advent-of-code

[earnest forge]

Good luck everyone

[woven oriole]

glhf

[shut trellis]

<@&518565788744024082> Good morning! Day 9 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/9. Good luck!

[long egret]

i dont even understand the problem

🗿

[compact wasp]

bruh @devout tusk you beat me in pt1

i've officially fallen off

15/16

[buoyant hatch]

is aoc python only

[devout tusk]

p2 ain''t workingbfor mw

[compact wasp]

rip

if i didn't go to the bathroom yesterday i'd be first 🗿

no regrets though

[obsidian anvil]

Yeah

[compact wasp]

😂

[pine tiger]

HAHAHAHA

[compact wasp]

🔥

bruh this mf seriously tied with me

[woven oriole]

okay there is somethign stupid going on

testcase fails wow

[wanton plover]

This one tricked me smh

[devout tusk]

13/338.
FML

[woven oriole]

what the sigma

why does it work on given case but not tc theres no room for edge cases?? smh

[pine tiger]

oh ru "Michael"

[compact wasp]

the 13 is for pt1

[pine tiger]

oops

[devout tusk]

no 😭

[devout tusk]

global 58

[pine tiger]

icic

not on pydis?

[devout tusk]

what's that?

[pine tiger]

the lb here

[compact wasp]

how many ppl here are on global

[pine tiger]

adavya, you, tesseract, oliver

[devout tusk]

tessAract. (I've intentionally misspelled it)

[pine tiger]

naruhodo

[devout tusk]

just to annoy people

[compact wasp]

if you really wanna annoy me then you should get to #1 on global

[devout tusk]

ryuichi?

[compact wasp]

and we'll need adavya or oliver to get #3

[devout tusk]

that ain't happening any time soon.

[compact wasp]

dw you have like 2 weeks to get there

[devout tusk]

and my ranking is going to drop even further since I may not even be able to work on AoC for 3-4 days straight since I'm going into a no internet area :/

[compact wasp]

need full pydis podium

[devout tusk]

how do you add yourself there? is there a bot/commands?

[compact wasp]

no i meant need top 3 on global to be pydis

[compact wasp]

i don't remember at this point

[woven scaffold]

||are we supposed to handle file ids >= 10||

[pine tiger]

||yes, they move around just like single digit #s||

[woven scaffold]

||but you also have to take into account the length of the id right||

||so like id 10 will take up 2 indexes of space||

[pine tiger]

||no, actually!||

[woven oriole]

nvm nvm

ignore

the problem is somewhere else

[woven scaffold]

||did I read the problem statement wrong or something lol||

[woven oriole]

||i am getting wrong on my tc||

|| i just use two pointers for swap||

||smh||

[woven scaffold]

||so if I take the example input and add another file at the end, and make the first free block larger: 273313312141413140203, I get the string 00.......111...2...333.44.5555.6666.777.888899101010. So after the first swap should it be 00101010.111...2...333.44.5555.6666.777.888899......?||

[compact wasp]

||where are you getting 101010||

||should be AAA||

[hollow wharf]

why doesn't the problem specify how to handle >two-digit ids?

[woven oriole]

what the sigma

[woven scaffold]

||ah||

[woven oriole]

what is ||AAA||

[woven scaffold]

||ok time to rewrite lmao||

[compact wasp]

it's merely representation for visual clarity, it's not like you're actually building the string

[peak dock]

gdi

i crashed vsc for 5 minutes because i forgot to remove a print in the loop before running it on the actual input

[compact wasp]

😭

ouch

[woven oriole]

what is this

ocntext

😭

[compact wasp]

literally look at the preceding message

[peak dock]

#aoc-solution-hints shoo

[woven scaffold]

whoops

[woven oriole]

||how did 10 - > A||

[compact wasp]

i unfortunately don't feel like figuring out how to explain

[woven scaffold]

||A is a way to represent the next number after 9 using a single character||

[compact wasp]

^

[thin topaz]

I had a stroke while doing part2 bro

[woven oriole]

WHERE IS THIS MENTIONED

WHER THE FUCK

[compact wasp]

.

[woven oriole]

okay fml

oh ok

got it

got it

[normal umbra]

The part 2 example has this at the end ||00992111777.44.333....5555.6666.....8888..|| this isnt the end result, is it? coz ||there are free spaces left like 4 spaces before 5555 so 8888 will be placed there right?||

[thin topaz]

||You tried to move 8888 right after 99 there is no spaces at the left for 8888 at that moment||

[normal umbra]

oh

so what I send will be considered the end result?

[thin topaz]

after you tried to move all the blocks

you would only try once

[normal umbra]

I see

okay thanks!

saved me a ton of time

[thin topaz]

could have got better ranks for part2 if I am not struggling so hard and refuse to just move the numbers around

[wanton plover]

oh thank god

[slate berry]

Can anyone help me with day 5 part 2? I wanted to optimise my sol...

[cyan sleet]

cute mascot

[magic oxide]

i just realized i've been trying to solve a harder problem

bruh

[cyan sleet]

I'm confused how this is the io module
https://hexdocs.pm/gleam_stdlib/gleam/io.html

I only see o

@magic oxide does gleam really not have a get_line?

[feral hazel]

i found this https://hexdocs.pm/gleam_erlang/gleam/erlang.html#get_line

[magic oxide]

i'm still doing today in python 😔

[cyan sleet]

I saw that, but it makes me sad to have to reach into erlang 🥴

[magic oxide]

read the problem wrong and solved a harder one

[cyan sleet]

I guess it makes some sense since it can be compiled to js or whatever

[feral hazel]

yeah looks like it https://github.com/gleam-lang/gleam/discussions/2689#discussioncomment-9025424

[slate berry]

anyone??

Im thinking || toposort|| but not sure

[floral hare]

||yeah it is toposort||

[hollow wharf]

first day left unsolved

[compact wasp]

what did bro do

[magic oxide]

p2 if ||moving a block could clear up space for other blocks to move into||

[compact wasp]

how would that even work

[magic oxide]

very slowly

i just ||kept iterating over each file and empty slot until there was nothing to be changed||

[normal umbra]

that is so much work

yikes

[covert arrow]

I skimmed today's part 1 and implemented part 2 instead

[slate berry]

Can you help me to understand it

like how it's working here

[floral hare]

well, for starters you can use ||kahn's algorithm to topologically sort the indexes you want||

[naive wharf]

god damn it

for this AoC, I'd do

for (from, len, id) in <censored>:
  ...

one of these is a keyword.
two shadow builtins.

WHY, Python?

[peak dock]

i wouldnt feel too bad about shadowing id tbh

[naive wharf]

still

[slate berry]

Then?

[limpid grail]

How long does your Day 9 solution take to run

[rapid wolf]

it takes ||6s||

[limpid grail]

mine is taking quite some time

oh its done

||part_1 took 104.20s to run||

[rapid wolf]

but I also operate ||on sequences of n blocks at once, not on each individual block||

[limpid grail]

and its wrong

[rapid wolf]

classic

[hollow wharf]

both parts take roughly ||24s||

[modern terrace]

about ||3s|| both parts

why spoil anyways-

[hollow wharf]

idk tbh

[modern terrace]

i found it easier to ||just work on the numbers and use arithmetic sequences|| than deal with the performance problems later

[cyan sleet]

🥴

[tender swallow]

first attempt took 2min, got it down to 2s now

[cyan sleet]

I think in python part 1 should be fairly trivial to do efficiently

[tender swallow]

part 1 was fairly fast even at first yeah
only slow thing was 2

[cyan sleet]

||walk with two indices from either end and swap things||

[woven oriole]

||did someone use bisect_left for p2 for optimization's sake||

||TIL insort_left exists , i couldnt make it work with bisect_left, although the ans is pretty close but still not correct , tuple comparator smh||

[woven oriole]

yes simple ||two pointers||

[oak flax]

fml, i finally did part2, worst code i have ever written probably, very ugly, but im not touching it again

[cyan sleet]

sadly I had to write in a functional language 🥴

[woven oriole]

which one

haskell or wot

[hybrid gull]

0.02s for part 1
1.64s for part 2

[unique osprey]

That's nothin. My p2 takes 40 sec

[hybrid gull]

Gotta get your money's worth outta that CPU

[unique osprey]

Thank you to whoever showed import tqdm the other day. Otherwise i would have given up halfway through running and worked on optimization

[shut otter]

why is my re.findall returning an empty list? for day 3

[distant igloo]

Might be a better question for #aoc-solution-hints

[shut otter]

oh ok also i go the answer i was missing an escape sequence \

[distant igloo]

Remember that this channel should not have any spoilers or hints! People might not appreciate the information you're putting here

[slender ravine]

Tbf the question is fine but answer should be spoiler

||nvm, that's was the one that you should use regex without a doubt||

[tender swallow]

||when I was solving it regex didn't even occur to me until I had already finished without it||

[slender ravine]

Oops

[naive wharf]

challenge: day 3 part 3 - allow muls to be nested

[thorn inlet]

regexers in shambles
recursive descent parser believers in profits

[dense gale]

i just signed into the website now lol that was alr too much effort i think i leave the programming tommorow

[dense gale]

well that makes me think im totally fucked now

[long egret]

i understand it now

[dense gale]

i havent even seen what the problem is yet

[unique osprey]

Start with day 1 if you haven't yet

[dense gale]

im tired for this tbh i need my last year motivation

[mild hill]

Guys, someone help because I am losing my mind

Isn't this what we are supposed to do in part two of day 9 ?

[hybrid gull]

Probably better for #aoc-solution-hints, but ||I don't quite get the math of what you are doing for the checksum. You need to multiple the block's current position by the ID of the file occupying that block. So 0's and 9's would be 0*0 + 0*1 + 9*2 + 9*3||

[mild hill]

bruh

Thanks, makes sense, I didn't read it correctly

[somber trench]

For day 9 will I have to account for numbers more than two digits and if so do I have to keep them the same double digit

[hollow wharf]

treat them as one digit

[somber trench]

So 17 turns into 71? Or stays as 17?

[hollow wharf]

stays as 17, also avoid building the string, it will cause unnecessary bugs like this one

[somber trench]

Yeah, i was gonna treat it as a list after getting that info

[shut trellis]

<@&518565788744024082> Good morning! Day 10 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/10. Good luck!

[gilded osprey]

bro shut up

unsubscribe me

[gloomy osprey]

#roles

[compact wasp]

31/29

gg

2 llmers ahead of me

[gilded osprey]

thanks

[devout tusk]

the question refused to open for me today! :/

[pine tiger]

there's a certain peace in knowing there's nothing you can do about it

[devout tusk]

24/45

[compact wasp]

bro still beat me p1

[compact wasp]

yeah, meh

at least everyone knows they're llming lol

[magic oxide]

at least you got a point

[pine tiger]

ffs hyper neutrino isn't on lb

[devout tusk]

OMG, I'm up to global rank #34 🥹

[compact wasp]

bikatr7 called out in the reddit https://www.reddit.com/r/adventofcode/comments/1ha7p80/the_bullshit_explanation_of_the_current/

[compact wasp]

sheesh

[devout tusk]

#34*

my bad

[compact wasp]

also i love how #1 and #2 both bought aoc++ today

[compact wasp]

they certainly didn't have it yesterday

[devout tusk]

wdym bought aoc?

like sponsored it?

[compact wasp]

the yellow thing

[compact wasp]

https://adventofcode.com/2024/support

[devout tusk]

ah!

[devout tusk]

all of top 3 bought it 👀

[magic oxide]

i need to stop misreading the question

at least today it wasn't too bad

[compact wasp]

yeah both of top 2 bought it today

let's just say my interpretation is cynical

[devout tusk]

the cynic in me believes that eric throttles speeds of opening questions for those who don't pay 💀

[compact wasp]

💀

[magic oxide]

i can't believe AoC is p2w smh

[humble copper]

i somehow solved part 2 before part 1

completely misread what it was asking for

💀

[magic oxide]

advent of reading comprehension strikes again

[woven oriole]

WHY DO I WAKE UP LATE ON EASY DAYS

also my ||soul died on omitting the visited set for p2 then i was like yeah its fine||

[covert arrow]

smh I accidentally did part 2 for part 1 again

[woven oriole]

i woke up 15 mins late

🤡

[magic oxide]

still beat my time

[hollow wharf]

same

[wind fractal]

accidentally solve part 2 first LOL

[woven oriole]

what was the rank in solving p1 <= 6mins

i wanna feel regret

[compact wasp]

not top 100

lb filled in 2 mins

[magic oxide]

leaderboard filled up insanely fast

yeah

[magic oxide]

wait u play amq?

[compact wasp]

a little bit

i'm bad though

[magic oxide]

cool

small world

[hybrid gull]

Feels like this year there's a way higher number of people convinced there's an error in the puzzle text. I know it happens every year, but this year feels inordinately higher.

[rancid creek]

9.5 minute solve -> rank 500.
maybe it's just me, but this years times seem faster than previous years, and I can't help but think there's a lot of people ai coding it, and not just in the top 100

[magic oxide]

tbf today was pretty cut and dry

but yeah i agree

[viscid elbow]

you made a difficulty chart last year, I follow your way to check this year's difficulty but wonder it may lost its accuracy this year

[lime fulcrum]

I mean i would agree. None of them where that hard yet

[thin topaz]

part1 is harder than part2 smh

[lime fulcrum]

Today?

[thin topaz]

||I tried too long to not just bfs from each 9||

yes

[long egret]

i accidentally did part 2 first before i even knew what it was

i fixed my code for part 1, then realized and un-fixed my code for part 2 🗿

[prisma fulcrum]

!paste

[marsh current]

Pasting large amounts of code

If your code is too long to fit in a codeblock in Discord, you can paste your code here:
https://paste.pythondiscord.com/

After pasting your code, save it by clicking the Paste! button in the bottom left, or by pressing CTRL + S. After doing that, you will be navigated to the new paste's page. Copy the URL and post it here so others can see it.

[compact wasp]

extreme demons are harder than insane demons smh 🙄

[rancid creek]

I'm sure it has. I wonder if I can find the link for the guy that made a live version of it.

[rancid creek]

the 2024 numbers seem reasonable so far. I think we might have had another medium in there a few days ago. But overall 2024 hasn't been too bad so far.

[magic oxide]

i would've had yesterday as insane

i was kinda out of it yesterday

[rancid creek]

I've been in the middle of moving, I've only started 1 day on time so far. And that day they had an issue with the website. lol

[compact wasp]

why didn't eric invalidate scores 🗿

[rancid creek]

Does mean I've been much more relaxed about solving though. Which is nice.

[rancid creek]

too few people impacted by too small an amount.

[compact wasp]

was 2020 much worse

wasn't there so i wouldn't know

oh

[rancid creek]

It's hard to tell, times have been getting easier as more people started doing aoc as well. So it's not just that aoc looks easier over time, it's also that aoc coder quantity has improved over time.

[ivory marsh]

The future is grim, the graph day finally arrived ```
d p time abs rel
...
10 1 376ms 38% 41%
10 2 366ms 37% 40%
sum 922ms 92% 100%

[compact wasp]

||you dped right||

[ivory marsh]

No #1315907361596112967 message

I did far worse

[compact wasp]

i see

[ivory marsh]

At least better than my initial solution which used || A* with edge goal and estimate cost both set to permanent 0 since I don't understand them||.

[rancid creek]

part 2 today, ||which was for me literally change 1 character,|| still took me 60 seconds to submit. Which shows how slow I am. =/

[compact wasp]

it's monotonic so ||we know there won't be any cycles||, thus it's just ||memoized recursion||

[ivory marsh]

Yes but graphs make my head hurt, and ||rust doesn't have a nice @cache built-in||. I'll probably optimize it tomorrow once I'm less tired since that is untenably long, and I want to keep going no deps.

[compact wasp]

🫡

[cyan sleet]

you could easily make a struct to deal with the caching

struct with a hashmap from argument tuple to result, impl the function and just do the caching manually

in python it would be something like

if args not in cache:
  ...compute value...
  cache[args] = value
return cache[args]

[warm fiber]

@mossy basin just wanna say that I love your visuals and I'm going to try myself something, thanks for sharing your library ❤️

[elder quest]

functools.lru_cache is basically cheating

I love python

[mossy basin]

❤️ show me your visuals

[warm fiber]

as soon as II understand how to start with your library, I will surely! (this day should be easy... should.. HAHAHA noob me!)

[mossy basin]

currently playing around with today:

[cyan sleet]

@minor cave congrats on picking fortran on the same day of the calendar and on another grid problem

so I could re-use my input reading code from last year

[blazing vector]

wtf is that 4.998873919248581e-07

[cyan sleet]

smol number

[worthy arch]

Hi everyone!

I made Advent of Code 2024 animations with Manim.
Kinda did this for fun at first(i was dared to and well, it's still fun), but I'm just going to see if I can animate all 25 days:

https://www.youtube.com/playlist?list=PLpBOB0SsPa0QpUTDXfHvPrsP1QJe5e8xB

I need your feedback to improve this : )
Eg: In the earlier days, the animation was quite slow but i've improved the speed XD

[warm fiber]

smooooth, very nice!

[warm fiber]

to not bother everyone on my problems with your lib, where can I ask for help? XD

[worthy arch]

Thanks : )

[tender swallow]

I think day 6 would've looked a little more polished with actual arrows instead of ASCII ^<>v
but great otherwise! you should post it in https://discord.com/channels/267624335836053506/1047673173447020564/threads/1312647387495534632 !

[worthy arch]

Yh, now that I think of it...
I'm a noob : )
Thanks!

[wary river]

hol up i missed something

how did you || END UP USING A*, FOR WHAT JOY ||

which problem statement

[modest sand]

i rn doing aoc day 1 part 1 but the number that i got is minus

is that also valid ?

[low condor]

Nope. Are you doing it in Python?

[modest sand]

yes

i going make it abs()

[low condor]

Ok, just checking you didn’t overflow an int in another lang

[low condor]

Have you tested against the example input?

[modest sand]

yes

can i drop my code here ?

[low condor]

#aoc-solution-hints
Do it there with spoiler tags

[modest sand]

after seen people code i found that my code also right

but the output is wrong i don't understand

[low condor]

Then your code isn’t right :)

[modest sand]

i did this line for get total distance

[tender swallow]

@modest sand don't post code here. #aoc-solution-hints

[modest sand]

alr

[mild hill]

easy day today

still stuck on yesterdays part 2 though, works for example not for input

[warm berry]

Yeah Day 9 Part 2 might just be a skip for me

[mild hill]

SKIP ⁉️

My ego won't allow me to skip a day that comes before the day I reached last year

[cyan sleet]

it's not that bad

if you handle things like ||(type, size) tuples|| it should mainly just be ||a nested loop||

[warm berry]

My brain just doesn't grok it very well

[naive wharf]

wait that was ||disk defrag 2||?

I just used ||2 separate lists, one for files (start, size, id) and one for space (start, size)||

[random lintel]

Man batgrl is fucking amazing

[warm berry]

That is sweet! You should also check out #1312647387495534632

[random lintel]

Oh, I hadn't noticed the channel, nice 😄

Also, thanks

[mild hill]

Did you guys see that ?
https://www.reddit.com/r/adventofcode/comments/1hawiif/yup_that_will_be_anniversary_theme_everywhere/

I really hope it's just a 10 year anniversary and not the last AoC

[paper folio]

wow hope not, this is my first year trying it out 🥹

[mild hill]

This is the first year I can actually commit to it

[cyan sleet]

I ended up with ||two lists, files and space and only files||

the latter for ||picking what I was trying to insert||

[limpid grail]

This is so good 😭

[limpid grail]

daym

i think you gave the guard some extra features 😹

very cool stuff!

[mossy basin]

insane use of the shadow caster

[random lintel]

Thanks Insane library you built there

[cobalt torrent]

yo that’s really cool

[ivory marsh]

I've figured out how to implement ||flood fill||. I still don't understand how ||caching paths|| could ever work, but this is fast enough for me. ||https://paste.rs/GmKm9.rs|| New times: ```
d p time abs rel
10 1 7ms .65% 3.2%
10 2 6ms .56% 2.7%
sum 206ms 21% 100%

[cyan sleet]

you were the one complaining about no cache decorator

[karmic coral]

Anyone working on day 3

[low condor]

I can help with day 3 if you need it

[torpid karma]

I solved part 2 day before part 1 because of a bug in my code that made it act like part 2 XD

Fixed the bug only to realize that I need to add it back

[hybrid gull]

Ctrl-Z was the theme for today

[clear nest]

Why was todays aoc actually suspiciously easy?

will tommorrows be hard or smth?

[hybrid gull]

Tomorrow will probably be Chinese remainder theorem or something like that.

(I kid, it's a bit early in the month for that.)

[torn sapphire]

you never know, maybe we'll be thrown the mid-week difficult curveball just to keep us on our toes

[hybrid gull]

Heaven forbid, maybe it'll be Beverage Bandits again.

[torn sapphire]

days 4, 6, 8, and 10 have involved maps or grids of some kind, so following the pattern, tomorrow will not involve any sort of grid, that's the only prediction i'm confident enough to make 😄

[hybrid gull]

I would be curious how the LLM folks do against a Beverage Bandits type of problem, though.

[clear nest]

https://tenor.com/view/non-canon-dc-worldwide-gif-25576546

[clear nest]

oh dang what problem was that

[hybrid gull]

One of the longest leaderboard times of all problems. Top 3 I think?

2h 23m to fill the leaderboard

It was a day 15 too

[cyan sleet]

I missed the leaderboards that day 🥹

[unique osprey]

What year?

[cyan sleet]

https://adventofcode.com/2018/day/15

[clear nest]

damn

crazy

[minor cave]

If we have a 60+ minute leaderboard fill this year, I'm just gonna open the thread and trust y'all to only post when the global has filled. Ya girl needs sleep

[compact wasp]

there's only like 6 people here who'd get onto that leaderboard 😭

[slender ravine]

Hope I am not too sleepy that I can complete it and then go back to sleep for a while

[magic oxide]

hopefully it's not too hard so i can focus on catching up with roulette

oh today is C##

[earnest forge]

Good luck everyone

[sour gyro]

also hoping it is not too hard, im a few days behind

[shut trellis]

<@&518565788744024082> Good morning! Day 11 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/11. Good luck!

[somber trench]

made a post in #1315547415381807114 for help

[devout tusk]

I was using a ||deque|| which was hellishly slow 😦

[minor cave]

#1316269903363440690

[devout tusk]

50/87

@compact wasp about to steamroll my ranking 💀

[compact wasp]

lol nah i'm too burnt out to fix my brute force

[devout tusk]

lol

[devout tusk]

i don't think it'll matter for you anymore. You'll still be top #15 at the very least even if you stop solving from today lol

[compact wasp]

nah i think i'll be like #40

judging by last year

[slender ravine]

||Oh f, part 2 bf gonna take forever||

[hollow wharf]

I think the puzzle creator is getting a bit lazy

[compact wasp]

lol yeah

[junior iron]

aaaaaaaaaaaaaaaaaaa

[pine tiger]

unlucky

but hey

[junior iron]

sorry my keyboard spazzed out

[pine tiger]

we power on

[hollow wharf]

it literally might

[slender ravine]

Just 39

[late frigate]

me submitting the demo output :|

[slender ravine]

Took 7 minutes from 39->41

[mystic gull]

everytime theres "after X amount of times" in the question, p2 most surely will be just "iterate more times"

i swear

[woven oriole]

ggs

[devout tusk]

p1 to p2 took quite a bit of time for me too!

[pine tiger]

ok king

[ivory marsh]

Today is one of the days I was dreading. Part 1, simple, trivial even. Now I have to somehow figure out the pattern for part two, or wait several years for my existing code to finish

[somber trench]

can someone help with this? no one in the channel is actually helping when i asked
#1316269903363440690 message
its the day 11

part 1

[magic oxide]

you would probably run out of memory before it finished

[random lintel]

We'll see about that, time to get on the campus cluster

[rancid creek]

mine ran out of memory around round ||40||. 75 was not going to happen even as an experiment

[ivory marsh]

Mine has made it to ||45||, but I now have to force close my computer since it’s completely frozen.

[random lintel]

Does AWS still do free trial months

[warm gazelle]

part 2 is going to take approximately 4 * 10^17 years to brute force...

[warm gazelle]

several years is a bit of an understatement lol

[ivory marsh]

Well they’re the same for me since I can’t wait several years or several universes. I wonder if the needed memory would fit in the universe.

[normal dragon]

I want to see someone run the brute force version on amazon

[ivory marsh]

Also since I had to restart my computer windows finally got to install updates 😭

[magic oxide]

quite easily actually

we're only talking about like ||petabytes, and google has way more than that||

[random lintel]

AoC scorns independent coders; new challenges heavily biased towards tech giants

[magic oxide]

can't wait for day 12 to be "solve the TSP for this graph with 1 million nodes

[kindred zodiac]

dont they have like yotabytes of storage

[kindred zodiac]

1billion rows typeshit

[compact wasp]

chat can i make it

[slender ravine]

I stupidly run both a python and c# version simultaneously (and also c# version is wrong) and my raspberry pi froze and I need to restart it

[slender ravine]

Btw the math is wrong, you cannot have 2.79×10^16 seconds then become 4.25×10^17 year,
You should divide by (60*60*24*365.25)
not divide by 60/(60/(24/365.25))
= 60*((24/365.25)/60)
=24/365.25

[slender ravine]

This would be amount of year if g is correct

[cyan sleet]

years is a vast understatement I think

[hollow wharf]

Today reminds me ||of the fishes challenge||

[normal umbra]

yo same

[hollow wharf]

||Will solve it the same way||

[oak pier]

.

[hybrid gull]

Are you thinking of the ||Lanternfish one from 2021? I thought the same. My pt2 solution felt very similar to that.||

[hollow wharf]

Absolutely yeap

[vivid finch]

Start it from today

[spice hemlock]

d4\

[bleak sable]

You don't have a botnet to help you bruteforce your solution? Amateur

[fervent anchor]

yea i know it's gonna make number ten

ten for ten years

why haven't I realized this sooner lol

[karmic coral]

bro is cooking

[hollow wharf]

||Solved day 11 exactly like day 6 of 2021 and part 2 was just about changing the amount of blinks ||

[woven oriole]

i am more curious about this expression derivation

[warm gazelle]

oops i didn't even notice my formula was wrong, thanks

[warm gazelle]

honestly i would've written some GPU code to see if i could brute force this, but unfortunately this problem can't really run in parallel super effectively

[warm gazelle]

i made this excel sheet (spoiler-ed because it has my answers to different blink iterations) but essentially i fit an exponential function to the time between each blink iteration (graph not shown in the image, but the data is column B), then found the integral of that exponential function to get the total time for a given iteration

[slender ravine]

I want to try find a value that can cause all code that use long or ulong for storing the number break (or eq.)

[woven oriole]

damn manual exponential regression

[slender ravine]

slight issue, your 2, would have already compute 1 and 3 would have compute both 3 and 2? or I misunderstand

because the next interation taking more double of the last doesn't sound reasonable

[warm gazelle]

oh, the time for anything under iteration 11 is filler data because it was so fast it was read as 0 seconds exactly. but you can't do exponential regression when there's a 0 in the data since k^x cannot equal 0 (where k is a constant > 0). so instead i just put the function 2^x * 0.0000001

it's probably not the most rigorous way to do this and would probably give incorrect results, but i'm not trying to get a super accurate time estimate anyway

[warm gazelle]

i'm not sure what you mean

excel does all the regression math for me

[woven oriole]

Yeah I didn't mean manual nvm

But regression

[slender ravine]

Im talking about 26->27->28

it each take like 2.5* the last time, which don't make sence because data could only maximum double in size

[karmic coral]

is anyone doing aoc in mojo

[hollow wharf]

got released now?

[dense gale]

i really dont want to drag myself into this i have no motivation lol

[unique osprey]

You keep talking about not starting yet and not having motivation to do it. What do you want from us? What do you have to lose to just start?

first step is click on the 1 and reading the problem

[velvet stag]

writing and sending the comment is most of the fun, I don't think there's something expected from us

[warm gazelle]

oh yeah that is weird

honestly i have no idea why that happens

[slender ravine]

I think you count the whole time from 1->28 instead of just 28, for what I meant

[warm gazelle]

actually i think i copy the stones list before working on it, so maybe that adds to the extra time when n is large

[slender ravine]

Could be

[warm gazelle]

that can't be since the time for 13 was less than the time for 12

it could just be that my processor was also doing other stuff during that time (e.g., another application was open)

[slender ravine]

Hmm fair, can you send the code in d11 spoiler?

Also could use CPU time instead

[warm gazelle]

#1316269903363440690 message

[slender ravine]

You did count the 1->28, which mean that you don't need the integral, just the point on the equation
It can because you count:
1->13 and 1->12 seperately, and they just too quick that cpu allocation time is more important then the actual time at that time

[warm gazelle]

||```py
def blink_n(stones: list[int], n: int, verbose=False) -> list[int]:
stones = list(stones)

for i in range(n):
    start = time.time()
    blink(stones)
    end = time.time()
    print(end - start)

return stones

this measures and prints the time for each iteration, no?

[slender ravine]

you need to do the whole process once,

like for having 13, it won't take the previous result of 12 and calculate 13, but it need to calculate 12, 11, 10 etc.

[unique osprey]

<@&831776746206265384>

[frigid mural]

!rule 5 We cannot assist with malicious projects, nor direct you to them. I'd strongly advise you against engaging with these, please do not ask for help on this or related matters.

[marsh current]

Rules

5. Do not provide or request help on projects that may violate terms of service, or that may be deemed inappropriate, malicious, or illegal.

[unique osprey]

Also this channel is for advent of code

[naive wharf]

if you don't want to do the AoC, I'd recommend doing the (in my opinion) more interesting Advent(2) instead

[slender ravine]

Tbf that sound fun, would check it out after AOC probably

[karmic coral]

i wanna do aoc in assembly. the older the language the better

[rapid wolf]

assembly is interesting for AOC, but quite painful when it comes to parsing the input.

[long egret]

you can solve all of them by hand. the oldest method of all

[distant igloo]

I solve every problem with an abacus and a chalk board

[unique osprey]

Is it related to aoc?

[frigid mural]

Just came backfrom a little hiatus to the rust discord and god the unsafe + SIMD war crimes they're throwing around are amazing

Take day 10 for example, people are actually hitting 58 microseconds??

[distant igloo]

Time to solve or execution time?

I guess it must be execution time since it's Rust

[unique osprey]

Nah it took me 58us to read the problem and come up with a solution

[distant igloo]

Well there are people out there solving puzzles with AI and solving in seconds but microseconds would be a bit much I guess

[sly shale]

you don't necessarily solve it everytime, but we can cache the results ongoing

[karmic coral]

Handwritten aoc with mathematical proofs sprinkled on top. Paper and pencil 😂

[sly shale]

there are commercial apis for hcaptcha available out there

[earnest forge]

GLHF!

[peak dock]

good luck have

[magic oxide]

good luck have?

[earnest forge]

Good luck have factorial

[shut trellis]

<@&518565788744024082> Good morning! Day 12 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/12. Good luck!

[compact wasp]

why's this one so tedious 🗿

[devout tusk]

That's not the right answer. If you're stuck, make sure you're using the full input data; there are also some general tips on the about page, or you can ask for hints on the subreddit. Please wait five minute before trying again. [Return to Day 12]
😦

[compact wasp]

so many edge cases bruh

[late frigate]

finally lb position...

[minor cave]

#1316635269855711253

[devout tusk]

WTH was this question?! 91/230

[compact wasp]

it's simple just tedious

hence i didn't feel like implementing

[devout tusk]

but still would've ranked above me

[compact wasp]

||just trace the edge and record direction changes||

also funnily enough i think this would've been trivialized with ||shapely||

could've gotten #1 if i remembered that existed in time

[pine tiger]

actually not that many if you store them ni a good way

[compact wasp]

yeah i realized that a bit later 💀

[feral hazel]

i'd hope so
the whole puzzle is counting edges :]

[compact wasp]

😭

[devout tusk]

how many of the global lb people are on this server?

is there like a private leaderboard for that?

[peak dock]

match usernames with https://paste.pythondiscord.com/STODH5JTTZCNATIUUKWL5IPBAE

💀

[woven scaffold]

It seems llms are struggling a bit more with part 2 today

[peak dock]

although this one is like 20 minutes old

[woven scaffold]

Seeing some more reasonable times and names I sort of recognise

[compact wasp]

lol yea

[hollow wharf]

MFW I'm opening my phone so early the advent of code ghost ping isn't a ghost ping anymore

[magic oxide]

i enjoyed today

[compact wasp]

masochist

[magic oxide]

i didn't do that well

but i got to use my favourite data structure

so it was worth

[compact wasp]

what's your favorite

[magic oxide]

||dsu extra whitespace||

[compact wasp]

whitespace

[magic oxide]

is the text not white

[hybrid gull]

lol wait a sec, am I reading this right? So 98 of the "people" who got lb for pt1 failed to make it on the lb for pt2? Haha, that must really get their knickers in a knot 😂

[thin topaz]

I counted 1 more sides out of all the regions in the puzzle input...

[oak flax]

todays problem fells hard

[cyan sleet]

wtf is purescript?

PureScript is a strongly-typed, purely-functional programming language that transpiles to JavaScript, C++11, Erlang, and Go.

huh

file extension .purs

[cyan sleet]

do you actually need it? at least for p1 it seems overkill

[magic oxide]

i mean ||i just prefer to work with a dsu myself than do a bunch of bfs/dfs||

||works very nicely with my usual dict[complex, ...] approach as well||

[cyan sleet]

now I need to figure out how to do graphs in a functional language 🥴

please have some hash map type

or some mutable random access type

[peak dock]

mfw we have a purescript coredev in pydis staff

[rapid wolf]

you get normal JS arrays

and objects, of course.

[spice hemlock]

E

[cyan sleet]

D was yesterday, so today logically should be E

[hollow wharf]

Part two today surprised me..

[low condor]

I can't say I was surprised by P2 today. But it was a bit of a pain to write

[hollow wharf]

yeah

[floral hare]

looks like I'm 21 on PyDis's 2024 leaderboard

yet I'm no where near the top 100 in the total lb

[buoyant slate]

what should i do with day 12

im not doing official but im doing just ideas

im doin next years

ignore day 7

[unique osprey]

--- Day 12: Garden Groups ---

[buoyant slate]

which is

[unique osprey]

The official name for the day's problem

That's what you're asking for right? A good name for the solution?

Another good one might be ElfZoningCommittee.py

[magic oxide]

AbnormalFencePricing

[naive stone]

any helpers out there feeling helpy? My day 12 part 2 code works for every example given, but not my input. I've tried coming up with my own examples, and it works for those too

[unique osprey]

Try this one
||

XYX
YXY
XYX

||

That's the one that got me

[naive stone]

oh wtf.. my code thinks one of the X's has zero sides

my code actually super failed all of these plants i think. weird

oh

damn

that is crazy

[unique osprey]

||yeah. mine was not detecting the corners when the diagonal was the same char||

[naive stone]

yep

dammit, i can't believe there really were more edge cases i didn't think of

[rancid creek]

good test examples can lead to exposing what you're doing wrong. 🙂 nice testcase.

[naive stone]

appreciate it @unique osprey

[unique osprey]

Hey happy to help

It's not like i do any real work in december

[rancid creek]

My code only checked ||orthogonal directions|| so I didn't run into that edge case.

[unique osprey]

How did you find ||inside corners||?

[rancid creek]

answering in the spoiler channel

[zinc kindle]

And I got 1306/18722 . The strategy of "sleep on the problem then come back in the morning" was successful eventually, but wasn't good in terms of leaderboard placement.

[unique osprey]

The problems release at 10pm my time. Last couple of years i tried staying up until i solved it, then i tried what you said and just reading the problems then go to bed, but that would always keep me up thinking about it haha

so this year i just do the whole thing when i get to my desk in the morning

[rancid creek]

Leaderboard seems worse this year. Even the more hardcore aoc's are down a bit this year. JP was around top 10 the last 6 years in a row, and is currently 83rd (although annoyingly number 1 on the private leaderboard we're on). It feels like there are a ton of people gaming the system, possibly with AI this year.

[rancid creek]

Getting '*'s and finishing the puzzles is an accomplishment in of itself. Be proud of what you've done and worry less about when you get it done. 🙂

[peak dock]

yeah, apparently llms struggled with p2 today and there was practically no overlap between top 100 first star and top 100 two stars

[rancid creek]

good. 🙂

[peak dock]

coincidentally i had my best rank so far this year on p2

[rancid creek]

nice!

[hazy ermine]

hahastinkypoop

[mossy basin]

i spent a lot of time trying to figure out how to trace edges from a batgrl.region

[rancid creek]

I had a hard to track error in my code (had the wrong variable listed in an equality check). Code looked fine, but it wasn't until I debugged line by line I saw it.

[mossy basin]

but i couldn't figure it out

[hazy ermine]

whats a batgrl

[zinc kindle]

I was already planning on starting at least one puzzle over 24 hours late, due to spending time with family. I'll be glad if I can just get all 50 stars before 2025 (and finish my backlog of other years before Dec 2025)

[unkempt solstice]

Hi, I just joined the meeting and this is first meeting, what are guys working on and how often do you do this

[rancid creek]

Pretty much every year. And welcome. Have you been keeping up with the first 12 days ?

[rancid creek]

https://github.com/salt-die/batgrl

[unkempt solstice]

No

What can I do to catch up?

[rancid creek]

https://adventofcode.com/2024/

[rancid creek]

9 hours until new puzzle!
/happy dance

[strange zinc]

after today's p2...

i was like "ah, simple enough". Then had to code it and was like: "nah, nvm"

[cyan sleet]

  Could not match type

    Int

  with type

    Number
```![pithink](https://cdn.discordapp.com/emojis/652247559909277706.webp?size=128 "pithink")

mfw

spago install integers

[rapid wolf]

classic

[cyan sleet]

any rouletters who can tell me how to read strings from stdin in purescript?

all things I've seen has been mildly cursed

[magic oxide]

cefqrn had this commented at the top of his sol https://stackoverflow.com/questions/35705155/how-to-read-a-file-in-purescript-readfile

[cyan sleet]

ah, yes

readTextFile UTF8 "/dev/stdin"

[cyan sleet]

want to elaborate how I mutate Array?

[rapid wolf]

oh, that you don't get (outside of ST), which is a shame.

[cyan sleet]

AAAAAAA

I just want a visited set 😭

with not absolutely horrible performance characteristics

I don't want to be linear searching through a list for this

[hollow wharf]

god i love discord's markdown parsing

you can't send a spoilered codeblock that contains ||

which is an operator used by quite some languages

||```
left || right

🦧

[zinc kindle]

testing: || a \|\| b ||

testing2 ||

a \|\| b

[cyan sleet]

this day sure is a day

somehow we've avoided graph problems in functional languages previously

[zinc kindle]

Yeah, the both parts:part1 ratio is the highest it's been this year so far. Still not as bad as 2013/day 21, but we've got 2 weeks left still.

[cyan sleet]

like, I know what I want to implement

but mutable state in functional languages is...not fun

[cyan sleet]

oh no

this is not what I want to see in a functional language

RangeError: Maximum call stack size exceeded

[cyan sleet]

oh, it was an actual bug

apparently trailing newline blows my code up 🥴

fun bug actually

bug: ||I fill the outside of the grid with ? and the zero width last row is incorrect, but my code things it's a full length row, so it traverses into infinity chomping up ?s||

[hybrid gull]

GLHF

[shut trellis]

<@&518565788744024082> Good morning! Day 13 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/13. Good luck!

[ivory marsh]

Wow, I think I’ve solved a lower dimensional equivalent of todays problem in APL before

[compact wasp]

30/25

meh

[magic oxide]

ah yes wrong answer

how tf do i debug this

[compact wasp]

i spent a while on pt2 to ||look up documentation for z3||

[magic oxide]

i hate ||floats so much||

i could've leaderboarded on p2

pain

[compact wasp]

lol ye it was slow

[devout tusk]

98/215 😦

[devout tusk]

pretty much same

[pine tiger]

haha thank god i ||converted everything to fractions||

[compact wasp]

lmao nice

[devout tusk]

my ranks keep dropping. I'm without internet for the next 3 days. I'll likely be out of global leaderboard

[magic oxide]

wtf i beat you for once for p2

[compact wasp]

i think the #1 stopped aoc so i should be able to get to #2 tomorrow

[devout tusk]

i hadn't used z3 pretty much since one of the questions in last year's AoC

[magic oxide]

i didn't use ||z3 either||

[woven oriole]

wtf strong

[thin topaz]

causally getting screwed by floating point imprecision

[mossy basin]

well, at least you know you aren't time traveling

[slender ravine]

Why would you need floating point 💀

[hollow siren]

This is my first project

[slender ravine]

#python-discussion
This channel is for only advent of code discussion only

[small solar]

hi

[torn sapphire]

hmmm

i think i see the solution to today but don't have the time to sit down & solve it 😭

[ivory vigil]

I used z3 but I think it might be a pretty simple ||pair of linear equations||

[torn sapphire]

that's basically what i was thinking the solution was :P

[tender swallow]

||it is||

[short sedge]

@umbral vortex hey remember me

[scarlet locust]

||it was, implementing cramer's rule made it a breeze||

i do wish part 2 showed the correct output for the example, though... would've made debugging easier

[lime fulcrum]

was wondering why my code stopped working. hmmmm

[wary river]

accidental backspace

[lime fulcrum]

yeah swapped the example input to the real input

[golden widget]

Replit it's not good

Pls user VSC or Pycharm or Github for Development code

[hollow siren]

Okay, man
Thanks

[golden widget]

How can you take part in the advent calendar?
I don't really understand

[unique osprey]

you can read through #aoc-faq

[tulip harbor]

I hate that I've once again reached the point with AOC this year where between work and other stuff I don't have the time to keep up. 😦

[ripe meteor]

You cant spare 30mins in a day to save Christmas?

[naive wharf]

There are other, faster advent calendars, like a 4x6-square bar of chocolate

[slender ravine]

Just sleep less, simple
<- Wake up at 5am, hopefully complete in an hour and hopefully can go back to sleep

[lime fulcrum]

bold of you to assume I get it done in 30 minutes. because usually i don't

[ripe meteor]

Less? 🙊

[tulip harbor]

I could speedrun a whole chocolate advent calendar, like one of those Tony's Chocolonely calendars 🤤

[tulip harbor]

I wish these took me less than 30 minutes. 😂

[distant igloo]

At least one of them is solvable in less than 30 minutes

[rancid creek]

Some of them take more time than others, but if you feel you're taking too long, time box yourself and then ask for ideas or start googling general methods. Maybe there's just something you don't know yet.

[tulip harbor]

That's a good idea. I'll hopefully get a little caught up over the weekend and finish more this year than I did last year.

Another problem I have is when the problems start piling up in my backlog it's hard to go back and pick one.

[warm gazelle]

just spent an embarrassingly long amount of time trying to debug my code only to realize i was using the sample input for day 12 part 2

[earnest forge]

Good luck everyone

[slender ravine]

what is roulette today

[magic oxide]

i just rolled out of bed

[shut trellis]

<@&518565788744024082> Good morning! Day 14 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/14. Good luck!

[long egret]

this one looks familiar

[onyx aspen]

Wth this is!😭

It looks like a values mixed with lists!

[long egret]

time to bring out the regex

[devout tusk]

What am I missing?!

[pine tiger]

omg

on plane

i literally can’t

[earnest forge]

Wtaf is this part 2

[compact wasp]

tf is p2

[pine tiger]

could you possibly tell me what it is

or would that count as spoilers

[compact wasp]

it's supposed to form a picture but idk what the picture is supposed to look like

[earnest forge]

(Spoilers for p2, obviously)

[pine tiger]

??????

[earnest forge]

Yeah exactly

[pine tiger]

thats actually so nasty

is this the

chatgpt repellent?

[compact wasp]

what does it mean by this

[pine tiger]

because goddamn it sure as hell is a human repellant as well

[compact wasp]

can someone send me a pic of what it's supposed to look like 💀

[magic oxide]

wtf is p2

[devout tusk]

p2 aint' working for me

[minor cave]

#1317359611270922311

[devout tusk]

WTF even was today.... 8/228 💀

[compact wasp]

i still don't know what they mean by christmas tree

[devout tusk]

christmas tree was all the friends we made along the way

[compact wasp]

😦

[slender ravine]

what is p2

[earnest forge]

I posted it in the spoilered image above

[earnest forge]

It's ridiculous is what it is (image in spoiler chat)

[compact wasp]

bro really expects me to sit through 8k cycles

[magic oxide]

?? i still don't understand with the image

[modern terrace]

what exactly is a christmas tree

[earnest forge]

HUGE spoilers for p2

||gur vzntr nccrnef jura gurer'f bayl bar ebobg va rnpu gvyr||

[devout tusk]

This statement is 100% true 💀

[magic oxide]

wait a second

oh ffs

surely not that

[compact wasp]

wild

i was literally just spamming return

waiting for it to show up

💀

[earnest forge]

(it comes at roughly t=||8000|| you're gonna be there a while)

[compact wasp]

yeah ik

||7572|| for me

[minor cave]

can we spoiiler a bit more in this chat?

[compact wasp]

my finger is tired

[devout tusk]

nah, around ||6400|| for me

[minor cave]

thanks <3

[covert arrow]

My flight got delayed so I just started 😩

[modern terrace]

delayed by OBOE

wow

i should place my print()s at more strategic positions

[devout tusk]

i've solve it and if this is actually a hint, idk what you're insinuating 💀

[magic oxide]

can i have a question answered to see if i'm on the right track: ||does this christmas tree include the base of the tree||

[earnest forge]

Luckily I caught that while I was still trying to figure out what on earth p2 meant lmao

[modern terrace]

2033/790

best for this year

[compact wasp]

either sit through everything or guess what eric wanted

[devout tusk]

||yes||

[earnest forge]

You're not on the right track

[compact wasp]

you ain't detecting it

[magic oxide]

???

[devout tusk]

actually best advice

[earnest forge]

It is a hint (it basically gives away the entire puzzle) which is why I didn't put it in plaintext lol

[modern terrace]

how do people figure this out in 2 minutes TT

what if they just have ||an ASCII text image recognition library...||

oh well

[compact wasp]

lol ikr

[woven oriole]

what language is this

[modern terrace]

probably the one used in import this or something similar

[compact wasp]

rot13

[earnest forge]

Genuinely I don't think that would help

[compact wasp]

omw to rank grids by embedding cosine similarity to christmas trees

[modern terrace]

technology isn't that far yet?? or is it just not as obvious to machines as it is to humans

[earnest forge]

Without already knowing ||what the image used is|| you kinda just have to make the assumption Eric is expecting you to

[modern terrace]

i'm not an eric expert so

[compact wasp]

if you had the money then it'd probably work

[modern terrace]

oh my god i didn't even realize ||it looked like it had a snowy or starry background||

[earnest forge]

I'm surprised it manages that but you should probably spoiler it lol

[compact wasp]

oops

[long egret]

chat what is part 2

[woven oriole]

thanks

[woven scaffold]

||I'm trying this but only getting christmas trees out? is there some other condition I need to look for for the easter egg or am I missing somethingelse||

[woven oriole]

how did you figure thatt out

intuition

[earnest forge]

??? ||that is the easter egg??||

[woven scaffold]

wtf

[earnest forge]

Throwing stuff at the wall to see what sticks

[woven scaffold]

||Bruh I thought I was supposed to be looking for an image of an easter egg||

Lmao

[woven oriole]

orz

[earnest forge]

It's the kind of assumption you've had to make before

[earnest forge]

Reading the problem explains the problem

[modern terrace]

||it does form an image, but finding it out programatically is the hardest part :p||

[long egret]

idk if im just really dumb but i dont get it lol

[woven scaffold]

wait

[magic oxide]

dw

[woven scaffold]

I think I have off by one error

gg

[magic oxide]

i didn't get it either

just try stuff

[woven scaffold]

Man

[modern terrace]

||you see that 2 there? that's a no-no for part 2||

[woven scaffold]

Huge reading comprehension failure

-_-

[long egret]

||so i need to find how long it takes for none of them to overlap?||

[modern terrace]

||basically, yes||

[long egret]

eric must've been in a trollish mood when he wrote this

[compact wasp]

i think that's just how he makes the inputs lol

[humble copper]

im going insane

[slender ravine]

||I mean, he probably built it in reverse starting from the tree, so It makes sense why it doesn't overlap||

[tulip harbor]

I'm having a lot of trouble getting the right answer on day 6 part 2 - is something off with my code here? ||I'm recording all of the visited spots along with directions and iterating through putting a barrier on each. Maybe I'm brute forcing it too hard or something, or am I just missing something? It works correctly for the examplehttps://paste.pythondiscord.com/EGRQ||

[slender ravine]

||what is the trouble, time? Or the answer wrong||

[scarlet locust]

||omfg i feel so stupid. i manually looked through 6,285 permutations until i found mine...||

[tender swallow]

1935/2960
personal second best so far, nice

[rancid creek]

if it's stupid and it works, it's not stupid. join the spoiler channel to see my stupid idea. 🙂

[rancid creek]

I'm wondering if he thinks this is something humans will do better at than an AI.

[humble copper]

im so very lost

😭

[scarlet locust]

honestly? not a bad idea

[compact wasp]

for aoc*

[tender swallow]

Eric has said he doesn't consider LLM-deterrence at all when creating puzzles... but this one would definitely do the trick.

[compact wasp]

for actual swe if it's stupid and it works it's probably still stupid

[scarlet locust]

this was a me-deterrence

[pine tiger]

:clueless:

[pine tiger]

may i ask where you got this?

[tender swallow]

https://old.reddit.com/r/adventofcode/comments/18bp8id/why_does_aoc_care_about_llms/kc646i4/?context=3

[pine tiger]

tysm!

[zinc kindle]

Is there a reason why the ||check for a state with no overlaps|| is a valid strategy, and can it be inferred from the puzzle? my strategy was ||look at every possible state and find the state where the total distance between robots is minimized, and assume that's the solution or within a few seconds of the solution||, which worked but had a runtime of ~1 minute (in rust release mode even) since my solution had a time complexity of ||O(n^2) per state, and with n=500 and there's lots of states||

[magic oxide]

might be better to discuss this in the spoiler thread

[rancid creek]

The problem is, that "most" people using AI are just using AI, not asking AI to interpret the puzzle. So it's going to be almost impossible to stop everyone.

[kindred zodiac]

I always use ai to Interpret and break down most my projects because it's actually really really good at it and more people shod do it to be honest

[rancid creek]

As Stickie said, probably best discussed in a spoiler channel. But that's not the way I solved it, and the puzzle wording does not seem to indicate any specific strategy at all.

[rancid creek]

for actual swe if it's stupid, it was the manager's idea. 🙂

[mossy basin]

||i made this assumption because if the robots had some unique arrangement, probably each robot had their own unique location in the arrangement... dunno, i had to make some assumption and this seemed like a reasonable first||

[compact wasp]

interns:

debatably actual swe ig

i'm not swe i wouldn't know

[pine tiger]

haha

i slip one day

and get knocked down to 3

it is what it is

[scarlet locust]

well now that that's out of the way, what would you all say is the hardest day so far for you personally?

[tender swallow]

12p2 because of all the edge cases

[wary river]

day 1 coz that's all I did 😭😭

[karmic coral]

anyone doing day 14

[rapid fog]

finished up a bit ago

whats up?

[karmic coral]

on part 2 is there an output easter egg or am i just looking for a time metricin seconds

[rapid fog]

you're looking for an actual tree lol

at some point in the input therell be a tree showing up

you might wanna check #1047673173447020564 for how it looks

do you mean like your answer?

[karmic coral]

mb

[rapid fog]

your answer should just be the timestamp when the robots make the tree

[karmic coral]

ill check spoilers thx thx

[tender swallow]

your answer is the elapsed time in seconds at which point a tree shows up

[coarse nymph]

me staring at my bruteforced terminal animation waiting for the f---ing tree to appear:

https://i.giphy.com/media/v1.Y2lkPTc5MGI3NjExcmI4OW1wbHRxY3g3bjF5ZnY2c2tkOGRsaHE4amswN245MWE5Ym1ycSZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/1zOggkwwUyB8c/giphy.gif

(i won the staring contest)

[rancid creek]

||I hope you don't miss it!||

[coarse nymph]

|| didn't miss it, gold star for me ||

[scarlet locust]

||good to see i'm not the only one who did this LOL||

[peak dock]

lmao part 2 is the funniest ive seen in a while

[magic oxide]

I'm gonna hope day 12 purescript isn't too hard

would be good to catch up tonight

and then i can stop spending my entire day reading language documentation

[vestal token]

are we allowed to commit aoc code solutions to github? Is adding input.txt in .gitignore enough?

[tender swallow]

yes, yes

https://adventofcode.com/about#faq_copying

[scarlet locust]

how i felt ||after putting 10k numbered iterations in a text file and somehow getting the correct answer from scrolling it||

[peak dock]

no shot

[vestal token]

do most people generally use one python file for each day of advent of code, or one python file for each part of each day?

[modern terrace]

a poll would be perfect for that

[scarlet locust]

my laziness knows no bounds : 3

[peak dock]

i see the first one more often

i have a notebook per day

[magic oxide]

i have both in my repo

[scarlet locust]

depends on the puzzle

[magic oxide]

because i started off doing 2 files and then switched to 1 file

[scarlet locust]

day 4 I split into two files because of how different my methods were :p

[woven oriole]

i generally try to calculate both p1 p2 in same function call, using flags and shit

[woven oriole]

strong

[slender ravine]

But I would copy the solution to a 2024.py at the end of the day

function name:
solution\d+[ab]

[karmic coral]

Do we all get AoC tshirts after

[fervent anchor]

my gosh i'm running day 14 part 2 really slow

[shell heath]

I will open staff application if my server will be complete 30 members dm me to join and get a chance to become a staff

[tender swallow]

this isn't the place for that

[humble copper]

idk why, but this years challenges seem significantly easier than the last time i got this far (2021)

[rapid wolf]

you may have just improved

[hasty wasp]

Just started the advent of code today at day 14 😅

The fact that day 1 is two one liners is wild

[tender swallow]

solutions should be posted in the respective #1047673173447020564 thread for that day, please.

[hasty wasp]

Oh sorry MB

Deleted the message

[rancid creek]

yes, you can share your solutions, just don't share your input file.

[karmic coral]

Just wait for day 25 finish that puzzle and then say you completed aoc 2024 🙌

[rancid creek]

heh. those that know...

[cyan sleet]

🙃

[unique osprey]

Gonna be honest it's my sixth year doing aoc and I've never made it to 25

[rancid creek]

You can always go back and finish ones you missed later.

[pine tiger]

hey does anyone know if i can access adventofcode in china

[shut trellis]

<@&518565788744024082> Good morning! Day 15 is ready to be attempted. View it online now at https://adventofcode.com/2024/day/15. Good luck!

[covert arrow]

Alright I think this is outside the scope of what I can do on mobile

glhf everyone else

[devout tusk]

can't solve p1. i think i'm missing some weird edge cases

[long egret]

try not to make a grid based challenge challenge (impossible)

[floral hare]

Hey doesn't today remind of sokoban?

[tall quest]

Was getting myself back into AOC then realized my brainpower is low, took considerable amount of shifting gears.

[devout tusk]

i'm out of ideas on what i'm missing

[wary river]

yeahh it does

maybe that's a hint?

[karmic coral]

or dont know 😂

[hexed axle]

https://discord.com/channels/267624335836053506/1317726648162062378

[rancid creek]

make sure you submit an int. I spent too much time after I submited xxx.0 and it said it was wrong, only to try xxx later.

[rancid creek]

reminds me of Robo Rally!
https://en.wikipedia.org/wiki/RoboRally

[woven oriole]

okay tf is p2

[rancid creek]

https://discord.com/channels/267624335836053506/1317726648162062378

[earnest forge]

Eh, not enough pits to fall down

[floral hare]

hmm

the way it's programmed does resemble it

but I'd say it's more or so || sokoban || since the problem literally says,

|| elaborate warehouse complexes operated by robots ||

[slender ravine]

Idk what wrongs with my code but the example (which is smaller than puzzle input) take comparable longer to run 🥴

[slender ravine]

Spoiler

[floral hare]

is it?

it's just from p1

[slender ravine]

Still, someone might reading the chat that haven't even open the question yet

So put a spoiler is always a great habit

[karmic coral]

bruh i got #2 on lb for day 15 part 2 sheesh

[rapid fog]

yea my part two just isnt working lol. Idk what to do abt it

[floral hare]

ok

[rapid fog]

im guessing a lot of people's sols dont either

[floral hare]

me: who haven't even started on p1

[karmic coral]

what doe the percentages mean here: Day ⭐ ⭐⭐ | %⭐ %⭐⭐

1. 395 389 | 33.00% 32.50%

[earnest forge]

% completed out of everyone on the board, I would assume

[rapid fog]

oof

[karmic coral]

what is the oof far what happened

[earnest forge]

...bearing in mind today's problem has only been out for an hour

[rapid fog]

yea but look at all the days

it is slowly reducing

[earnest forge]

That happens every year

[karmic coral]

yea the leaderboards are subjective. it doesnt account for difficulty of programming language or ease of

[rapid fog]

ik im usually one of the ones who drop out lol

decided to stick it out this year atleast

[pine tiger]

HAHAHAHA

[earnest forge]

:P

[karmic coral]

i started 10 mins late im coming for you 0o31 😂

[earnest forge]

'I'm coming for you'
most of my p1 solves being less than 10 minutes

[karmic coral]

they dont account for hardware too some may have old old laptops or computers

[earnest forge]

To be fair those people are probably not trying to compete for global

[peak dock]

ehh the bottleneck is dev time not runtime

[earnest forge]

After the first week or so, yeah well actually it's kinda both for the first week

[karmic coral]

well software is also subjective too ive been doing aoc on command line

yeh dev time

i have a headache need agua

[pine tiger]

ok 30 pt diff

it is what it is

[earnest forge]

Tbh persistence is how I won the pydis board (and got 78th global) in 2019 lol

[pine tiger]

in retrospect intcode seems fun

at the time it was awful

i still think that network design spec was really vague

[modern terrace]

1786/1819

first time since day 5 i had both ranks below 2k

[magic oxide]

honestly it's somewhat nice to have just gotten home

i can take my time instead of trying to rush and not getting leaderboard anyways

[earnest forge]

I enjoyed it at the time too

¯_(ツ)_/¯

[earnest forge]

Also I liked both the lockstep and delayed networking specs

[compact wasp]

damn i took a nap and missed it

[karmic coral]

I’ll take a nap over aoc any day you won pretty much

[tender swallow]

937/2499
personal best for p1, nice

[slender ravine]

My best is this:

  3   00:05:21   1671      0   00:09:05    883      0

ik not great

[tender swallow]

<1000 for p2 is great

[slender ravine]

that was just typing speed probably

[spice hemlock]

e