advent-of-code

[cyan sleet]

:/

so basically don't write prolog for what prolog does well

[hidden musk]

rip

[noble skiff]

I'm kinda regretting going back and doing day 5 part 2 the efficient way

this is pretty annoying code to debug

[woven sable]

what'd u write it in?

[dense osprey]

took a break

will solve part 2 now

optimal solution inc

[noble skiff]

rust. but the annoyingness has nothing to do with the language

[hidden musk]

this one might be a >24hrs ngl

[dense osprey]

i dont even know how to approach this in a non naive way

i just cant come up with a way

ive always avoided range style problems

havent solved anything similar yet

[tardy vale]

I got stumped a little by part 2 as well, but doing a little data exploration on my input and I have some new ideas

[hidden musk]

maybe you should activate your inner monoprologue

[hazy ermine]

same, hopefully we won't be getting more ranges...

[cyan sleet]

ffs, =< and => are the most cursed ways to write inequalities

[hidden musk]

yes

[weary kelp]

lol, erlang has =< and >= 🙃

[cyan sleet]

it's =< and >=

[hidden musk]

i'm about to manually min the output ngl

[cyan sleet]

I'm just dumb

[cyan sleet]

do you have a list?

[hidden musk]

no i just have the query results. idk how to make a list of them either 🥴

[cyan sleet]

wdym query results?

[hidden musk]

like, i'm doing it interactively

[cyan sleet]

🥴

[hidden musk]

shh

[cyan sleet]

oh, I managed to make something that finds all shifted intervals

now how tf do I find the non-shifted ones?

[noble skiff]

ugh, finally got the fast solution for part 2

smh

31 us for part1 and 2

[hidden musk]

i got part 1 finally 🙄

[cyan sleet]

prolog recursion is such a mindfuck

append([], List, List).
append([Head|Tail], List, [Head|Rest]) :-
    append(Tail, List, Rest).

[hidden musk]

🥴

yeah this is a >24 for pt 2

[cyan sleet]

I'm missing like one part to be able to start making an attempt

inverting ranges

[low condor]

I could not even begin to imagine doing today in an unfamiliar lang

[hidden musk]

i think if i was familiar with prolog it would be a breeze

[low condor]

if you lot can do it, very impressive

[lusty rover]

I still haven't gotten part2 in Python

[terse comet]

part 1 is not too bad
I still haven't really started working on part 2 though

[low condor]

I only just did p2

[dense osprey]

wrote a naive solution

[low condor]

took me hours and rigorous tests

[lusty rover]

Wanna help me?

[dense osprey]

i hope it finds the solution within the next 3 hours

[low condor]

I kinda want to refactor my code first

[dense osprey]

idc abt optimal solution

i hate range stuff

[lusty rover]

good luck

[low condor]

I don't need luck - I have tests

[dense osprey]

spoiler sol

idk how to name my variables

[terse comet]

bruh

[dense osprey]

bruh contains the mappings

[hidden musk]

bruh

[low condor]

looking at the reddit memes I'm mildly surprised today is brute-forceable

[honest nacelle]

okay, back to work on day 5. Wish me luck

[low condor]

i took one look at the input file and thought the brute force would be a terrible idea, but seems you can make it work

[hidden musk]

i think i'm just gonna run the brute force overnight and see what happens

wait part 2 ain't even that hard

[cyan sleet]

either you're a genius or you missed something

[dense osprey]

then why do bruteforce

[hidden musk]

latter it is

[dense osprey]

this check shouldnt affect my runtime right?

1e9 of those should fit in 1 second

?

but division is slow

[hidden musk]

reiterating my dislike for the name of itertools.batched

[cyan sleet]

chunked

[dense osprey]

2.5seconds for 1e6 checks

[hidden musk]

that's quite slow 🤔

[dense osprey]

yea

it is

1sec

[cyan sleet]

I hope you're using optimization flags

[dense osprey]

25 for 1e7

4:10 for 1e8

ok ill let it run while i sleep

41:40 for 1e9

assuming that sol is below 5e9

thus around 3 hours

i just want the golden star

[covert arrow]

maybe num & 0x7FFFFFF == 0 would be faster? although it won't be a round number anymore

[dense osprey]

removed the statement completely

just wanted to time my program

without using a buildin timer

but thank you

[cyan sleet]

f(A, [X], [range(A, X)]).
f(A, [range(L, R)|Tail], [range(A, L)|Rest]) :- f(R, Tail, Rest).

:- f(-100, [range(1, 2), range(4, 9), 100], L), writeln(L).
```that'll do it

[range(-100,1),range(2,4),range(9,100)]

my head cries

[compact wasp]

i think that's how crying usually works

[dense osprey]

i want a math heavy problem

not advanced math

[compact wasp]

can i send you my algebra hw

[dense osprey]

20$ and ill do your homeworl

[compact wasp]

nah not worth

3 lowest hw grades get dropped and there are only 2 left

i could get 0s and still ace the class

[dense osprey]

cant you just cheat on your homeowrks

i wish i had homework grades

i love chatgpt

[dense osprey]

whenever my teacher asks something, i ask chatgpt the same question

and then use chatgpt's answer

[compact wasp]

[dense osprey]

and get an a in class

[dense osprey]

oh abstract

[compact wasp]

yeah

[dense osprey]

tried öearning it last year

gave up

[compact wasp]

rip

[honest nacelle]

Man I thought I had a good solution, part 2 taught me different

[dense osprey]

3min runtime

or 4

[hidden musk]

wild

[dense osprey]

i love easy leetcode problems

[honest nacelle]

easy

[simple raven]

3min runtime wouldn't solve a leetcode problem 😄

[dense osprey]

like todays daily leetcode

[dense osprey]

im not talking abt this problem

[simple raven]

ah k

[honest nacelle]

gonna let this brute force run while I try to think of something better...

WAIT

[river hollow]

Part 2 is 34 milliseconds for me

[honest nacelle]

I'm so dumb

WAIT

I'm not dumb :9

well I am but not in the way I thought

[river hollow]

How is the bot leaderboard here different than the AoC leaderboard? One has me at 116th, the other has me at 16th.

[cyan sleet]

arguable

[honest nacelle]

so it turned out to be 😦

[cyan sleet]

we all be dumb

I'm bonus dumb

[honest nacelle]

WAIT I'm not NOT not dumb maybe

[cyan sleet]

Schrödinger's dumb

[dense osprey]

3 jane street guys in top 100

[feral hazel]

wasn't there a site with the entire aoc leaderboard for the python server or was that taken down in favor of the command?

[dense osprey]

i thought they work 24 7

[dense osprey]

you meant he private leaderboard?

its still avaiable

[feral hazel]

yeah the global command links to the global leaderboard

[simple raven]

I finally figured out the solution

for i in range(1e20):
    response = submit_answer_to_AoC_and_parse_response(i)
    if response:
        break
``` see you in a week

[terse comet]

range(1e20) is a TypeError lol

[honest nacelle]

Mmm... is it a good solution if it only works because of the specific values in the source data?

[dense osprey]

yea good enough

[simple raven]

eeeh how do I write it correctly?

[terse comet]

10**20
(1e20 gives you a float)

[dense osprey]

how do ppl solve a problem in 90seconds

or a part of a problem

[feral hazel]

yeah the bot used to link to a full leaderboard
#aoc-bot-commands message

[simple raven]

good spot

[river hollow]

Lots of pre-existing libraries they glue together, particularly for parsing.

[terse comet]

how specific?
usually the input is more constrained than the problem describes, but oftentimes includes one or two edge cases not included in the example

[honest nacelle]

I'll let you know if this works

[river hollow]

You can find such a leader and click through to their github and find their code ... and libraries.

There's at least one person who wrote their own AoC DSL.

[honest nacelle]

DSL?

[river hollow]

Domain-specific language

[dense osprey]

someone got day 4 in 1:22

how is that possible

[river hollow]

Answered above

[honest nacelle]

Honestly though I can't even understand the problem or physically write code that fast

it's wild

[dense osprey]

ah

[dense osprey]

oh

[dense osprey]

damn

[dense osprey]

understanding the problem already takes more than 3 minutes sometimes

[simple raven]

when you watch videos of people solving the google coding challenge I would expect they are the same people that some these in minutes, crazy to watch

[river hollow]

Here's one guy who's in the top 5 (For day #4) : https://github.com/nthistle/advent-of-code

[dense osprey]

codejam?

[river hollow]

There's a lot of libraries, very compact code. So it's lots of practice code golfing and doing competitive stuff.

[dense osprey]

ah

[simple raven]

that might be the name

[dense osprey]

watched his video on day 4

[ripe meteor]

we've escaped the great filter once more, see you tomorrow peeps

[river hollow]

I found one person who had tools that hits the AoC website and grabs the sample data, then the input data; downloads it, builds a skeleton project with their AoC library, and then pops open the IDE onto that project.

I got Day 6 in 2021 in 9 minutes, 36 seconds. I was ranked 2838.

Haven't done any better. It's futile, since I'm not going to write my own library or any of these tools.

[hidden musk]

prolog when?

[ripe meteor]

Never

Im struggling enough as it is

[hidden musk]

huh

python?

[ripe meteor]

Ye

Idk if im just out of practice but this year feels much harder

Maybe im just dumb, who knows

[hidden musk]

you fell off

[ripe meteor]

Peaked many years ago

Tbh today's part2 wasnt that hard

I think i just panicked when i read it lmao

Put it off the entire day

[dense osprey]

i mean i could write soem functions before

but idk how useful they would be

[hidden musk]

i know what i want. i don't how to get it

[dense osprey]

since üroblems so far didnt really require anything that couldve been prewritten

[ripe meteor]

You're using prolog today?

[dense osprey]

except reading input

[cyan sleet]

I'm going all recursion, screw this

I've found a tool that can actually do stuff

[hidden musk]

ye

[hidden musk]

oh?

[cyan sleet]

recursedion

[plush grove]

are we allowed to use libraries such as regex in advent of code?

[honest nacelle]

BRUTE FORCE FTW

[dense osprey]

why shouldnt you

[terse comet]

yeah - you are discouraged from using AI tools like GPT, but other than that you are allowed to (and incentivated to) use any tools available

[dense osprey]

i hate regex

had to use it once

[honest nacelle]

regex is a great tool, but not a hammer

[dense osprey]

a year ago

i still remember reading the docs and not understand them

the pain

the suffering

regex

[honest nacelle]

it comes with practice, personally I don't write any regex without using regex101 or a similar site

[river hollow]

Jimi Hendrix had the same problem. And he turned out okay.

Mostly.

[dense osprey]

i rarely use it for searching

found out that the ctrl+f search thing has regex

[river hollow]

Of course. You can't write a program with zero libraries.

[honest nacelle]

I use it a lot for search and replace, can be really handy for cleaning up input and stuff like that

[jovial jay]

false, harder, but feasible

[river hollow]

It seems dumb to debate it.

But I look forward to your live-stream of flipping switches on your Altair to get one of the solutions running.

[pine ravine]

whats the library people are using to import the data and submit results

[river hollow]

This one? https://pypi.org/project/aoc-lib/

[compact wasp]

i use my own

[placid lake]

i use my own as well

but salt's is quite good https://pypi.org/project/aoc-lube/

[river hollow]

Yeah, personal bespoke ones abound

[compact wasp]

i could probably publish mine on pypi

[placid lake]

i need to go on site tomorrow so i can't stay up to do AoC tonight :(

[compact wasp]

rip

[fleet fox]

F

[compact wasp]

i'll try to leaderboard for you

[placid lake]

cheers

best of luck

[compact wasp]

ty

[feral hazel]

oh hey today was my 200
at least on this account

[magic oxide]

nice

[woven sable]

i got zero for my answer wtf 😭

it works for the test data but not my input data

and i made sure to ||clear all the old seed values j to get the final location mappings at the end||

[sleek cave]

sees day 5 part 1
"what a beautiful day for another ||naive simulation||"
answer passes :D
sees part 2
oh.

time to ||overlap intervals!!||

[plush grove]

i'm going for the minimalist approach this year

[onyx flicker]

fuck i'm just now getting started. part 2 gonna be interesting

[magic oxide]

I had like -600mil for my answer at one point

[onyx flicker]

i now have ideas
brb

[jovial jay]

Day 6 is going to be better, right?

[river hollow]

Ooo! Where can you get that page?

[feral hazel]

https://adventofcode.com/2023/events

[river hollow]

Thanks! I've never found that before

Oh, derp. It at [Events] in the top menu

[woven sable]

i'm confused for part two

||so even if you have a map that entirely remaps a range, you still keep the original range bc it stemmed from the initial seeds?||

||bc now i j have a 0 in my ranges which kinda fucks up my code||

[zinc kindle]

||No. The number of seeds is constant after each remap, you don't keep the old ones.||.

[woven sable]

how would it be constant though?

||ranges can be split into multiple new ranges no?||

[onyx flicker]

i have a plan of action af but i require sleep
must save til tomorrow 😔

[zenith elbow]

total number of ||seeds for all ranges|| should be constant ||even after splits||

[cyan sleet]

I mean sure, but ||the number of seeds is quite big||

I think ||the number of ranges will grow at worst linearly with the number of shifts there are||

[modern terrace]

||multiple new ranges that should have the same total length as the original range||

[sharp panther]

Language Roulette: Day 6
The language is ... F#

[hidden musk]

bruh

@ripe meteor

[magic oxide]

Gb moment

yet again am outside

I'm gonna hope my parity theory is true and day 6 is of normal difficulty for day 6

[earnest forge]

GL everyone

[hidden musk]

f# is surely a lot better than prolog

[shut trellis]

<@&518565788744024082> Good morning! Day 6 is ready to be attempted. View it online now at https://adventofcode.com/2023/day/6. Good luck!

[magic oxide]

gl

ooh math

yay

[noble skiff]

Weird question

Could actually do this by hand in less than 5 minutes probably

[sharp panther]

oh good, today is a normal speed day

[noble skiff]

The data set is so small part b must be something really evil

[jovial jay]

It's more bearable I think. I might actually try

[hexed axle]

That was faster than normal I'd say, wow

[sharp panther]

yep, that was really fast

[earnest forge]

That wasn't too bad

[tepid cairn]

finished already?

[earnest forge]

Messed up my comparison for p1 though which probably cost me at least 20 places though lol

[tepid cairn]

I need to catch up to the previous days first 😩

school needs to stop schooling for a bit

[humble copper]

lmfao

i wish i didnt think too hard about this

def could have shaved some time

[compact wasp]

fuck i missed the start

[woven solar]

what the fuck is this day compared to the previous ones

this should have been day 1 or 2

[compact wasp]

yeah

[pine tiger]

god i'm so bad

[modern terrace]

i have no clue how this works

[pine tiger]

BRO I THOUGHT I HAD TO ||SOLVE THE QUADRATIC FORMULA||

[jovial jay]

I like comprehensive short challenges better than hard to read challenges

[magic oxide]

lmao

I finished today in desmos on my phone

[left pawn]

it wasn't

[woven sable]

still stuck on day 5 lmaooo

[pine tiger]

got baited tbh

[left pawn]

I ended up entering the data manually

faster I swear

[zenith elbow]

would be so sick if today's language was maple or something

[stable shale]

I swear it would be

[covert arrow]

nice

[stable shale]

Was sitting there going it can't be this easy...right?

(especially after P2 last time)

[pine tiger]

the numbers could have been a bit larger 😭

[untold briar]

Well today's challenge is really... light?

First time I've seen such a small input

[pine tiger]

there's been puzzles where's the inputs like a single number

there is a precedent

[untold briar]

I think if they make the input bigger, it would be more interesting

[pine tiger]

!!

[untold briar]

Ppl would then find ways to optimize their code since it would take forever to brute force today's problem

I saw the small input and just decided to brute force it haha

[deep ferry]

💀 i used the exact same code for part 2

and it worked

[candid brook]

Knowing when to brute force is also a skill lol

[untold briar]

Cause the problem is exactly the same

[deep ferry]

yes but the speed requirements

[untold briar]

Yeah

[deep ferry]

my code still ran fine

maybe cuz i'm using an m2 mac

[untold briar]

I'm wondering how fast did today's 1st place do lol

Is there a way to see that on the website?

[deep ferry]

i was about to start coding a smart solution until my old code worked within 6 seconds

[untold briar]

My laptop is kinda weak so it took 10 secs to run the bruteforce solution

[minor cave]

you can see the times here: https://adventofcode.com/2023/leaderboard/day/6

[deep ferry]

||a smart way was quadratic formula, no?||

[untold briar]

Well dang, 2 mins for both stars

😂

[naive stone]

you saying this made me try myself. I was so certain it'd be too slow.... took 5 seconds 😑😑😑😑😑

[untold briar]

And 5mins if you want to be in the leaderboard

[deep ferry]

lmao

[zinc kindle]

Yep, that was fairly easy. Brute forced part 1, wrote more efficient method for part2.
Yep, makes sense, looking back at the input, brute force might have worked.
(also I ended up using ||bisection, an O(log(n)) method, and in retrospect using squares and square roots might have run faster and been written faster||. Eh, still ran too fast for me to notice a delay).

[deep ferry]

in USACO usually they have numbers well over 10^9

[naive stone]

the speed of python amazes me sometimes

[deep ferry]

ikr

[deep ferry]

||square roots... would be O(1), no?||

[pine tiger]

actually

[deep ferry]

for non-bruteforcable problems at leat

[deep ferry]

and i doubt ||"bisection" is O(log(n)). I wonder how you implemented it? is this for pt2?||

[untold briar]

Today's equation had a bell-shaped graph

So we just need to find 1 end of the solution and mirror it

[woven sable]

i finished day 5 omg

after 5 hours of trying it

[deep ferry]

day 5 pt 2 is cancer

[woven sable]

my issue was the way i split ranges 😭

[deep ferry]

ikr

[untold briar]

Also we can take bigger steps to make finding the first solution much faster

[woven sable]

never gonna spend this much time debugging again

[untold briar]

I struggled for 2 hours with that one lol

[pine ravine]

whats the point of part B if i didnt even need to change my code?

[pine tiger]

very good question

we're all trying to figure that out

[pine ravine]

maybe authors forgot a couple digits on the input

[untold briar]

I think they made the input a bit too small

[zinc kindle]

Yep. ||For part 2, used basic variant of bisection algorithm. used 0 as low end, time/2 for high end. Distance travelled is O(1), number possible speeds is O(1). Then return if t is fast enough but t-1 isn't, otherwise standard bisection (except I didn't bother with cacheing||

[untold briar]

If the input is bigger, part B could become quite a struggle

[pine ravine]

you dont even need that... i can do it with a piece of paper in 7 seconds

no code needed

[pine ravine]

there is plenty of room for optimizations

[untold briar]

But because the input is small enough, I think people mostly could just bruteforce it

[pine ravine]

right

[untold briar]

Much faster than sitting there thinking of a smart solution

[pine ravine]

i expected it to be hard so i wasnt online at exactly 11:00 and my coworkers beat me 😠

[untold briar]

Maybe they intend this as a breather? Since yesterday's part 2 was a disaster 😂

Wow yesterday's first place took 8.5 mins to earn 2 starts

[pine ravine]

is that high?

[untold briar]

Well... I took 3 hours

And yes that's really high

2.5 mins faster than 2nd place

[woeful topaz]

Anyone got solution for day 5 part 2

(yesterday)

[scarlet wasp]

There's always #1047673173447020564

[woeful topaz]

I'm sorry for asking that, I just spent way to much time breaking my brain

[woeful topaz]

Thanks

[magic scaffold]

scared for tomorrow 🥹

[magic scaffold]

change you input 🙂

i just removed spaces from input.. rather than map to str and join

[weary kelp]

the calm before the storm

[magic scaffold]

difficulty level of AoC this year

'Fucked' if i%current_day else 'EZ'

[pine ravine]

thats a no no word

[magic scaffold]

the calendar looking nice... i can see gondola

[untold briar]

My favorite part is still the trebuchet

[terse comet]

this channel should remain as spoiler-free as possible, delete it and ask in #aoc-solution-hints or #1047673173447020564

[dire notch]

ok

[jade atlas]

so day 6 was easier than day 1.. why would they do that?

[earnest forge]

My guess is by putting hard days near the beginning he's hoping to dissuade people who might use LLMs

[naive stone]

from the about page on the website:

Why was this puzzle so easy / hard? The difficulty and subject matter varies throughout each event. Very generally, the puzzles get more difficult over time, but your specific skillset will make each puzzle significantly easier or harder for you than someone else. Making puzzles is tricky.

[earnest forge]

I mean yeah but even among the variance in advent of code day 6 was clearly much easier than day 5

[naive stone]

yea, i mean, as soon as i saw day 5 i knew day 6 was gonna be easier lol

[digital whale]

hows day 6?

oh easy?

[peak dock]

yeah

[digital whale]

good for yall

[naive stone]

daw

[digital whale]

daw?

i tried redoing day 3 part 2 this morning but i got so irritated i gave up temporarily

[flat mica]

is there any private leaderboard for this server?

[untold briar]

You could go to aoc-spoilers if you are really stuck

Treat it as a learning experience and do better in the next challenges

[untold briar]

I tried the /aoc join command from the aoc-faq channel but there are only 10-20 active ppl there.

[flat mica]

ah thanks, I am very bad at reading

although tbh they should just put the join code in the channel descriptor

[dim pivot]

I’m really hoping, no monkey map this year

[peak dock]

we have multiple leaderboards because of the limit

[flat mica]

oh i see

[untold briar]

Well we already got 2D challenges

So a 3D or more will probably come

😏

[digital whale]

Tried to see but it's all in a different lang sadly

I'll try to check tho thanks :)

[long jacinth]

well, todays very easy...

i need to return to yesterdays, didnt want to bother in the evening after work

[flat mica]

i like the cube one from last year

[ripe scroll]

I wonder why today's was so easy

[narrow rover]

i applaud people who get this done in a few hours. it takes me all day sometimes. and sometimes i dont even get it./ But I suppose thast what learning and chalenges are all about. I agree that todays challene is so much easier. Not that i have done it yet.

[simple raven]

ah day 6 part 1 is a bit of physics, finally I can feel at home a bit 😄

[simple raven]

who knew that the good old ||quadratic equation|| would make an appearance

[devout turtle]

While the thing does look familiar, I can't remember the last time I actually did something with it.

[simple raven]

well today's your lucky day 😄

[devout turtle]

As in solving day 6, you mean?

[simple raven]

well I don't want to spoil too much but: ||yes, you can solve day6 with it||

[devout turtle]

Hmm... Okay.
I did already solve day 6, but it might be nice to look at another approach, if I have some time left today.

[timber stirrup]

Hey everyone, I'm stuck on day 5 part 2, my solution is to use ranges and split these ranges depending on the values, I have created tests such as:

assert test(range(80, 90), 85, 10, 1) == [range(80, 85), range(1, 6)]
assert test(range(80, 90), 70, 30, 10) == [range(20, 30)]
assert test(range(80, 90), 70, 15, 10) == [range(20, 25), range(85, 90)]
assert test(range(80, 90), 80, 10, 0) == [range(0, 10)]
assert test(range(80, 90), 81, 2, 0) == [range(80, 81), range(0, 2), range(83, 90)]
assert test(range(0, 10), 9, 10, 100) == [range(0, 9), range(100, 101)]

And they all pass, but the number I get is too low, is there any other corner case I could have missed?
The example data passes and I've traced it and everything seems to be correct there.

[covert arrow]

can you send the rest of your code?

[timber stirrup]

I'll send it here, I'll just clean up the code as it's a big mess, give me a few minutes

[covert arrow]

also, we should move to #aoc-solution-hints

[halcyon pulsar]

This year is throwing me around the place

Couldnt even do part 2 day 5

[simple raven]

that one's hard 😦

[devout turtle]

My PC is still looking and counting..

[long jacinth]

day 5 part 2 is easy if you have an eternity to solve it! 😄

i should hook up my raspi and see if it can manage to solve it using my current code before december ends

[devout turtle]

And the RAM, I guess.

[long jacinth]

ram is fine

my solution used 4MB of ram

[devout turtle]

😮

[devout turtle]

That's the right answer! You are one gold star closer to restoring snow operations.

You have completed Day 5! You can [Shareon Twitter Mastodon] this victory or

Whoohoo.

[wet yoke]

Side note by the way, Sir Robin incorrectly tells folks that &aoc join is used to join the leaderboard, when in fact it's a slash command, /aoc join. Using the & form just throws up the help info.

[dense osprey]

input file is small

hope todays problem is easy#

[untold briar]

So that tomorrow we can be roasted to well-done

Or to congratulations

[dense osprey]

tmrw will be an unsolved problem

[long jacinth]

just

solve chess

[hollow wharf]

Hey, I'm new to python

[woven solar]

...that's how I did it. was there another way?

[earnest forge]

||brute-force runs in under 5s||

[woven solar]

meh

that's a weird choice, to make the numbers that small, actually

make it even more of a "what is this day 2 task doing at day 6"

[earnest forge]

||closed-form has off by ones to worry about, brute-force is easier||

[woven solar]

||I did an if-statement to correct the off-by-one, but yeah, it did take a bit of time||

[earnest forge]

Oh yeah I agree that the puzzle is more of a day 1/2 task than a day 6 task

[earnest forge]

Yeah - yours is the better solution but mine is easier to implement correctly and quickly 🙃

(that said, testing against the example screwed me out of several places because of an edge case in the sample that isn't in actual inputs)

Literally 14 seconds from leaderboard (p2) - less than I spent debugging that edge case in p1 :')

Just realised I'm at most the 4043rd person to get 400 stars in all of AoC (4043rd on 2016 day 25, which was my 400th star)
I'm surprised that more people haven't done it tbth

[magic oxide]

I would but no time

yeah today was definitely an anomaly

[simple raven]

||agree, but I was still happy I could put my physics knowledge to good use||

[tall wasp]

today being so easy makes me fear tomorrow.. hell is gonna break loose

[noble skiff]

https://www.reddit.com/r/adventofcode/comments/18bp8id/comment/kc646i4/?utm_source=share&utm_medium=web2x&context=3

so much for the theory that AoC is hard this year to dissuade AI usage

[woven solar]

ugh, dirty link

so much for the theory that AoC is hard this year to dissuade AI usage
oof

[noble skiff]

?

[woven solar]

well, without that theory it just seems like an unforced error

[noble skiff]

unforced error?

[simple raven]

interesting, though I'm quite surprised day5 puzzle came before day6 because even when I could figure out the correct approach on my own and code up 90% of the solution, I still couldn't figure out the correct order of applying my functions which made me fail day5 part2, but perhaps someone with better abstract thinking had no issues figuring out the last bit of the solution

[honest tangle]

hey guys

can somone send me screenshoots of results

[hollow wharf]

thats not even close to topic of the channel

!rule 7

[marsh current]

Rules

7. Keep discussions relevant to the channel topic. Each channel's description tells you the topic.

[honest tangle]

can someone send me results of this code

ok i know sorry

buy i need results

[hollow wharf]

..

[warm berry]

stop posting this large code block

[marsh current]

:incoming_envelope: :ok_hand: applied timeout to @honest tangle until <t:1701875549:f> (10 minutes) (reason: chars spam - sent 4734 characters).

The <@&831776746206265384> have been alerted for review.

[noble skiff]

yeah, I mean I got day 5 part 2 but it was definitely a grind I thought

the algorithmic concept wasn't trivial, and the implementation was definitely not trivial to iron out all the bugs

for a day 10-15 it would be pretty typical

[honest nacelle]

Brute force win every time

[simple raven]

I was lucky in that I've seen the range concept before, but the implementation was beyond what I could imagine in my head... I'm happy for people who got it

[noble skiff]

took me a while. should have stuck with brute force 😛

[honest nacelle]

I noticed that a lot of people skipped day 3 intially, that one seems harder than I'd expect an early one to be

[hollow wharf]

3 wasnt so hard seeing what was in day 5 now

[honest nacelle]

My Day 6 solution was also a bit brute forced, but in what I'd like to think is a slightly smarter way

I tried calculus first and gave up

[jade atlas]

day 3 and day 4 have almost same number of 2 stars.
day 5 and day 6 have almost same number of 2 stars.

[ripe meteor]

in 2022, the first day to have ~46k completions was day 14

they really messed up the difficulty this year

[rapid wolf]

to be fair, 2022 has had a bit more time to accumulate.

[tall wasp]

yeah day 6 hasnt even been open for 12 hours yet, of course the numbers are lower than day 6 for a year ago

[ripe meteor]

idk, you think it would double by next year?

[torn sapphire]

wayback machine shows these as being the stats as of around this day last year

[rich delta]

Bruh tried yesterdays part 2 code using dict

lmao

[lime fulcrum]

Well give it more memory

[ripe meteor]

smh why you gotta bring facts into this

[rich delta]

anyone knows any online compiler or something with larger mem?

[woven solar]

than an actual PC? no.

(also, calculate first how much memory you'll need, to see if it's feasible :p)

[rich delta]

Well just these dicts which store the initial numbers and the end numbers (not the whole numbers)

[odd wigeon]

I found you don't need to actually know the name of the mapping. They're all processed in order.

[woven solar]

Well, sure, so you can estimate how big these dicts will get, multiply this by the about 80 bytes that an int-int pair inside a dict takes, and thereby estimate the memory consumption.

(||funnily enough my estimates aren't that high|| - ||only in the low hundreds of gigabytes||)

[rich delta]

Ok thanks. Will try

[mental pagoda]

✨️

✨️

[hazy ermine]

isnt it just ||quadratic inequalities|| though

[peak dock]

okay but how do you have different syntax highlight for each one

[rich delta]

Which IDE/Editor do u use?

[hazy ermine]

yoo i was wondering something was off

[tepid cairn]

it’s like rainbow indents

syntax highlighting, but solely for aesthetics

cat | lolcat

[peak dock]

vsc

[rich delta]

i use some plugins solely built for pycharm pro. But there are some geberal one for vs code too

[kind oasis]

day6 part1 awakened the mathematician in me
only problem is that he has been sleeping for decades

xD

[untold briar]

The data says it all lol

Day 3 and 5 make people drop like flies

[untold briar]

Sounds like an endless loop

Or you did just a bit too much brute force

I accidentally made an endless printing loop in one of the challenge and an error similar to this appeared.

[rich delta]

Nah not an endless loop. Works for the sample

[summer kraken]

Isnt the difficulty level too random this year?

Day 5, Day 6 💀

[onyx flicker]

day6? more like day0

[naive stone]

people will say this literally every year without fail

[onyx flicker]

lmao

[glad tartan]

day 6 can be solved with paper and pencil

[compact wasp]

actually justified this year though

[onyx flicker]

tbf there always is one problem much easier

[naive stone]

i don't wanna hear it from global leaderboard ova here

[compact wasp]

😔

[onyx flicker]

take it from non-leaderboard 😤

[lime fulcrum]

||these quadratic equations are a bit rough for pen and paper||

[glad tartan]

fine, paper, pencil, and a calculator

[naive stone]

you guys need to understand that eric doesn't care about this being a perfect climb in difficulty. Puzzles are ordered for lots of reasons, one of them including fitting in with the story. Day 6 is when we got to that boat race area, so day 6 is when we get that puzzle. Yes it may be easier than all the puzzles prior, but that doesn't matter to eric

[onyx flicker]

yeye i'm not mad at him lol. i think we all get difficulty goes up and down and is also subjective. just fun to poke fun at it

[naive stone]

lol yea no for sure. There are plenty who are legit criticizing though, which is just crazy to me. It's a single man making a christmas game for hundreds of thousands. Give the man a break lmao

[onyx flicker]

does bot react on christmas in message?

[naive stone]

yea wtf lol

[onyx flicker]

lol

[hazy ermine]

it reacts for tree as well

[naive stone]

TIL

[hazy ermine]

I kinda believe this is very common in software in general, we take free stuff for granted

[lost star]

tree

[onyx flicker]

ohshit more highlight words

[lime fulcrum]

i mean today could have used a larger input

just enough so that brute force isn't viable

[onyx flicker]

it is what it is 🤷‍♀️
i wrote my solution assuming that the input would be too large to brute so it ended up being fast either way

[minor cave]

every tree is a christmas tree in december :D

[naive stone]

i won't lie, i was surprised when my brute force finished in 5 seconds. But eric also is a merciful man, and it's likely 6 was meant as a break from the insanity that is F I V E

[naive stone]

yea i'm scared for 7. I really need to finish up day 5 part 2 before midnight

[hazy ermine]

i just bruteforced it lol

[naive stone]

how long did that take?

[hazy ermine]

and while it was crunching numbers i was thinking about optimising but it spit out an answer before I could come up with one

took about 4m30s

(in Golang)

but saw another Python bruteforce that took about 5m

[naive stone]

i'd be curious to see that

[onyx flicker]

oh anything over 30s to brute feels too icky for me to do

[naive stone]

regardless, i'm currently trying to write a clever solution

[hazy ermine]

I'd need a whole day for that so I saved it for after AoC ends

[onyx flicker]

ye i have an idea for what i want my clever solution to be. but like writing it is a tad annoying and i only have so much time to work on it

[hazy ermine]

ye exactly

[lost star]

What is AoC?

[naive stone]

same yea. currently at work

[onyx flicker]

advent of code

[lost star]

what is it though?

[onyx flicker]

fortunately was able to do today on my break :P

[minor cave]

#aoc-faq explains it~

[lost star]

ty

[hazy ermine]

not too good at naming variables but here you go

[woven sable]

o spent like 5 hours on it last night 💀

second most i've ever spent on a. problem

[onyx flicker]

i was mentioning in my work chat whether or not i wanna spend time writing the clever solution in python or just brute it in go lol

[naive stone]

sweet, go is a crazy language

[hazy ermine]

C bruteforce took 60s as someone here said

[compact wasp]

alexandria ocasio-cortez is the representative for ny's 14th congressional district as a member of the democratic party

[hazy ermine]

https://tenor.com/view/umm-confused-blinking-okay-white-guy-blinking-gif-7513882

[naive stone]

lol fellow new yorker enira?

[minor cave]

I will die on the hill that AOC != AoC 😤

[naive stone]

haha. i did think of advent of code when someone first used that abbreviation though

[onyx flicker]

lol it's so hard for me not to answer with her when asked about aoc

[compact wasp]

nj

[ripe meteor]

F# is based and elegant and I wish I was better at it

[naive stone]

ahhhh

[ripe meteor]

tho it did take me a while to figure out I cant have

{ 1 .. time }

if time is int | uint64

silly error messages are so cryptic

also ionide shat the bed on me today

[earnest forge]

is that your cat?

(/gen)

[pine tiger]

what lalng is this

[hidden musk]

https://learn.microsoft.com/en-us/dotnet/fsharp/language-reference/units-of-measure @cyan sleet gonn use these 👀

this is actually based ngl

[earnest forge]

Yeah type systems with units are kinda cool

[ripe meteor]

F#

Pretty much instant pt2 today even though i brute forced it

[earnest forge]

Yeah p2 was surprisingly small

[cyan sleet]

related, units enforced by the type system is one of my favorite things to see

wrong units → compile error

[hidden musk]

mhm

i saw a similar thing in rust

very neat

today was quite a reprieve

[woven sable]

kotlin was instant too

[noble skiff]

I suspect it must still be a bit annoying to implement over there since const generics are still pretty limited

in C++ it's actually somewhat straightforward (depending how many bells and whistles you go for)

it's actually being proposed for the standard library in C++, though I doubt it gets in

[hidden musk]

idk how well it works, but that seems reasonable

[silver wyvern]

I love getting the wrong answer because I printed a number and assumed that was the answer. But I actually forgot to print the answer

[cyan sleet]

@hidden musk F# feels like ocaml but with batteries included

[modest ermine]

day 6 was much nicer than day 5 lol

[cyan sleet]

day 6 is the easiest task so far

quite boring really

[modest ermine]

yeah

i mean my part 2 would be slow if the numbers were larger

so i could optimize that but i honestly don’t care

[cyan sleet]

you could have made twists to make it a tad more interesting

[hidden musk]

I think you just need to install Core

F# is great because .net though, no denying that

[cyan sleet]

right, you can make brute force infeasible to try to make people ||binary search||

[modest ermine]

i love brute forcing 😩

[cyan sleet]

stuff like |> double just working

[hidden musk]

yeah true

calling functions is weird though

[cyan sleet]

it's just so straightforward https://github.com/algmyr/aoc-2023/blob/main/day06/main.fs

though maybe i've been functionally primed now 🥴

[hidden musk]

ain't no way you wrote Delta

[cyan sleet]

correct, I wrote Δ

[hidden musk]

\Delta

[pine tiger]

decided to hunker down and

[lime fulcrum]

holy moly. thats a lot of work. did you have fun?

[pine tiger]

depends on which day

i get angry every time i have to hardcode something

[cyan sleet]

only 4 days left

[pine tiger]

what

[cyan sleet]

8 more stars

[pine tiger]

haha

[lime fulcrum]

No.

[lime fulcrum]

how do i get to this window?

[cyan sleet]

events

[pine tiger]

there's an events thing

@ the top

[lime fulcrum]

ah. i've been enditing the url to change years. that does make it more pleasant

[plush grove]

HOLY COW I COMPLETED DAY 3 PART 2 FIRST TRY

[low condor]

nicely done :)

p2 is not so difficult after p1; it's getting day 3 p1 done which can be tricky

[plush grove]

hmmm i see

[cyan sleet]

wait what day was 3?

groovy, ocaml, haskell, ...

gears?

[low condor]

yup

[cyan sleet]

I am starting to forget the order of languages 🥴

..., lisp, prolog, F#

[hidden musk]

🥴

[sharp panther]

they should be in the previews in #1047673173447020564 (well, except for the first two days)

[cyan sleet]

well yeah

I could also just look at my repo

I didn't name the dirs by language, but...

[hidden musk]

i don't think i've seen a single f# job

[ripe meteor]

Are you looking for one?

[hidden musk]

no the language looks sad

[ripe meteor]

😔

I think it looks great

[plush grove]

im gonna be honest, trying to do the minimalist approach (doing everything in as few lines as possible) is the hardest thing i've ever done

[hidden musk]

ocaml still on top for me ngl

haskell close second

of the languages we've done so far ocaml is prob my favorite

prolog takes the cake for "interesting"ness

it's probably the most unique of the bunch

[low condor]

Golfing is a difficult pastime

[magic oxide]

fewer lines isn't really golfing tho
one liners should always be possible here

even excluding uses of exec and stuff

[compact wasp]

for sure

[plush grove]

I kinda just crashed my computer doing advent 😭

All of my memory

[magic oxide]

that's why you don't brute force

[untold briar]

I only brute force with small input like day 6

My laptop only has 8GB RAM so I have to be extra careful when brute forcing

[plush grove]

Oh no I’m doing day 5 part 2

[hidden musk]

brute forced day 5 pt 2 in 5.724 seconds 😩

[wet yoke]

Noice

[covert arrow]

this is gonna have to go now

[rancid creek]

what will it be today, path finding? mini compiler? chinese remainder theorem or other modular arithmetic? game of life in 12 dimensions with sea monsters?

[minor cave]

Language Roulette: Day 7
The language is ... Visual Basic

[rancid creek]

reminds me I shold install networkx and learn how to use it again

[earnest forge]

GL everyone

[rancid creek]

hey no esoteric languages!

[wet yoke]

🫡

[earnest forge]

COBOL tomorrow 🙃

[rancid creek]

That would rock.

[hexed axle]

That was the first programming language I ever used, VB is ... Factually speaking a language

[rancid creek]

Not as much as Rockstar being a chosen language, but still

[wet yoke]

That one elder member of the server "YES"

Is Excel on the roulette?

[magic oxide]

good luck to everyone

[covert arrow]

GLHF

[rancid creek]

The number of reddit mods here outnumbering the number of cobol devs here.

[magic oxide]

won't be able to aim for leaderboard again sadly since outside rn

[zenith elbow]

vimscript would be nice

[royal gull]

I’m tired tho, Gluck I’ll hit it in morning

[covert arrow]

I have a flight in 2 hours

[rancid creek]

emacs lisp!

[earnest forge]

Smh imagine touching grass /j

[magic oxide]

[covert arrow]

at the airport rn

[rancid creek]

You should at least look.

[magic scaffold]

glhf

[shut trellis]

<@&518565788744024082> Good morning! Day 7 is ready to be attempted. View it online now at https://adventofcode.com/2023/day/7. Good luck!

[royal gull]

I tried that before and went to bed 2 hours later 😉

[magic oxide]

ooh nice

wonder what p2 is

[woven solar]

...tbh this one just looks painful and boring.

[magic oxide]

it looks easy

or at least p1

which is what I need

[hexed axle]

There's dozens of us!

[magic oxide]

because I do not have the time to struggle with aoc

[hidden musk]

bruh

naptime

[magic oxide]

good night pub

[hidden musk]

@ripe meteor this one's for you

[magic oxide]

have a good sleb

[hollow wharf]

this make brain hurt

😢

[sinful pewter]

Looks like Poker

[hollow wharf]

i'd rather play it in real life with a person on top of a camel

[pine tiger]

dopamine.exe

[earnest forge]

14/16

[earnest forge]

Wow GG

I should have just pushed it for p2 I was thinking 'surely this isn't fast enough'

[pine tiger]

?

[earnest forge]

(spoiler: it was and I could have sent it around 20s earlier)

[pine tiger]

was there any clever optimization?

[earnest forge]

I don't think so but it was basically instant anyway

[pine tiger]

lmao see here's the thing

[earnest forge]

I should have just pushed it immediately

[pine tiger]

||like i had a debug print printing out every comparator||

[earnest forge]

Ah

[pine tiger]

||and even with that my sol ran in time||

[earnest forge]

Yeah that will slow you down

[pine tiger]

anyways, i just ||brute forced all jokers, wbu||

[earnest forge]

Wait why did you need to debug your ||comparator|| lol

[earnest forge]

Yeah, ||combinations_with_replacement|| mvp

I should have just shipped it because ||n is at most 5 so it really isn't that bad||

[pine tiger]

i mean, ||it's always optimal to have all J's be the same value||

[earnest forge]

That's completely fair

[pine tiger]

||i was replacing one joker at a time 💀 ||

[earnest forge]

You literally just acknowledged why that isn't better lmao

[pine tiger]

well i supposed it's time to start cleaning up my code

[pine tiger]

i act up under stress idk man

[minor cave]

#1182188764278304809

[magic oxide]

could someone tell me what p2 is?

[jovial jay]

Day 7 look fun

[magic oxide]

just anytime after leaderboard fills up

[pine tiger]

||wild cards||

[magic oxide]

||how does that work?||

[pine tiger]

||they can replace any card for a better -of-a-kind|| i spose

[earnest forge]

I told him in solutions btw

[covert arrow]

i choked again

[pine tiger]

most unfortunate

[covert arrow]

I completely missed this line: ||If two hands have the same type, a second ordering rule takes effect. Start by comparing the first card in each hand.||

||I was comparing the repeated cards instead (so 25557 > 33327)||

[earnest forge]

Lol rip

[covert arrow]

advent of reading comprehension is back 😔

[pine tiger]

💀

[zenith elbow]

welp that was slightly tricker than I thought ||JJJJJ||

[humble copper]

not using ||regex|| is probably what messed me up on this

[covert arrow]

what would you need regex for?

[naive stone]

the way i determine the value of a hand is so scuffed lol. Also part 2 is clever, seems like it'll be fun

[humble copper]

idk just glancing over it it seemed like it would be easy to use it to find hand types quicker

[earnest forge]

Having thought about this, yeah regex could do it fairly easily. Still wouldn't be my first choice though

[covert arrow]

I think the ||counter method|| is the most straightforward

[naive stone]

spoiler tag?

[untold briar]

Well day 7 is the middle ground I guess?

Not hard, not easy

For the jokers dont panic, a couple more if-else would be able to handle it

[covert arrow]

Oops mb

[naive stone]

all good!

[humble copper]

yeah that's what i ended up going with, easiest to implement

[weary kelp]

Maybe the next one will be hard...

[untold briar]

https://tenor.com/view/robert-downey-jr-tony-stark-gif-26471287

[long jacinth]

so whats the value of a JJJJJ hand? five of a kind? full house?

[ripe scroll]

yep yep

[long jacinth]

or is it a 'high card' so the weakest hand?

or does it turn into a 'AAAAA' but its still just four of a kind?

[pine ravine]

Do yall find this one hard compared to other days?

[pine ravine]

Five of a kind

[long jacinth]

five of a kind of not a defined type though

[pine ravine]

Of type joker

Joker lives matter

JJJJJ is the weakest of all the 5 of a kinds

[long jacinth]

god, it is

im blind

theres a bug somewhere, but time to work

right answer, yay

can i maybe get someone whos a pro to look over my code to check for places to improve?

[mystic gull]

isnt 33333 weaker?

[long jacinth]

no
"To balance this, J cards are now the weakest individual cards, weaker even than 2."

[woven solar]

five of a kind (the weakest five-of-a-kind possible).

[long jacinth]

yeah, i forgot 5 of a kind was defined at the beginning

considering you dont get 5 of the same cards in a normal deck

but saw JJJJJ in my input and it confused me

[mystic gull]

ah i see

thats what ive been missing 💀💀

[thin blaze]

Fun, nobody on the subreddit used the same method than me.

I found it elegant enough.

[spice hemlock]

Aid

[jade atlas]

I feel sad now because I over engineered the 2nd part.
I thought J should ||be one of the dealt cards, instead of seeing that it is just the most occurring card.||

[thin blaze]

What does it change?

[jade atlas]

I had to find all possible k combinations with replacement of n unique cards

[thin blaze]

I have the feeling I didn't face that kind of difficulty.

My way to solve this was pretty straightforward.

I just counted jokers card and added them to the score of the other cards.

[covert arrow]

Because you don't have to deal with it if you ||correctly assume jokers become the most common non-joker card||

[jade atlas]

yeah that's what I meant

[lime fulcrum]

do you tend to use modules or try solve it without anyting extra?

[thin blaze]

I used collections.Counter but you can do it without.

[lime fulcrum]

yeah it is really powerful in this case

[long jacinth]

so far ive used no imports

although maybe i should start using them, i feel like im not learning anything from this

[thin blaze]

To be fair, Python builtins are already powerful enough.

[halcyon pulsar]

Ever get it when your solution works on the test input but not the whole data

[devout turtle]

Yup. Been there.

That's my main issue with day 3.

[dapper pine]

Today (part 1): When the hand variations are explained, what does this one mean:
High card, where all cards' labels are distinct: 23456
Does it mean all labels should be different? Or should it be in some series, like 23456 is? Thanks

[ripe meteor]

It means all labels are different

A2678 is also a high card hand

[dapper pine]

Thanks :)

[halcyon pulsar]

Could anyone run their solution on my test input and tell me how many hands get sorted into the 7 different types?

1 93 100 173 171 261 201 is my types supposedly where the 1 is 5 of a kind and it goes down respectively

[normal umbra]

Anyone know of a good discord bot for a private leaderboard?

[slate berry]

ok part 2 done too nice

but not very nice implementation

[gloomy kernel]

missed day 3

ahh

[somber trench]

I've given up on this all together, dear lord it's hard

[lime fulcrum]

which day?

[somber trench]

I believe I stopped day 2 part 2 or maybe day 3 idr

[lime fulcrum]

yeah day 3 is fiddly. took way too long but I had fun. have you considered #aoc-solution-hints ?

[somber trench]

It's not that it's just that the way it's explained is confusing and I have no idea where I would even start

[lime fulcrum]

can I give a slight hint?

[somber trench]

Sure

[lime fulcrum]

1. ||put your input in a nested list so that each character has its own coordinates||

2.||get a list with only the numbers||

[somber trench]

This is day 3?

[lime fulcrum]

..

crap

[somber trench]

Nw I'm still on day 2 part 1 lol

[lime fulcrum]

no its right

[somber trench]

Just checked

[lime fulcrum]

ah.

that explains it

[somber trench]

That's for day 2 right?

[lime fulcrum]

no that was 3

thought you were on 3

[somber trench]

Ah

[lime fulcrum]

did you process your input already?

[somber trench]

Like get my input data?

[lime fulcrum]

yes and put it in a list where you have your values seperated

[somber trench]

That should be easy enough, I've done it before

[slate berry]

😓

[kind rapids]

128 pixels meme

[hidden musk]

can I do vb.net

🥺

[hollow wharf]

wjhar 💀

[slate berry]

yeah?

[minor cave]

Yeah that's fine lol

[hidden musk]

I have 6 months of experience with vb.net 😩

[slate berry]

where is the code golfing section for advent?

[earnest forge]

https://ptb.discord.com/channels/267624335836053506/1180008645384216606

[cyan sleet]

isn't vb just the new name for vb.net?

Visual Basic (VB), originally called Visual Basic .NET (VB.NET)

[hidden musk]

they have different runtimes

oh they call them the same thing now?

it feels like cheating because .net is so based though

[ripe meteor]

Do it in excel or else

[north atlas]

👀

Or else what

[slate berry]

wait how?

in excel...

[rapid wolf]

excel is surprisingly good at AOC

[hidden musk]

indeed

[ripe meteor]

VBA

[cyan sleet]

that's a different thing 🥴

[ripe meteor]

basically the same thing

[hidden musk]

nah

[untold briar]

Well excel can be used to solve certain problems

For more complicated ones, its faster and easier to just write code, unless you are really adept in excel

[slate berry]

and what's that

[royal gull]

Did anyone post an Excel solution for day 7? Or even VB?

[hidden musk]

I'm cooking a VB.net one

[sharp crown]

Im doing day6 part one. Id like to verify my understanding of the task so far. I am supposed to move up all cards of the same rank with the secondary rule. This would mean that every rank execept for the last rank has only one member?

[royal gull]

Card? You mean day 7?

[lime fulcrum]

6 was the races

[simple raven]

the secondary rule applies within-rank, not between ranks

[simple raven]

if I have '4444T' and '4444K', they are both in the rank with 4 of the same number, but then I need to sort them based on highest unique card within that rank

[sharp crown]

Yes sorry, had a long day 😅

[royal gull]

Maybe open a thread over in the hints channel?

[sharp crown]

maybe later

[jovial jay]

just realized one of my earlier code might be useful here. does the order matter?

[half tree]

Is it just me or do the AoC problems this year come with a lot less useful test data? I keep having situations where my code works for the test data but not the real data and I think it's happening more than in previous years

It's hard to tell if this is an intentional change or not, but I'm finding it a bit more frustrating this way

[rancid creek]

I found that happened a lot in previous years as well. This year seems ok so far. You could always add more test cases to your example input I suppose.

[noble skiff]

I think it's pretty much the same as always too, fwiw

the test data is good for basic sanity checking, but every year there's generally at least one day, often several, where at some point I'm failing on the input but succeeding on the test

some edge cases just don't show up in the test, it's always been like that

[rapid wolf]

I do like that quite a few problems can just be checked by hand, its just compute a thing for each row.

[noble skiff]

One thing i will say is that for a lot of days, he really walks through how the test data translates into the solution, step by step

for example, he might give the value for every row for a row based problems. Or for day 5, he explained every single step of the mapping

So when you write unit tests based on the test input, you don't have to write a single unit test that just does the same thing as running the whole problem

[raw quest]

im still on the first day is it too late for me 💀

[noble skiff]

you can have much more fine grained unit tests. that helped me a lot when I was trying to get the optimized solution for day 5.

[rapid wolf]

there is no time limit

[noble skiff]

also you don't need to do previous days

[raw quest]

i was put up to this by some sysadmin friends the first challenge looks easy enough. is this something you guys put on your resume?

[noble skiff]

you can just jump to day 7

[raw quest]

it doesnt get harder?

[noble skiff]

though do keep in mind they get more challenging

[raw quest]

ah

[simple raven]

nah a friend started today as well

[simple raven]

and skip day5? 😄

[noble skiff]

best part!

[onyx flicker]

you can add anything. though idk if resume would be best. you can definitely put your solutions on a repo and put the link to your github or something on your resume

[raw quest]

ah. How are we faring? no spoilers i wanna do this on my own

[simple raven]

bad, good, bad, meh, terrible, great, good

[raw quest]

lol

ok cya guys around

[slate berry]

6 was great?

why such randomised difficulties

[cyan sleet]

6 was boring

[slate berry]

true

[simple raven]

I was thinking in terms of difficulty for me

[slate berry]

I didn't write even proper code for 6

[simple raven]

too bad you could ||bruteforce the thing in under a second||

[rapid wolf]

APL can invert functions, so it was truly trivial.

[slate berry]

APL?

[rapid wolf]

a weird language I am learning for this AOC

[naive stone]

alien programming language

[simple raven]

APL stands for A Programming Language 😄

[slate berry]

If I wanted to make my own language to get my own syntax what would it take?

like what do I need to learn

[naive stone]

the first sentence on its wiki page is practically "APL (a programming language) is a programming language...", i love it

[slate berry]

I could use hell lot of macros in c++ ig but maybe could get a lang. of my own lol

[rapid wolf]

https://craftinginterpreters.com/ is the best resource I know of

[hazy ermine]

i coded the entire solution assuming go's regex package supports backrefering.... it doesn't

this hurts

[lost needle]

I just signed for the first time(so I'm new to this) into the advent of code and completed part one of day 1.
But I just don't know how to do the part 2 in python.
I would really appreciate any help.
Thank you!

[royal gull]

Ask over in #aoc-solution-hints , no spoilers here 🙂

[dense osprey]

todays problem is nice

isnt part 1 just writing a custom comparator

[mossy basin]

you can make a functools-type compare function, or a function that just returns something that sorts like the cards

[bright haven]

"just" a custom comparator: there are details to get right inside that comparator

[ripe meteor]

These edge cases are really kicking my butt

I should look at my input more closely

[torn sapphire]

edge cases in part 1 or part 2?

[ripe meteor]

Part 2

[torn sapphire]

aah, ic

have fun with that :P
how did you solve part 1? i think certain solutions to part 1 end up with much easier/more difficult solutions to part 2

[ripe meteor]

Implemented a comparator and a bunch of str.replaces to make sorting easier

[simple raven]

this might cause you trouble in part2

[dense osprey]

oh i shouldnt haved used the word just

[ripe meteor]

My two solutions are pretty similar, i just hadnt considered ||JJJJJ|| and ||J dominant|| inputs

[dense osprey]

but what i meant was, that the solution is a custom sort function

the difficult part would probably be checking what type of hand you have

[ripe meteor]

I get that part of the puzzle is to figure out other edge cases but i hate it all the same 💀

[dense osprey]

part 2 is edge case handling

spoiler :||or you could just insert every possible card for j and choose the strongest||

[woven sable]

that's what i did

[dense osprey]

another spoiler: ||but if every of the 1k hands had 5 js, then youd have 14^5 * 10^3||

[woven sable]

ran instantly

[woven sable]

would spoiler this btw

[woven sable]

what i did was j ||if there's more than one j, j assume all the J's will be the same||

[royal gull]

One thing you should check is: || whether there are any duplicate hands ||