer days this year I'd like to prioritize languages people are interested in trying. So feel free to post requests here!

[white arrow]

can i get help with my code for the challenge or not?

also i need to know how i download this file lol

[tender swallow]

yes, you can ask in #aoc-solution-hints
please use spoiler tags appropriately!

[tender swallow]

right click > Save as should work I guess

[white arrow]

its the one with all the L, R stuff on it

from the website

[tender swallow]

yeah, right click on the "get your input" (etc) link and choose Save As

alternatively just select all and copy-paste it into a file in your editor

[white arrow]

awesome thank you

i also have a question about git

if my file structure is

aoc-2025
  - Day 1
    - input.txt
  - .gitignore

if i put input.txt in .gitignore

when i push, will it be ignored even though they're not in the same directory

[tender swallow]

it will be ignored yeah

[white arrow]

awesome

[bright spindle]

Hm, don't you need to specify the path?

Maybe not

[tender swallow]

nope
my gitignore for AoC is just

__pycache__/
*.py[cdo]
input
small*input*
others/

[bright spindle]

Seems like there's special handling for filenames without leading paths

[white arrow]

i just added input.txt to mine

[tender swallow]

also TIL (based)pyright treats module resolution differently based on whether you're in a git repo or not

[white arrow]

also, does the dial always start at 50?

[tender swallow]

yes

[white arrow]

okay cool

[wanton plover]

Today's pt2 stressed me out to no end i just gave up and decided to do the bruteforce and it annoys me i couldn't figure out a cleaner way it would always be off by like 30

[worthy horizon]

I just did it the very dumb way

[bright spindle]

I kinda ended up feeling like the brute force was the cleaner way

[shy tulip]

I'm trying it in a new language and i'm not sure if my logic is wrong i'm just not used to how this works 😅

[earnest forge]

||divmod|| is the 'secret' to doing it the cleaner way. My comp solve also did it the brute force way lol

[tender swallow]

a lot of times the brute force is the more... I'm not sure, naturalistic? skeuomorphic? way

[earnest forge]

Intuitive is the word I'd use

Or, depending on mood, naïve

[bright spindle]

||I mean, the code for the naive approach is easier to read than even nicely written non-naive solutions. And in this case it doesn't need to be optimized at all.||

[earnest forge]

Also I assume everyone noticed that [not solution related] ||the password method we used is the 'CLICK' method||

[white arrow]

can anyone help me lol im having problems with types

[tender swallow]

I mean that it closely follows whatever real-world object we're trying to represent

and in that sense it feels more elegant than a clever shortcut to sidestep the real-world

[bright spindle]

Indeed, though the point of a lot of AoC problems is that you have to be more clever than that. ||Not this one though.||

[earnest forge]

||Not this one though.||
I mean, typical of D1

[bright spindle]

As usual, it's quicker if you just ask.

[white arrow]

i put it in solution hints :)

[bright spindle]

I suppose

[dusty trout]

finally

the second part was tough

[void forge]

Awesome! You should!

[prime vector]

&aoc link aocleaderboardnamegoeshere

[shut trellis]

You have linked your Discord ID to aocleaderboardnamegoeshere.

[prime vector]

&aoc link OneidaGui

[shut trellis]

Your linked account has been changed to OneidaGui.

[cyan sleet]

imo it's a bit sad when there is a lesson to be learned that gets missed because brute force is feasible

[bright spindle]

Well, it's also nice that AoC is accessible.

There's pros and cons.

[cyan sleet]

I just hope we get a good handful of problems to silence the "I kept brute forcing and it kept working" crowd

[tender swallow]

we'll get punishing inputs eventually I'm sure

[bright spindle]

There usually are.

[warm berry]

inb4 ||recursion stones||

[tender swallow]

I really enjoyed that one last year

[cyan sleet]

which one was that?

[tender swallow]

2024 day 11

[cyan sleet]

ooh, dlang

I recall that one being fine

iirc the main insight was just that things are order independent and that you should handle things in bulk

https://adventofcode.com/2021/day/6 was a good early one to mess with people

https://adventofcode.com/2019/day/22 is probably my favorite in that regard

[remote mauve]

Wohooo

just finished day 1

btw @minor cave is it 12 stars for @Advent Of Code Completionist role?

[bright spindle]

I'm re-reading my solution to that one and I don't get it, lol.

[earnest forge]

Would be 24

[remote mauve]

so both parts everyday?

[earnest forge]

Yes, previously it's been 50*

[cyan sleet]

basically work with counts per timer stage rather than a huge list

[earnest forge]

And tbf it is called 'completionist'

[cyan sleet]

completionisht

[bright spindle]

Yeah, I know that part.

I just don't understand the way I'm calculating the counts.

[cyan sleet]

now I'm curious

[bright spindle]

fn evolve(data: &Vec<i128>, max_gen: usize) -> i128 {
    let mut data = data.clone();
    for i in 0..max_gen {
        data[(i + 7) % 9] += data[i % 9];
    }
    data.iter().sum()
}

Ah, hmm

I guess (i + 7) % 9 represents that which will be 8 on the next day on day i

And i % 9 is 0 on day i

or something like that

I dunno man

[earnest forge]

This looks like a DP solution to me

[bright spindle]

Speaking of solutions that don't correspond to the natural representation of the problem...

[cyan sleet]

here is the relevant part of my code

  for _ in 0..256 {
    let zeroes = counts[0];
    counts[0] = counts[1];
    counts[1] = counts[2];
    counts[2] = counts[3];
    counts[3] = counts[4];
    counts[4] = counts[5];
    counts[5] = counts[6];
    counts[6] = counts[7] + zeroes;
    counts[7] = counts[8];
    counts[8] = zeroes;
  }

[ripe meteor]

part2 took way longer than i would have liked it to

washed gang

[cyan sleet]

you just make use of things being cyclic

[bright spindle]

That's what everyone says

[worthy horizon]

(btw, you should almost never take &Vec as a parameter, it's better to take &[i128])

[bright spindle]

I haven't written a line of Rust since AoC 2021

[earnest forge]

Oh wait no this is actually really simple you're just doing the whole 'shift everything down' but you're rolling around the array without actually shifting the elements. Your array has 9 places, which i cycles around, and then i+7 is to reset the timer on the current slot, which has effectively been replaced by its children. It's actually a super clever optimisation, but yeah rather inscrutable unless you really think about it

[cyan sleet]

so yeah, you are adding the counter 6 fish, the the value at i%9 automagically becomes the next 8 counter

it's clever

[bright spindle]

Yes. I am clever. Or well, past me anyway. Not current me.

[earnest forge]

In this context I can see why, .clone() is more obvious than .to_owned() if you're not too familiar with the language and you can't .clone() a &[T] to get anything useful

[worthy horizon]

slices have .to_vec() I believe

[earnest forge]

This is how I feel when looking at my overall trend on global AoC leaderboards (until this year...)

[bright spindle]

I wonder how I came up with this, tbh.

[earnest forge]

They do indeed. Definitely did not remember that and that kinda helps my point tbh

[bright spindle]

I did it in Rust to learn the language, so I definitely fall in the "not too familiar with the language" category.

[earnest forge]

I mean it's a reasonably logical combination of a circular array data structure and the counter-based method of solving the problem, it's just not immediately obvious to make that optimisation

[worthy horizon]

(sorry if I'm being pedantic, I thought you were doing AoC in Rust this year so that might have been relevant informaiton)

[bright spindle]

Oh yeah, actually, if you compare your solution with mine, it's pretty easy to see that they're equivalent.

[bright spindle]

It's ok to be pedantic, it's a programming education server after all.

[worthy horizon]

to be pedantic, I wasn't being pedantic, rather I was providing unsolicited advice

well now I am being pedantic

[bright spindle]

I wasn't asking!

🙂

[worthy horizon]

soul staring smiley 🙂

[earnest forge]

I mean in fairness it is one of the things that cargo clippy -W clippy::pedantic would tell you about :P

[bright spindle]

I think perhaps I did something like what @cyan sleet did initially and then collapsed it into the above.

[earnest forge]

That seems likely, I can't imagine anyone would think of the circular array thing first lol

[cyan sleet]

I was focusing on learning rust that year

(and the solution was so fast that I didn't revisit it to optimize)

[open elbow]

is somebody able to come to #aoc-solution-hints to help me with my problem?

[limpid grail]

next AoC challenge soon

is this how i link my acc for the aoc completionist role too
#aoc-faq message

/aoc join

oh its aoc link

[mossy basin]

gonna fall asleep before todays problem

[lusty rover]

SLACKER

[open elbow]

ok this time for today's puzzle I'm going to reread it 10 times

[storm rampart]

[lusty rover]

oh hey

Based

[sullen adder]

yippee day 2

[shut trellis]

<@&518565788744024082> Good morning! Day 2 is ready to be attempted. View it online now at https://adventofcode.com/2025/day/2. Good luck!

[open elbow]

oh this ones a good one

a really easy but slow way to do this would be in o of n squared time

[late frigate]

time for the yearly lurk on the python discord server until AOC

[hollow summit]

95-115 has one invalid ID, 99.

what

isn't it 99 and 111

[open elbow]

no because the pattern has to be repeated twice and only twice

so for 1111 - 2222 it would ofc be 1111 and 2222 or 1919 or 1818 or 1717 or 1616 or 1515 or 1414 and so on

this is a tougher one due to the ambiguity of the combos

[sullen adder]

wow part 2 really stumped me @~@

[magic oxide]

clojure is not bad

[lusty rover]

Okay parentheses man

[open elbow]

ah i see a pattern now

[lusty rover]

It can get pretty terse

You probably love that you math guy

But I just can't read it

[open elbow]

oh nvm aghghg

oh lmao I already found a working case

[half swan]

regex ftw

[open elbow]

aww damn it part 2 actually kinda has me stumped rn

[naive wharf]

can you tell me the changes?

[onyx aspen]

what is advent of code ._.

[atomic cairn]

Check #aoc-faq

[ivory vigil]

That felt easier than d1

[limpid grail]

is someone done with part 1 and part 2

[magic oxide]

there are plenty of people done in the spoiler thread

[limpid grail]

i dont want to spoil it for myself, but id like your input and outputs for this personal leaderboard im hosting for my hostel mates

[ivory vigil]

Yes

[limpid grail]

i mean i can post my input but its kinda pointless without the outputs heh

and i need to solve it

i like the stories

does anybody else enjoy reading the stories

[mossy basin]

nope, you're the only one

[limpid grail]

but its so cute

i documented 2015's
https://github.com/blankRiot96/aoc15/blob/main/story.md

i stopped documenting after day 14 but i did read the rest too

[limpid grail]

so uhm, any takers

[hollow summit]

stupid elf kid

why would you let your elf child go on the shop computer

i just had to filter out invalid ids using python

[limpid grail]

well would you like to share your input and outputs with me

https://tenor.com/view/cat-wait-waiting-cat-wait-waiting-cat-waiting-gif-9780709586447195996

[ivory vigil]

No. Eric explicitly asked that people not make public collections of inputs.

[potent echo]

But you can just go to https://adventofcode.com/2025/events to see all of the things (apart from the part twos)

[hollow wharf]

what would u even do with inputs

lmao

and outputs

[limpid grail]

its for my leaderboard

[limpid grail]

i think that's for public stuff right

[half swan]

Do u have more than 200 ppl doing it?

[limpid grail]

well, if you dont trust me thats fine, but i could have shared my own to the public if i wanted to heh

[ivory vigil]

How exactly does input+output get used for a leaderboard?

[limpid grail]

evaluation

[hollow summit]

every year there is an advent of code because santa and his team are way too incompetent
last year he was mining "advent coins"?? are you serious???
and now this year there's a child with unrestricted access to the gift shop computers

[limpid grail]

uh just my hostel mates unfortunately

[half swan]

Use the private lbs then

[ivory vigil]

Eh? Are you trying to clone the website?

[limpid grail]

its based on runtime

[half swan]

Ah

[hollow summit]

i mean, if you like, you could generate your own inputs and outputs
but that's another problem

[ivory vigil]

Use your own input 😛

[limpid grail]

how would i check against the output 😛

but fine, ill just solve it myself then

they'll just have to wait

[coarse wagon]

I mean if nobody else in your group has solved it what's the rush, can't you just take the input and output pair from the first one?

[limpid grail]

they have solved it

:c

[coarse wagon]

Then ask them

[limpid grail]

im the slow one

[limpid grail]

i cant have them printing the output and cheezing the leaderboard :ferrisClueless:

well, the one who has solved it wouldnt do that tbh

cant say the same for the others

ill do that ig

but i was hoping to find a fren here who would help me maintain the added fairness

[ivory vigil]

👀 Why would it matter if you get the solution from one of us vs one of them? And/or what's the rush if it's based on runtimes. Just solve it.

[limpid grail]

theres no rush, id like to offer the lb faster for them is all

so they can compare

an anonymous person who shall not be named has helped me
thank you!! ❤️

[vast jetty]

2 hours wasted

[mortal bronze]

how you guys solve the part 2 of the day 1

[potent echo]

go to #aoc-solution-hints for hints, #1047673173447020564 for complete answers

[acoustic dock]

Whats the expected complexity for day 2 part 1?

[half swan]

time complexity?

or to solve?

[acoustic dock]

Yeah

Like the optimal time

[tender swallow]

mine was just O(N) where N = number of ranges x average span of each range
but apparently ||generating the repeats|| can be faster? salt-die did something like that

[acoustic dock]

Interesting.

Thanks

[naive wharf]

[part 2] can I reasonably ||store all invalid IDs that have been added to the sum?||

[tender swallow]

||yeah there's <1000 of them||

[naive wharf]

oh wait I can just make a set and sum at the end

[tender swallow]

||or do you mean if the sum itself will fit in an int/long/etc? Python ints shouldn't be a problem, but I'll say my sum was 11 digits||

[tender swallow]

||why though? you just need the sum itself||
||making it a set might cause problems if the ranges overlap anywhere (haven't checked if they have)||

[naive wharf]

||```py
ids = set()

for i in invalids2(max_id):
if check_id(i):
ids.add(i)
print(sum(ids))

worked for me||

[magic oxide]

||i personally feel like eric would've made the inputs not overlap

They've even checked most of the product ID ranges already; they only have a few product ID ranges (your puzzle input) that you'll need to check.

wouldn't make much sense in the lore if they overlapped||

[naive wharf]

same

[naive wharf]

I tried joining inputs, some do overlap

[magic oxide]

joining inputs?

[naive wharf]

i.e. folding adjacent/overlapping ranges into one

I deleted the code tho because I am given 33 ranges and I have cycles to spare

[magic oxide]

interesting

mine didn't have overlap, and i think ||a lot of solutions just sum over invalid ids in each range||, so must be rare

[naive wharf]

I consider adjacent ranges as overlapping

[sage bobcat]

do you guys make a repo with your solutions?

[naive wharf]

Wait people actually solved this by... iterating through all numbers and checking if they repeat?

[naive wharf]

what the hell

[magic oxide]

do you mean something like [a, b] and [b+1, c]? or wdym by adjacent

[naive wharf]

exactly

[half swan]

the ranges aren't that big

[magic oxide]

||adding to a set or just summing wouldn't change anything for those tho||

[naive wharf]

✨efficiency✨

[magic oxide]

||but storing it in a set would be less efficient?||

[half swan]

||hash retrieve sum is worse than sum as you go since you don't have repeats||

[tender swallow]

yes

[sage bobcat]

where to discuss solutions?

ah #1047673173447020564

[noble flower]

cracked part 1 but with an O(n^2) solution

[mortal bronze]

well that's better cus my solution on the day 1 part 2 involved consecutive adding/subtracting 1 ._.

[eager dove]

is there others there also feel the part 2 is really hard ( think I have used almost 1½ hour and cant wrap my head around a solution atm )?
Where part 1 was really easy for me
And no I'm not looking for a solution at lest yet which is why I won't go into the hints chat and spoilers chat

just curious if others also find it hard

[pulsar drum]

i accidentally did part 2 in part 1 :D i misread the instructions

[dense osprey]

cpp returning 100 when computing 1+pow(10, 2) made solving this problem so much slower

and it printing 101 when doing cout << 1+pow(10, 2) << endl; made it even worse

[dense osprey]

i did a more or less naive implementation because the ranges combined only cover around 1e6 integers

s = '...' # input
        arr = s.split(',')
        x = 0
        for i in arr:
            a, b = i.split('-')
            x += (int(b)-int(a))
        

        # x = 1949801  

[eager dove]

not sure if its hints for it but remember this channel are not for hints or spoilers for the AOC

[eager dove]

hihi nice 🙂 well done

[dense osprey]

how did you solve it?

because for me part 2 only required me to change a few lines from part 1

[potent zealot]

yeah p2 is entirely different from p1 for me and idk how to start implementing at all

i did p1 in the mathy way, i cant see how p2 can be done like this

[eager dove]

Yeah for me I did pat 1 in a non math way though still wont directly work for part 2

[eager dove]

Wee finally got the part 2 done even that I don't like my solution

[potent jacinth]

Yo folks. I don't understand this.

How can those be invalid IDs?

Where are 99 and 1010 coming from? 🤔

[sick musk]

yeah part 2 made me rewrite whole thing, but i glad that it can be heavy optimized by math knowledge

[warm berry]

invalid IDs by looking for any ID which is made only of some sequence of digits repeated twice

[tender swallow]

they're ranges.

[thick bluff]

it states in the question that it's a range

[potent jacinth]

Oh good Lord.

[warm berry]

Ah thats the part

[potent jacinth]

I understand now.

It's more tedious.

Thanks guys. I'll go grind the question again.

I'm still in doubt: does the sequence have to be repeated exactly twice? For example, would the number 1010610 be an invalid sequence? And should it "occupy" the full number string? What I mean is: should the repeated sequence be exactly half the amount of digits of the whole invalid number? If so, 10106 would be valid (not to be considered), but something like 101101 shall be added.
I believe it's this way, right?

[shy tulip]

For part 1 its exactly twice

Like you split it in the middle and left and right are equal

[potent jacinth]

Beautiful.

Thanks.

[vast jetty]

What new math facts did you learn

asking because for me I just abused itertools did the easy thing

[sick musk]

ill write in hints channel

[vast jetty]

as you should

how long does yall's code take

for both parts

i just ran mine and its not good

lemme time it

[tender swallow]

0.15s for p2

[vast jetty]

30 repetitions should do?

holy shit

nice

im assuming you didnt do lazy stuff

[tender swallow]

my first go at p2 was just ||checking each number in range||, which took ~3s
a while ago I tried ||generating all possible invalid IDs for a given number of digits and checking in the range|| which brought it down

[vast jetty]

INTERESTING

[sick musk]

~0.3-0.4ms for d2p2

[vast jetty]

my core logic is just:

which does take a bit time to run

well

not the function itself

but so much looping

i have ~5.2 seconds for both parts

5.3 more like

p1 is like <1 sec apparently

[tender swallow]

can't deny, that is elegant

[vast jetty]

p2 is the excessive looping

[bright spindle]

❯ ./run.sh 2 2
Running days.day2.star2()...
Time taken: 0.0004239 s

[vast jetty]

alrighty

lol

[vast jetty]

yeah not fast tho lol

ill throw the whole solution onto spoilers just for tradition's sake (throw them on the subreddit anyway)

[potent jacinth]

Uhm.

Is it safe to make the assumption the number of digits of the upper boundary is equal to the amount of digits of the lower one?

I don't think it's written anywhere, but I wouldn't know, for now, how to include the problem of different amount of digits in my code.

[tender swallow]

nope, they are very often different

[potent jacinth]

ARGH

[dense osprey]

||

for(ll i=a; i<=b; i++) {
            string ts = to_string(i);
            ll ls = ts.length(); // length of s
            for(int j=1; j<ls; j++) {
                if(ls%j!=0) continue;
                ll div = 0;
                for(int k=0; k<ls/j; k++) {
                    div += pten(j*k);
                }
                if(i%div==0) {
                    sol2 += i;
                    break;
                }
            }
        }

||

[potent jacinth]

Save me, someone. 🙁

[vast jetty]

unreadable fr

[dense osprey]

readable

idk what batched does in py

[vast jetty]

its from itertools

[dense osprey]

oh ill look it up

[vast jetty]

def batched(iterable, n, *, strict=False):
    # batched('ABCDEFG', 2) → AB CD EF G
    if n < 1:
        raise ValueError('n must be at least one')
    iterator = iter(iterable)
    while batch := tuple(islice(iterator, n)):
        if strict and len(batch) != n:
            raise ValueError('batched(): incomplete batch')
        yield batch

[hollow wharf]

use of ll

for long long

are you by chance competitive programmer

[dense osprey]

havent touched it in a long time

but i still use the same template and writing ll doesnt make it less readable than using long long

[hollow wharf]

fair

[mossy canyon]

hey, anyone know if there is a forum where we can create threads for aoc? I want to generate some images in the style of the story following the storyline of the puzzles and compile them in a thread

[hollow wharf]

following the story line of puzzles

sounds like a fun idea

[hardy relic]

Haha shit just looked at day2 and got that dunder Mifflin effect or whatever it is

Yesterday was ezpz

[minor cave]

Feel free to pop it in #1444747649084424202

[hardy relic]

are these designed to be done without any extra modules?

[minor cave]

No, use whatever you need

[drifting radish]

you decide, it can be solved with almost any programming language and with as little or as much libraries for your chosen language as you want as long as you can solve the puzzle

[hardy relic]

ok sweet as

thansk

[vast jetty]

panicked python user noises

anyway no you just need the answer

found this which is amazing

[solar jay]

yeah true, doesn't really matter what language you use to get it, as long as you get it

[drifting radish]

you only need to produce the correct answer in any way you like
by hand with pen and paper is probably a little bit extreme though and tedious with such large inputs though 🙂

[vast jetty]

people literally do it in assembly so

[vast jetty]

2023 day 5 vibes

[hardy relic]

i was guna say that, day2 seems like it would be a lot easier just writting it out on paper

[vast jetty]

or smth along that lines

[solar jay]

I did it on my school whiteboard when i had frees

[vast jetty]

thousands of entries

or tens of thousands

[solar jay]

for the idea for the code

then I spent like an hour coding because Im not the best at it

[hardy relic]

day 2 has like, 30 entries

maybe100

[vast jetty]

no it doesnt

[hardy relic]

oh your right haha

lucky i got a pencil sharperner

so stab myself with

[vast jetty]

this is my invalid list

[hardy relic]

i really dont know how i would do this, i think regex but ill need to do some research

[vast jetty]

got part 1?

[hardy relic]

i diud day 1 was pretty easy, but no i havent really started day2 yet. Just going over it in my head how i should do it but i dont know how to do the part where your looking for a pattern lol

you know, the main part that ya need to do

[potent jacinth]

I'm the stuckest person in this challenge.

I can't beat challenge 2 part 1.

Got my 170 lines of code (half of which are comments so I can track what I'm doing), but God knows why I get a wrong result.

At this point I might generate random numbers for ever until I get the answer right.

[warm berry]

Are you passing with the example input and example answer?

[minor cave]

If you're still stuck after the examples, feel free to make a thread in #aoc-solution-hints and people can help get you un-stuck

[potent jacinth]

No.

The whole building is a mess.

Well, very interesting.....

[potent jacinth]

Oh sorry LOL.

I thought I was in another server.

[maiden drum]

hello

[hardy relic]

my ternimal is running for like a minute just to try get the lists of numbers between the start id and l;ast id lol, i think im doing something wrong

[normal loom]

maybe, the largest gap I get between the start and end ids is in the order of hundreds of thousands, and it shouldn't take that long to iterate through all of them

[hardy relic]

yeh i think its the printing to console thats taking so lojg

now i have my gigantic lists of numbers, but i dunno how to looks for that pattern

yeh its the printing, the itarating takes 1 sec

LOL

[warm berry]

Could you instead check each value as you iterate instead of adding them to an ever-growing list? Would also save you from having to re-iterate over that list after it is made

[hardy relic]

yeh, but im still very stuck how to check for that pattern

[warm berry]

Which pattern? Part 1 or 2

[hardy relic]

im having fun tho haha

part 1

[warm berry]

So you need a sequence to be repeated twice

[hardy relic]

yeh totally stuck on that

[warm berry]

Is there any symmetry to that

Also, probably better to be discussed in #aoc-solution-hints

[hardy relic]

ohhh it has to use the entire number. that changes it. i was thinking th pattern could be like 568121285 ( that has repeating "12"s in it ) and i was like HTF can i ever figure that out

[vernal coyote]

helo

i just had s doubt can anyone solve it for me

like on day 2 puzzle how do i store that input like do i need to give it in a scan function or i can take it as a string with that input ?

[hardy relic]

save it as a txt file

[vernal coyote]

oo so i have it as a txt file and then scan character by character to get the number values in a way ?

[warm berry]

The provided values are comma separated start and stop points for a range of numbers

11-22 means a range of numbers starting at 11 and ending at 22

e. g. 11-22,95-115,998-1012 are three different ranges

Generally, saving the full input as a text file and then reading it in is a good way to handle the full inputs. The example inputs are usually short enough to keep as a multiline or single line string

[void forge]

Done

[vernal coyote]

i see
thank you for the explanation

[hardy relic]

Seperate them onto seperate lines then for can do line by line

[vernal coyote]

will give it a try now
since i don't know python completely
imma use c

[void forge]

What determines the total score?

[warm berry]

From the faq: #aoc-faq message

[void forge]

ah

thanks

[echo wave]

day1 part1 for me so far

[warm berry]

Could you post in the #1047673173447020564 thread instead?

thank you

[white arrow]

day 1 made me give up

[bright spindle]

There's no time limit. You can come back to it later and try again.

[white arrow]

just frustrating lol i couldnt figure out my type problems

[bright spindle]

I often just shelve something when it gets frustrating and do something else, then come back to it later when I feel more motivated.

I feel like I often get insights about the problem when I'm not actively working on it.

like in the shower or something

[ripe meteor]

➜ Measure-Command { py .\solve.py }

Days              : 0
Hours             : 0
Minutes           : 0
Seconds           : 21
Milliseconds      : 524
Ticks             : 215240285
TotalDays         : 0.000249120700231481
TotalHours        : 0.00597889680555556
TotalMinutes      : 0.358733808333333
TotalSeconds      : 21.5240285
TotalMilliseconds : 21524.0285

I do not care AoC a star is a star

[dusty trout]

seconds: 21

[ripe meteor]

a win is a win

i've got time

[pseudo fulcrum]

I,m feeling badly, I can't share in advent of code due to I,m beginner 😩

enjoy my friends

[shy tulip]

You could have a go at day 1 part 1 even as a beginner

[hardy relic]

YEH im a begineer and day1pt was fun!

[white arrow]

can anybody help with day 1 pt 2?

[remote mauve]

#aoc-solution-hints

[lusty rover]

Chat whats more important?
Sleep, or AoC?

[low condor]

Why not both? I dream about lanternfish after all

[ripe meteor]

dont tell me you stay up for aoc shen

[hollow wharf]

AOC def

[ripe meteor]

you have all day

[hollow wharf]

I do stay up

Till 1 am

For it

Atleast that’s what it took at max to solve all the problems

I can stay longer

🫡

[lusty rover]

I do
Opens at 11, generally in bed by 12:30, hopefully asleep by 1
Then get yelled at to get my ass up at 5:00 for work

[hollow wharf]

Opens at 11

Huh

[ripe meteor]

What timezone is that

[hollow wharf]

You are 1 hr ahead of est

[low condor]

I do actually prioritise sleep - puzzles open 5am for me and no way am I waking up for that

[lusty rover]

Texas
It’s midnight UTC-5
I’m UTC-6

[lusty rover]

Behind*

[hollow wharf]

Sorry behind

[hollow wharf]

Or you can sleep early

[ripe meteor]

In the spirit of xmas they should move the opening time to like 10am UTC

[hollow wharf]

To wake up at that

Time

[hollow wharf]

💀

[ripe meteor]

Think about it

[lusty rover]

Well that just screws the other half of the people

[ripe meteor]

When you open your gifts you dont do it at 00:01 xmas day

You do it when you naturally wake up

And i naturally wake up at 10

[low condor]

I kinda wish the opening hours would change every day before

[lusty rover]

Why so late??

[hollow wharf]

You see that’s normal human behaviour

But we are programmers

We don’t do normal

[low condor]

But since there’s no global leaderboard I care less

[ripe meteor]

Sleep is good shen, i like sleep

[lusty rover]

Yeah but sleep when there’s no sun

Don’t sleep the sun away

[hollow wharf]

I also sleep during the day

So that I can keep up

Late

[ripe meteor]

Theres no sun when i sleep

The gf has black out curtains up

[hollow wharf]

there is no sun

Any time of day

For me

[lusty rover]

Get a new GF

[hollow wharf]

😭 😭 😭 😭

[ripe meteor]

But all the time i invested in this one

[hollow wharf]

Okay you have a hard choice now

Gf or AOC

[ripe meteor]

gf lmao you nerds

[hollow wharf]

😭

☠️

[magic oxide]

yeah no leaderboard has been nice
i'm surprised i'm actually not stressing myself to do it on time

[low condor]

Stressing oneself out over Advent of Code seems like the opposite effect of what the author wanted

And yet I have done that myself on occasion

[ripe meteor]

Idk about that

Midnight release + leaderboard= what else could have happened

[hollow wharf]

Lmfao

True

[ripe meteor]

You gotta be new to humanity to not see whats going to happen

[open elbow]

AoC gang I'm a completionist ✌️🥺

[hollow wharf]

💀

[hollow wharf]

I was thinking pinging you right now ngl

And you replied

great minds think alike

🗿

[hollow wharf]

Did you do the day 2

[open elbow]

almost

im at part 2 right now

i'm very close

[hollow wharf]

I saw you did part 1

Super quick

[open elbow]

yeah I finally got my part 2 parsing working, it's not working completely for a certain number 😔

[low condor]

Debug time 🐞

[open elbow]

nvm it's missing like 4 values

[low condor]

Who is Uncle Albert’s older brother anyway 🤔

[open elbow]

i have no idea 😔

[low condor]

I wish your display name was a reference to the British show “Only Fools and Horses” - that show has an Uncle Albert

But it’s just referring to Einstein isn’t it

[open elbow]

yeah

can someone help me in #aoc-solution-hints? I genuinely have no idea what's wrong

[hollow wharf]

Did you got it

Part 2

[open elbow]

YEAHHH

[hollow wharf]

Yoooooo

[open elbow]

I solved this one entirely by myself with some hints

[hollow wharf]

Hell yeah mate

🗿

[open elbow]

I used a big brain technique to just turn the lists into sets and union both of them 😎 ( #1445278429372219422 )

[open elbow]

1 min

[ivory vigil]

Tick... tock...

[shut trellis]

<@&518565788744024082> Good morning! Day 3 is ready to be attempted. View it online now at https://adventofcode.com/2025/day/3. Good luck!

[open elbow]

this one's weird lemme reread it

[waxen hound]

Alright

[open elbow]

oh my god what the HELL is the input 😔

[glass ridge]

☕

[ivory vigil]

That was fun

[earnest forge]

That was cool

[ivory vigil]

Now I need to figure out what I did 😆

[woven oriole]

ggs

[vast jetty]

lol

[tepid galleon]

I give up

[vast jetty]

tfw the brute force approach wont work for part 2

[open elbow]

this one was actually pretty easy for part 1

[woven oriole]

i really wanted to know the global rank for this one

[vast jetty]

wdym i gotta write actually good code

[vast jetty]

whats your time

[woven oriole]

4, 8

[vast jetty]

this is def not gonna work LMAO

good on part 1 tho!

[tepid galleon]

Won't work for pt2

[vast jetty]

part 2 you gotta write actual logic

yeah it wont

lol

[tepid galleon]

I gave it a solid 6 minutes

[vast jetty]

LMAO

did we all do this

[woven oriole]

i am very glad it doesnt work on p2

some value given to day 3

[vast jetty]

i dont feel like writing actual logic rn

yeah lmao

i was prepared to even gave it half an hour if needed

but its not even doing a single line

theres prolly a way to make more optimised brute forcing tho

[vast jetty]

damnit faster than me

FCK

gonna be stuck in vim now lmfao

[reef harness]

Just notice this. Can I start from day 1?

[woven oriole]

sure

whos stopping

[vast jetty]

sure

honestly day 3 is easier tho 💀

at least part 1

[tepid cairn]

it's advent of code already??

i didn't even notice 😭

[vast jetty]

do we think ||recursion|| sounds stupid

i think it could work

or at least a loop calling a function over and over again

[ivory vigil]

Tsk!

[humble copper]

length of input makes it extremely difficult to do this easily

[vast jetty]

no?

it seems efficient enough

im writing a prototype for example lets see if it works there first

[humble copper]

oh wait for which part

[vast jetty]

ah

well fuck

[humble copper]

yeah

[vast jetty]

omw to debug

[vast jetty]

part 2

part 1 was done in like 5 minutes

[peak dock]

100C12 is ~10^15 btw

while 100C2 is like 5k

[vast jetty]

pretty quick debug lmfao

yeah i think this is gonna work

ill tidy up code later lemme just find the answer first

[humble copper]

light work

let me just bust out my botnet

[vast jetty]

it works

[obsidian shell]

part 1 is so easy bro

but

part 2..

fucking hell

had to literally spend an hour LMAO

[vast jetty]

it was reallyyyy fun tho

[obsidian shell]

yeh

true

[vast jetty]

5 some minutes lol

[vast jetty]

i was eating while doing it

mightve gotten it done before otherwise

lol

[obsidian shell]

lmao

[vast jetty]

what did you do

[obsidian shell]

i was eating tangerines lmao

[open elbow]

damn it part 1's already stumping me because I have no idea what method can actually do this

[rancid creek]

explain your problem in #aoc-solution-hints maybe we can give you an idea.

[half swan]

me when aoc is leetcode: ||https://leetcode.com/problems/remove-k-digits/description/||

[open elbow]

just did

don't know why I'm struggling on something so easy 😔

I guess this is a new type of challenge (that I haven't done before)

[sullen adder]

...i spent half of an hour assuming the actual input was the same length as the example before i read my input file 🤡

[rancid creek]

lessons learned from the past: the input can be important.

[sullen adder]

i'll take it as a sign to stop attempting AoC at midnight

[magic oxide]

dang i forgot about aoc

[vast jetty]

hm

fucked up timing i think

ok much more believable

yeah ok the other one is more shit

good to know

[mortal bronze]

Bruh I still can't solve the day 2 part 2 maybe I'll just brute force it

[winged trellis]

i brute forced it, took me about six seconds of runtime 😄

[bright spindle]

I kinda got into the habit of always taking a look at the real input and checking if there's some kind of significant difference from the samples

after making my code work with all the samples

[vast jetty]

i check actual input almost immediately after reading instructions

maybe after getting parsing to work

[vast jetty]

how did you do that

[bright spindle]

Yeah, sometimes I do as well.

In order to grasp the full extent and intention of the puzzle.

And identify any potential traps.

[winged trellis]

ask in a spoiler channel

[winged trellis]

you can ask in the spoiler channel

[teal field]

the first thing i do is run my little script to download the inputs, the size in bytes also gives me some idea of what to expect compared to the example input when i go read the problem 😄

[pulsar drum]

My part 2 for today
Execution time: 0.003069162368774414 seconds

[winged trellis]

all relavant parts or just the core business logic?

[pulsar drum]

its the timing for my main()

also i use just pythons time, so it might not be super accurate

[winged trellis]

fair, i just time the time for everthing, starting the interpreter included

perf_counter is fairly ok

[pulsar drum]

what command to use to time the literal python command? for unix

[winged trellis]

time python main.py

uv run day3.py 0.02s user 0.01s system 97% cpu 0.028 total

[pulsar drum]

python3 2.py input.txt 0.01s user 0.01s system 64% cpu 0.040 total

TIL: about the time command

[winged trellis]

its pretty nice to know about

[coarse wagon]

There's also more dedicated benchmarking tools that do multiple runs similar to timeit

Like https://github.com/sharkdp/hyperfine

[winged trellis]

i like timeit!

is hyperfine a general tool?

[pulsar drum]

i downloaded hyperfine jusdt now, looks cool

[coarse wagon]

Yeah similar to time it benchmarks CLI commands, but doing multiple runs with support for stuff like warmups, and setup/cleanup steps. (For example if you are benchmarking builds you could add the cache cleanup as a setup command)

[winged trellis]

❯ hyperfine --warmup 3 'uv run day3.py'
Benchmark 1: uv run day3.py
  Time (mean ± σ):      26.0 ms ±   1.0 ms    [User: 18.3 ms, System: 7.5 ms]
  Range (min … max):    24.2 ms …  30.5 ms    108 runs

running it with uv has an huge improvement over just running the python interpreter normally

[tender swallow]

how come?

[coarse wagon]

Are they using the same python version?

[winged trellis]

yes

[warm fiber]

but really... I have to ask... was the description of the part 1 problem enough to understand the example 818181911112111 -> 92 ?
I had more problems cracking the hidden rules than the actual problem.. (problem is actually trivial)