advent-of-code

[noble skiff]

there's a pretty clear difference between today's problem, which many people have commented on how long it took to simply read

vs e.g. the folding question, which was much quicker to read, and still led to needing an insightful solution (for the second part)

[sleek cave]

I couldn't agree more. in fact, I like that it has complex instructions and prompts that actually require carefully analyzing the questions, so that it feels like an actual working environment or something like that. plus, reading is a part of the fun, as demonstrated by the lovely creator of http://adventofrealizingicantread.com

[quaint hare]

.aoc unsubscribe

[hidden copper]

Okay! You have been unsubscribed from notifications about new Advent of Code tasks.

[rough siren]

.aoc lb

[hidden copper]

:x: Please use #aoc-bot-commands for aoc commands instead.

[honest nacelle]

FINALLY finished day 15, that one really kicked by butt

[woven oriole]

.aoc c

[hidden copper]

Day 17 starts <t:1639717200:R>.

[humble copper]

30 seconds

[woven oriole]

gl

[humble copper]

good luck everyone

and godspeed

[hidden copper]

<@&518565788744024082> Good morning! Day 17 is ready to be attempted. View it online now at https://adventofcode.com/2021/day/17. Good luck!

[sinful pawn]

oops

[foggy zealot]

second

[uneven hollow]

gl

[humble copper]

everyone except for me apparently

i have no idea how to begin

[uneven hollow]

physics

[velvet stag]

heck ya

this is the advent i know

this is 2019

this is the day i skip

[tulip harbor]

it's your ||cartesian ||whatnots

[sonic silo]

this is cute

[hazy pendant]

wtf

gravity stuff

[iron reef]

I've just finished reading this and I feel like this can just be solved on paper

[woven oriole]

exactly

vector analysis

but can anyone just confirm the drag

[hazy pendant]

∑

[woven oriole]

changes by 1 toward the value 0

what value

[hazy pendant]

this is a very ∑ moment

[still tangle]

what the converstion

wath's the convesation

hello

[humble copper]

yknow what this was fun but im gonna put this as a "do after a night of sleep" thing

[tulip elbow]

I really wish I knew how to read

[silent knoll]

https://media.discordapp.net/attachments/624014259751485441/860366856903458836/image0-69.gif

[hazy pendant]

lmao

that one computer when solving aoc:

[maiden brook]

looks more like cobra

[ember cargo]

How do I know the velocity?

Ohhh I get it now

(6,4) means it moves 6 right 4 left, then it turns into (5,3) (4,2) and so on

I thought we had to work out m/s or sum shit lmao

[stable heron]

oh no... I- I did...

[tight dune]

i agree, the program anaconda is just as annoying as a real one

[dense ibex]

could someone help me with day 5 part 2 I get the correct answer with the example but not the actual puzzle data

please ping if you can help

[blazing kraken]

Again I make my opinion of "python is slow of aoc"

[tight dune]

for aoc python is fast enough

it depends on ur algorithm

[eternal shale]

Help with this code

#python-discussion i have attached code in this channel

[dense ibex]

@eternal shale #❓｜how-to-get-help

[unborn trout]

Python 2 ??

[tight dune]

and what is that os

[proud cosmos]

Wondering why I can't just go 6,9e9e9e9e9e9e9 for the example in part 1

[muted aurora]

on the way back it would not hit the target area

[proud cosmos]

y not

it'd fly straight down over the target area

[muted aurora]

it's not continuous time

[proud cosmos]

see the graph for 6,3

so why can't I just set the height to the integer limit

[zenith mulch]

[muted aurora]

yeah but it can skip over the entire box in 1 unit of tim

if the velocity is too high

[proud cosmos]

when would it ever skip over it

it'd literally go right until exactly above it and go up billions of units

then slowly drop back down

oh.

you mean cause y always increases

I see

[muted aurora]

this example:

x = 0
y = 0
vy = -100
target area is from (-1,-1) -> (1,-2)

[proud cosmos]

Figuring out 6,9 is max though sounds kinda difficult

oh well

[muted aurora]

you'd just straight from y=0 to y=-100 and never be in the box

[lilac ivy]

.aoc subscribe

[hidden copper]

Okay! You have been subscribed to notifications about new Advent of Code tasks. You can run .adventofcode unsubscribe to disable them again for you.

[lilac ivy]

.aoc join

[surreal sky]

.aoc unsubscribe

[hidden copper]

Okay! You have been unsubscribed from notifications about new Advent of Code tasks.

[surreal sky]

.aoc subscribe

[hidden copper]

Okay! You have been subscribed to notifications about new Advent of Code tasks. You can run .adventofcode unsubscribe to disable them again for you.

[sleek cave]

currently i have no approach other than brute force, which I assume is too inefficient for the input size

the puzzles are getting spicy

[crystal shard]

i used brute force and it worked
just like a minute runtime

[sleek cave]

Oh

#advent-of-code-spoilers-archive please :)

[glad tartan]

oh sorry i thought I was there! 😦

my bad lol

[sleek cave]

it's alright lol, mistakes happen

[dusk cove]

a minute? weird, mine finished in 0.2s

[devout tusk]

the timing can be significantly brought down based on your implementation and language. Last I saw was a 9.75 microsecond time

[crystal shard]

Different build structure

[humble copper]

i have honestly no idea what i can do for this other than ||bruteforcing||, like if i can just ||make an efficient enough function that goes through every x and y -100 through 100 and test that out|| then i think it could work?

[muted aurora]

There's some discussion on #advent-of-code-spoilers-archive if you're interested in a better way

but even with brute forcing ||you can be smart about which values you check by doing some math||

[humble copper]

oh yeah absolutely, im gonna consider that too

[low condor]

Day 17 done!

I liked part 1 in particular

[crimson marten]

I did days 1 and 2.

Now I need to finish the rest and catch up!

I feel so behind haha.

[low condor]

good luck! And don't worry about it, the most important thing is learning along the way

[crimson marten]

Haha for sure.

It's just good fun at the end of the day.

[humble copper]

WHAT

this feels cheap

i didnt do this right wtf

[low condor]

let me guess ||you used brute force 💪 ||

[humble copper]

||and it worked??????||

[low condor]

I did so and fully expected it to

[crimson marten]

Haha.

Now to re-do 😉

Ohh wrong chat.

[jade atlas]

.aoc unsubscribe

[hidden copper]

Okay! You have been unsubscribed from notifications about new Advent of Code tasks.

[noble skiff]

i kind of feel like today's puzzle can be done mostly with pencil and paper... maybe I'm wrong though

[low condor]

||part 1 definitely can||

[noble skiff]

actually yeah that prob doesn't work

[fossil cipher]

||so could part 2 if you're really patient. And have a lot of paper :D||

[noble skiff]

i actually just did part 1 in my head lol... no pencil or paper

first time I've ever seen an AoC like that

[low condor]

I needed paper to twig the concept

[sonic silo]

||for bruteforcing it just took me 30 secs per run lol||

[low condor]

with code?

[sonic silo]

ye

||mind you, per run PER part||

||3:30 mins, 3 mins to add the change and run it twice||

[low condor]

||For Day 17 that is super slow to take 30 secs||

[woven oriole]

nice dedication

towards cpu

:D

[sonic silo]

||my CPU is slow and i used a range of 0 to 1000 for dx and -500 to 500 in dy||

[woven oriole]

||all you can do is minimize the randomized bruteforce range||

||to some mere two digit number iirc someone did that in subreddit||

[sonic silo]

||i know||

[woven oriole]

||hm||

[sonic silo]

||there's closed solutions to this||

||'nother thing is whether i wanna bother lol||

[woven oriole]

||lol||

[low condor]

this is quickly become Advent of REDACTED again 🙂
||I would say it's worth knowing how to do part 1 properly||

[woven oriole]

||yes||

||definitely||

[low condor]

||Part 2 on the other hand - yeah, I did a brute force just with minimum ranges of x and y calculated beforehand||

[woven oriole]

||\👌||

[humble escarp]

😄

[sinful pawn]

are we speaking morse code in here or what lol

[wooden jewel]

||-. --- / .-- . .----. .-. . / -. --- -||

[strange zinc]

Any helpful advice for day 16? Mind is pretty boggled at this moment

[low condor]

I'll ping you in the spoilers channel

[strange zinc]

Thanks

[humble escarp]

Done for today 🤘

[wooden jewel]

nice

[mossy basin]

Trick shot visualization!

[hidden musk]

:O

[devout tusk]

holy moly 😮

[woven sable]

Woah

[placid lake]

niiice

[hollow wharf]

Might finally have the energy to try AoC again. Had a final yesterday and wound up sleeping for about 14 hours 🤣

I guess I was tired

[sleek cave]

absolutely sick

okay, gonna clean up yesterday's solution and then it'll be time for day 17

wonder what awaits me today

[mossy basin]

i wonder how many days i can string together to make a submarine game

gonna bonk lantern fish with drones

[sleek cave]

that will be insanely cool

[mossy basin]

that would probably be too crowded on a terminal, but this has been a great way to debug my library

[sleek cave]

||okay perhaps i could use quadratic functions to solve d17? (or maybe just brute force lmao)||

[woven sable]

||i just bruteforced, it took 5-10 seconds or so||

[bright haven]

||I didn't use quadratics. I used brute force, but with well-bounded limits. it took about .5sec||

[woven oriole]

||call it grid search not brute force 😎 ||

[velvet stag]

seben

[woven oriole]

sex

.aoc c

[hidden copper]

Day 18 starts <t:1639803600:R>.

[woven oriole]

not once again advent of algorithms taught in uni

🥴

[woven sable]

I wonder if the ||pathfinding|| day was just a fake peak, and this weekend will be the real challenge

[velvet stag]

i bet it was real peak

[hidden musk]

this weekend is traditionally the hardest 👀

[prisma sundial]

gl

[woven oriole]

what about sunday smh

[hidden copper]

<@&518565788744024082> Good morning! Day 18 is ready to be attempted. View it online now at https://adventofcode.com/2021/day/18. Good luck!

[sinful pawn]

oops

[crystal shard]

No!

[pallid shoal]

[signal grotto]

Woo weird math

[velvet stag]

i did thing lyk this once took me maybe 6 hours

[humble copper]

the burnt-out urge to ||parse this as a string and not as a list|| because ||im too lazy to add depth checking and i know how that ended up last time||

havent finished it yet and i dont plan on solving it today but thats how id do it

also has anyone bothered to like graph the length of the questions as the days progress lmao

[devout tusk]

too much artificial difficulty on this one. I can't be asked to parse this shit rn

[humble copper]

amen

if it's late and the problem is too difficult i just start it the morning after

[woven sable]

I thought of a nicer way to do this than ||nested lists|| if anyone wants to know

At least in my eyes i would prefer it over the depth of a nested list

[stable heron]

i just woke up... 😭

i dont feel like doing it rn

[humble copper]

ouch

[weak torrent]

me not understanding anything in this question 😔

[muted aurora]

I keep getting this for the bigger example, but all the smaller ones seem to work on their own. Anyone got any ideas?

  [[[0, [4, 5]], [0, 0]], [[[4, 5], [2, 6]], [9, 5]]]
+ [7, [[[3, 7], [4, 3]], [[6, 3], [8, 8]]]]
= [[[[7, 8], [5, 0]], [[0, 7], [6, 0]]], [[[8, 0], [9, 9]], [[0, 7], [0, 8]]]]

[velvet stag]

i think it can be anything

a bug in split

a bug somewhere

[pallid quail]

Hey folks - I wonder if someone can help me talk something through - I'm new at programming in general, and I'm wondering if what I am thinking will work.

I'm on the second challenge of day 1.

And to get the sums, in the sliding window, I am wondering if I can somehow write a recursive function that would take in the length of the list, then sum the first 3 index positions, and then subtract 3 from the length of the list, and call itself with the new length, repeating until the length reaches less than 3

I know it might be overcomplicating things - but I'm learning and hard is a good way to learn, no?

[uneven hollow]

if you subtract the first 3 index positions, and then subtract 3 from the length and call it again, it will once again do the same thing ( subtract the first three index positions), so you wont be getting any progress

[pallid quail]

I forgot to mention, I was going to pop the first index position out each run

[uneven hollow]

yeah then u dont really need the length of the list

just recursive call it with the new list

[pallid quail]

How do I get it to stop when there are less than 3 items left in the list

Oh wait

nm

[uneven hollow]

if len(list) == 3:

dont pop

[pallid quail]

Yeah

just thought of that

[uneven hollow]

👍

[pallid quail]

Sometimes just writing my questions into discord, helps me figure things out haha - I should just keep sublime open, and type out my questions in there, noodle them a bit, then come on here and ask after.

[uneven hollow]

good idea

[pallid quail]

Holy crap, it worked.

My first recursive function 🙂

https://tenor.com/view/damn-that-feels-good-kayce-dutton-luke-grimes-yellowstone-it-feels-good-gif-23736985

[hollow wharf]

idk if it still helps but ||make certain all the explodes happen before all the splits, on both left and right of pairs. This was my problem for like 45min because splits on the left were happening before explodes on the right.||

[broken pasture]

hey guys

i am doing advent of code Day 16 Part 1 currently

[regal gazelle]

https://paste.pythondiscord.com/ozorigawem.lua
why does the .py find the nums.txt file but the .exe doesnt find it
( pls ping me )

I always have problems with .exe

[broken pasture]

and currently my code correctly parses all the examples and gives the right answer but gives the wrong answer on my real input

[velvet stag]

are you padding every letter to 4 bits?

make sure you;re not storing 6 as "110" for example

i guess that would show up in examples

[broken pasture]

yeah

[velvet stag]

there's not really a big difference

in the real input

[broken pasture]

i will paste my code here

[velvet stag]

ok

no wait

that's a spoiler

[broken pasture]

should i dm you the code then

[velvet stag]

sure go ahead

[broken pasture]

hey guys can anyone help me with advent of code day 16 part 1

[rancid creek]

poste it using !paste in #advent-of-code-spoilers-archive

[rancid creek]

have you asked for advice in the spoilers channel? It's the place for help.

[broken pasture]

nevermind figured it out @rancid creek

[muted aurora]

Thanks! I did figure out the issue after debugging for an hour, turns out I had ||copied a self.left and forgot to change it to self.right||

[restive imp]

Some instructions for operator packets on Day 16 aren't clear to me:

• the example with length type ID 0: Why is part A 11 bits long and part B 16 bits. I know the total is 27, but how do they split?
• the example with length type ID 1: since it doesn't specify, is every sub-packet always 11 bits long?

[abstract bluff]

the packets have a recursive nature to them, you don't know how long each subpacket is until you read it.

[restive imp]

ohhh... so every sub-packet can be a literal or operator ... so the sub-packets would start with version and ID again ?

[strange zinc]

We have another essay today 😆

[abstract bluff]

yes

[proud cosmos]

me: why is this recursing infinintely what

also me: giving it itself as argument

[hollow wharf]

so for yesterday

it seems like there is ||a quite simple mathematical|| solution

is that worth exploring for part two lol

[maiden brook]

worth I'd say, it can get quite efficient

[bright spindle]

||From what I understand there's a way to reduce the bounds for part 2, but you can't just solve it outright like with part 1||

[maiden brook]

you can look into ways that ||don't require simulating the projectile, but still require a loop or two||

[sleek cave]

evil puzzle

why nested data structs

[woven oriole]

||parse string > making tree hehehe||

[proud cosmos]

Honestly might've been easier

[sleek cave]

okay, my to-do list for today:
d17 appetizer
d18 main course (one of the hardest problems of the entire aoc)
d15 dessert

[proud cosmos]

gl on that

day 18 seems hard to me but idk

maybe I'm just slow

I'm stuck on something

whoknows what

Something is definitely wrong though

[sleek cave]

oh no it's not just you, usually d18 and d19 are when the hardest puzzles are released

[hollow wharf]

is there like a leaderboard for AoC, so i see which implementation is the fastest?

[proud cosmos]

I'm like actually convinced one part of this day 18 part 1 is 10000000x easier if I just do it as a string instead of my whatever

[maiden brook]

not that I know of, but people are talking about bigger inputs and efficient solutions in this reddit thread https://www.reddit.com/r/adventofcode/comments/r9s5pz/2021_big_inputs_for_advent_of_code_2021_puzzles/

tag @muted aurora since I asked about this a few days ago

[hollow wharf]

meh

idc about bigger inputs

i was curious about optimized solutions

performance

[maiden brook]

well, to process bigger inputs in a reasonable amount of time you need faster solutions

[hollow wharf]

doesn't matter

you can still do them with slow implementation

and with python 😂

[bright spindle]

It's easier to measure the performance of a solution with larger inputs.

[hollow wharf]

sure, but implementation may need to be changed if you're working with non dynamic structures

which very much may be the case if you're optimizing

[bright spindle]

My naive solution solves part 2 in ~150 ms, if I change something and run it again, it's hard to know if any measured difference comes from the change I made or any confounding system factors.

[hollow wharf]

150ms is huge

[woven solar]

huh, that's really fast, is that in Python?

[bright spindle]

No, Rust.

[woven solar]

ah, that explains it

[hollow wharf]

C++ solutions for 2020 were in microseconds https://github.com/Voltara/advent2020-fast

[bright spindle]

Those were different problems.

And like I said, this is my naive solution.

The point is that it's easier to distinguish performance improvements with longer runtimes.

[hollow wharf]

again, yes, but it also may change implementation

which takes away from original challenge

at that point you're implementing your own solution for your own problem

[bright spindle]

What type of properties of the input do you have in mind that would significantly change the type of implementation you would choose?

[rapid wolf]

an example is the lanternfish day, where the matrix solution was slower for 80/256 days, but for 1mm+ days it was vastly faster

[bright spindle]

Well, fair enough.

[hollow wharf]

this for example, this can easily fit in a static array of 10 by 10 char

[bright spindle]

Still fun, though.

[hollow wharf]

(C char, otherwise uint8)

[bright spindle]

Right

[hollow wharf]

day 11

[bright spindle]

I see what you're saying

[hollow wharf]

so no need to bring in list into the story

and keep allocating

(which'd be dynamic)

[low condor]

Question - for day 18 - if there are multiple pairs to explode, but exploding causes a pair to split - do I ensure I split all the pairs before doing anything else?

coding explode and split isn't so bad but order of operations is killing me here

[bright spindle]

||You always apply all possible explodes before you try a split||

||Then after you've done a split, you again try as many explodes as possible before trying another split||

[low condor]

right - that's what I thought - yet for some reason can't get the order right D:

[dark slate]

||it's like written on the site:

1. if can explode then explode
   else if can split then split
2. goto 1)||

[bright spindle]

Please spoiler this.

Thanks.

[low condor]

is that a spoiler really? clarification of requirements doesn't seem spoilery to me

[dark slate]

ok but it's written on the site as well

[low condor]

that's why I'm asking here, not in the spoilers channel

[bright spindle]

Interpreting the problem description is part of the challenge.

[low condor]

I'd argue otherwise, but nvm

[fossil cipher]

Might be my brain not in the mood for this today, but today's puzzle description seems really bad. Can't make sense of it and the examples don't illuminate at all

[proud cosmos]

I think I am stuck completely

every single test works

part 1 works

[bright spindle]

You are free to not consider it a spoiler if you wish, but other people might, so we are erring on the side of allowing the maximum number of people to enjoy the puzzle the way they want to.

[proud cosmos]

but my partt 2 is wrong

the test for part 2 works though

No idea what I am supposed to do

Besides personally checking every single line added to eachother I don't really see what to do

and i think i'd rather cheese the solution than doing that

[woven solar]

my problem was mutability

[bright spindle]

||You could try a bottom-up solution and construct your own test cases (for all subroutines), starting from the simplest possible ones and gradually adding more complex features. Eventually you might encounter a test case that fails. Something not covered by the ones in the problem description.||

[woven solar]

my part1 changed the data used for part2. make sure that doesn't happen.

[proud cosmos]

I forgot aoc is not that dumb and I can't get any more info

lol

[bright spindle]

||Alternatively, you might spam a bunch of asserts all over your code and hope one of them breaks||

[proud cosmos]

Yeah that sounds like a pain

Haaa

I mean I guess I know which lines produce magnitudes too large

I should probably add those together and see what should happen and what is happening

[bright spindle]

||There aren't really that many different possible configurations for the input to be in (that are categorically different, I mean), so you could just construct test cases for those and see if you get what you expect||

[hollow wharf]

yesterday's puzzle makes me uncomfortable

I have no way of knowing whenever I truly have the best possible case or not

or do I

[proud cosmos]

> vs >=

my add method was bonked

||basically if it added a 3 deep and a 4 deep together it thought the result is 4 long not 5, so it didn't reduce it||

[bright spindle]

ic

[hollow wharf]

I hate day 17

[bright spindle]

||For which part?||

[hollow wharf]

part 1

I think I have an idea actually

[bright spindle]

||I didn't find the O(1) solution myself either||

[hollow wharf]

I don't even get the example

[humble copper]

is it guaranteed that in any list there will never be a pair that is nested in more than 4 pairs?

[hollow wharf]

oh

[maiden brook]

||you can check that for yourself||

[hollow wharf]

ungh

I'm not sure how I am supposed to find the upper bound for the iterative method

is there even one?

[proud cosmos]

yes

pretty simple one i think

[bright spindle]

||Yes, because an explode can only reduce the depth, and a split can only increase the depth by one, and after a split you will always perform any possible explode before you do another split||

[bright spindle]

||I just guessed the bounds and it worked||

[mossy stream]

lol I think he got tired of this challenge

[pallid stirrup]

well just woke up so hope today's isnt too hard

[proud cosmos]

Depends

If you do it like me probably yes

If you do it normally probably fine

It's quite managable tbh but choosing how to do it might bring pain

:(

[pallid stirrup]

i did -200 to 200 and it was overkill but worked lol

[proud cosmos]

Also 20 pages

[pallid stirrup]

uh oh is this another one like the fish where it's gonna be slow

[bright spindle]

||I did -1000 to 1000, lol||

[pallid stirrup]

lol

[proud cosmos]

You can do it

[finite cove]

for day17, ||I'm trying to get the x velocity towards 0
I was thinking about something like x_velocity -= towards_zero(x_velocity)
where towards_zero looks something like this
def towards_zero(nr: int) -> int: if nr < 0: return -1 elif nr == 0: return 0 else: return 1||
but I feel like there's something built-in for this already, and if not, that this at least has some sort of "official name" in comp science
anyone got an idea?

[bright spindle]

I don't think today's is super difficult, but it's a lot of work.

[bright spindle]

Please either spoiler this or discuss it in #advent-of-code-spoilers-archive .

[bright haven]

Python 2 had a builtin called cmp() that did that. Now you can define it with: def cmp(a, b): return (a > b) - (a < b)

[finite cove]

oh, my bad

[mossy stream]

it is the mathematical -sign function, so basically -abs(x) / x, I don't know if Python has an implementation for this

[bright spindle]

No worries.

[bright haven]

oops, not what you asked 🙂

[proud cosmos]

your definition of the sign function is not defined for x=0, theirs is

[maiden brook]

there's ||from numpy import sign||

[mossy stream]

:P

[proud cosmos]

There does exist math.copysign though

Which is not equivalent but exists

[mossy stream]

||x and -abs(x) / x||

||-copysign(1, x)|| should do the trick I believe but that's ugly

[proud cosmos]

[finite cove]

yeah, I also considered copysign, but that failed for 0 😅

[mossy stream]

uuhh

oh, and

[proud cosmos]

also if would prob be more efficient anyway

it looks maybe large but i'd imagine the compiler is quite happy with it

[finite cove]

ah, that's exactly what I was looking for... why'd they remove it from the stdlib?

[mossy stream]

because now we have fancier comparison dunders

#advent-of-code-spoilers-archive please

[pallid stirrup]

oops throught i was over there

[mossy stream]

haha, fair

[finite cove]

aahh, I didn't realize what the purpose of cmp actually was, that makes sense

[humble copper]

so in trying to do the explode part ||i accidentally made the function handle every instance of explosion at once (which also works) when my original idea was to do them one at a time, at least that saved me a ton of time trying to set up a loop||

[woven solar]

I'm not too sure that works
though, if it's ||only explosions||... huh.

[humble copper]

yeah nevermind it doesnt lmao

but i fixed it

[humble copper]

apparently i didnt

[sleek cave]

lmao, happened to me as well on day 12

[sleek cave]

woohoo, d17 is a great warmup to a day of intense programming

[low condor]

have fun with day 18 🙂

[sleek cave]

99 percent of the bugs on d17 were + 1, - 1, y1 vs y2 and < vs > lmao

[woven sable]

For fishes they sure do have a complicated addition system

[sleek cave]

i sure will

[low condor]

are you starting day 18 today?

[woven solar]

The only reason they haven't conquered the planet yet is because they have no need to. Their schoolchildren are as our adults, and their adults are as our Einsteins.
When the snailfish get bored with us, we won't stand a chance.

[woven sable]

We're lucky they're far down in the ocean trench for now

[low condor]

I'm more worried for the lanternfish

imagine if they could come on land - they'd overwhelm us all

[sleek cave]

yep!

[woven sable]

Hopefully the elves' packet transmission keeps them busy

sending useless messages

[humble copper]

achievement on its own for me

wow that was easy ok

[hollow wharf]

Today's?

[low condor]

I suspect they mean ||part 2 for today, having done part 1 already||

[sleek cave]

||i see nested, i smell recursion||

at this point i might as well do every single aoc problem with ||recursion|| lmao

[signal grotto]

Hey if anyone here has solved 2018 day 10 (https://adventofcode.com/2018/day/10), can you DM me your output? Looking for copies of the various letters, like these:

[low condor]

why is it that you need the output?

[proud cosmos]

Didn't someone here make a advent of code letter reader thing?

Probably has all of them in it

[signal grotto]

Adding support for those taller letters to my library https://github.com/bsoyka/advent-of-code-ocr, they just have to be different from every other day lol

[proud cosmos]

Forgot who it was

Oh that was you

I see

[signal grotto]

(Meee (probably))

[low condor]

fair enough, one sec

[proud cosmos]

Sadly haven't done that

[signal grotto]

Every other day that has letters that I know of use 6-pixel-tall characters, but this one is just different 😂

[sleek cave]

||i see checking magnitude recursively, there is no doubt: i can taste the recursion||

[proud cosmos]

Tastes lemony

[low condor]

I'll send my output on spoilers channel

[signal grotto]

As an update, if you have any letters not in this list can I have them lol ABCEFGHJKLNPRXZ

[bright haven]

I have PANLPAPR

[signal grotto]

I think I have those ones already from the internet, will use yours to double check though thanks!

[sleek cave]

i just had a moment of realization

||day 18 is a binary tree, the nested levels are depths, dfs is used to traverse the tree and bfs is used to determine "first number to the left/right"||

oh wait

[woven sable]

I'm finally finished!

[sleek cave]

WOOHOO!!!

[sleek cave]

recursion broke my brain. will reattempt the puzzle once my brain recharges

[woven sable]

Pretty massive drop today

even though the day isn't over, there's a notable drop

[sleek cave]

brain recharged. day 18, I'm coming for you

[woven sable]

How'd you recharge? Was it all the energy from the glowing lanternfish

[sleek cave]

mostly a good supper, but the lanternfish helped

[humble copper]

good luck

you got this

[fervent talon]

Question regarding Day 8 Part 2. I am struggling with it. My inner Alan Turing is running wild.

Will alphabetical sorting of segments input/output break logic of given puzzle?

[woven sable]

||Alphabetical sorting will break the logic. Say with the example it gives you that deafcbg works, if you alphabetically sort it deafgbc would also work when it shouldn't because it's the same characaters||

[fervent talon]

Could you clarify a bit more?

[woven sable]

Sure, so ||when you alphabetically sort, different types of combinations might be the same alphabetically but yield different results, because there's an order to the string.||

|| ddddddd
e a
e a
e a
fffffffffffff
c b
c b
c b
gggggggg makes deafcbg||

[fervent talon]

|| I see. Is this order derived from puzzle explanation or from solving it? I mean can you come to this conclusion without knowing the solution or not? ||

[woven sable]

||This order is derived from the puzzle explanation||

[sleek cave]

slight progress, small victories

alright i have a party planned and i'm late so no more breaking my brain today

will attempt 3 tomorrow

[fervent talon]

Could you point me to that place in puzzle explanation. Because I read and re-read and even after your explanation, don't understand where it comes from.

[woven sable]

Sure! One second

[fervent talon]

Thank you

[sleek cave]

they really ain't kidding when they said these two days were going to be hell

[woven sable]

||The segment display above would form abcdefg, I actually just realized it wasn't explicitly mentioned but it was more an inference from how it was set up||

of participants who completed Part 2 of that day after every AOC day from the first day till now

[low condor]

thought you said no more breaking your brain 🙂

[fervent talon]

||
I looked on wikipedia seven-segment display reference and assumed the order is different, e.g. https://commons.wikimedia.org/wiki/File:7_Segment_Display_with_Labeled_Segments.svg#/media/File:7_Segment_Display_with_Labeled_Segments.svg
||

|| Do I assume correctly that either way I will need to trail and error different combinations for all segments to get correct one? ||
Basically what author says:

After some careful analysis, the mapping between signal wires and segments only make sense in the following configuration

[woven sable]

You do assume correctly regarding that

[fervent talon]

Okay

[woven sable]

if you have any more questions feel free to ask by the way (but just a side note: we won't have to put spoilers around everything if we move to #advent-of-code-spoilers-archive )

[fervent talon]

Okay. Got it. Good point.

Have to sleep on it. Will probably ask more tomorrow 🙂

[violet raft]

im so behind on aoc

stuck on day 10 : (

does it just keep getting harder after day 10?

[low condor]

in general problems get harder, but you might find something hard that others find easy and vice versa

[low condor]

don't stress over it - every puzzle you do means you learn something

[low condor]

would you like hints on this?

[violet raft]

yea, thats true im gaining a lot from this just reading other ppl's code and stuff

[violet raft]

that would be great!

[low condor]

OK - will post in spoilers channel. Part 1 or 2?

[violet raft]

part 1

[sacred geyser]

day18 is hard

[sleek cave]

day 18 is so tough

i can't get the explosion working 😔

tomorrow, attempt 4

[abstract bluff]

what is not working?

[noble skiff]

thank god I'm done with this

[sleek cave]

ggs!

I'll be whipping up a call stack tmr, similar to how I did with d12

Hopefully that helps me

if I still can't solve it I might ask for a slight hint

[prisma sundial]

gl

[hidden copper]

<@&518565788744024082> Good morning! Day 19 is ready to be attempted. View it online now at https://adventofcode.com/2021/day/19. Good luck!

[hazy pendant]

i give up

anyways

[sinful pawn]

oops

[foggy zealot]

third

[sonic silo]

oop

[woven sable]

Talk about a long problem

[lost dock]

cool

[pearl needle]

god numpy wizards are gonna love this

[woven sable]

I wish I knew numpy 😔 I'll learn it some day

[sleek cave]

heh, I stayed up so late today the new puzzle released. gotta sleep, bye

[humble copper]

yep im not doing this today

well its been fun

say hi to the people at the top of the leaderboards for me

[pallid shoal]

What's this?

Explain a bit

[modern harness]

Nice job making it onto the global leaderboard on part 2!

[modern harness]

I was pretty close behind you:

[lapis jay]

what language did you run it in? I had to wait a while for my python to run the second part (and had to run it twice due to typo)

so mine likes like 2-3 min to run

[modern harness]

Python 3.10. Mine took 4 seconds

for part 2

[lapis jay]

I think your code is slightly more efficient

[modern harness]

An earlier version had a lot more itertools.product until I realized I slightly misunderstood the problem (couldn't get scanner 2 to match in the example), and ended up getting rid of most of my product stuff

[pallid shoal]

What's this? :(

[pallid shoal]

What's this? :(

[modern harness]

Are you doing https://adventofcode.com ? You can see your personal leaderboard here: https://adventofcode.com/2021/leaderboard/self

[dense herald]

I'm working on an encoder for day 18 (for debugging, not the solution). Can anyone help me convert my encode function from recursive to iterative? https://replit.com/@jmcantrell/RosybrownOtherConnection#main.py

[flat sequoia]

What you can do is parse the input character by character

and push things onto a stack

so for [[1,2],[3,4]]

it'd go

[dense herald]

I'm trying to go the other direction. Turning the tree back into a string.

[flat sequoia]

ooh

sorry i misread

you can do the same thing though pushing things onto a stack

[lapis jay]

look up inorder traversal

[flat sequoia]

but there's really no point since when you recurse that's basically what it's doing anyway

[dense herald]

I figured it would look something like that. I just can't figure out when to add the , and closing ]

[lapis jay]

, when going from left child to right, ] when you go up a level

    def __repr__(self):
        if self.value is not None:
            return str(self.value)

        return f'[{self.left}, {self.right}]'

that was my node class repr

[dense herald]

that's essentially recursive postorder, right?

[lapis jay]

the , is inorder, ] is postorder

[dense herald]

how so?, repr is getting called for the left and right child before the f-string is constructed

[lapis jay]

ah, yeah I suppose so

I was thinkin literally about the charectors

[dense herald]

i've been sort of obsessed lately with coming up with iterative versions of recursive code. i find it irritating that i don't have better intuition about how to do this

[gloomy kernel]

Yye

I won't continue aoc

[modern harness]

Weekends are harder. The weekdays and the 25th are easier

[gloomy kernel]

O

[devout tusk]

What is the question even trying to say? I don't think I understand it

....

[void ocean]

need to catch up soon

haven't done aoc in a week

[muted aurora]

man y'all too good giving me self confidence issues

[digital ember]

time for day 15 :)

[proud cosmos]

I've read half of today

And it reminds me so much of 2020 day 20

This will be pain

[twin fog]

I think this says it all

[proud cosmos]

It is still in the middle of the day

[twin fog]

The completion times also tell the story. It's some problem.

[proud cosmos]

It's hard for sure yea

Still haven't even done part 1

[stable heron]

i havent done like

a lot of

challenges

😭

[primal vapor]

hi

my code works for todays test

but it takes like 13 seconds

to complete this test

did you face performance issues today ?

[steel bear]

no worries I am still on Day 17

🥲

[stable heron]

ha im on day 15

🥲

[frigid zephyr]

I just did day 16

[sleek cave]

new day, new attempt at day18

[muted aurora]

Good luck!

[sleek cave]

thank you!

[muted aurora]

Got some free time today morning if you want to bounce off some ideas lmk, I'm down

[sleek cave]

||my biggest problem rn is explosion and traversing the parents of my binary tree nodes lmao||

[sleek cave]

tysm, I'll be trying to write out some pseudocode first, if I get too stuck I'll ask for a hint :D

[proud cosmos]

JS is litearlly juist ignoring my -0 and turning it into a normal 0

:(

y

[peak dock]

?

[sinful pawn]

I'm like 99% sure that most languages do this

[proud cosmos]

Fair

and my solution kinda needs it but not always

It's a mess for sure

[sinful pawn]

In [1]: -0
Out[1]: 0

[proud cosmos]

I mean the funny thing is it kinda keeps it but not really

[sinful pawn]

oh. gross

nim> -0
0 == type int

[peak dock]

is there any language which doesnt do this lol

[sinful pawn]

Apparently JS to some extent

which is weird but... then again, it's JS

[peak dock]

yeah

js doesnt count

[sinful pawn]

JS belongs on esolang tbh

[peak dock]

!otn a js is an esolang

[marsh current]

:ok_hand: Added js-is-an-esolang to the names list.

[sinful pawn]

:O

[proud cosmos]

Lol

Yeah this day definitely giving me give up vibes my solution idea is kinda garbage

I guess I'll just do something else and try later

[humble copper]

i dont understand a single thing this is saying

[proud cosmos]

lots of random points on 3d plane that may be rotated randomly in any of the 24 possible ways to have a point rotated on a 3d plane

and you need to find points that are at the same place except the reference changes

Pure pain for me

[twin fog]

I can see how you'd do it

Like, || make a list of edge||

[proud cosmos]

I have an idea too it is just really bad or smth

Cause I am not getting closer to anything

Also at some point I noticed my code doesn't even correctly output all 24 rotatons

:(

[hidden musk]

for floats, there's a -0 and a +0

[proud cosmos]

Okay lets go

I think I had an idea as well how it won't matter if -0 == 0

my initial idea was kinda stu pid

and I think I even managed to create all 24 permutations this time

[olive glen]

nim> -0

how do u guys understand this coding stuff like me I don't know how to do all this stuff u guys do I wish I could code to XC

[sleek cave]

it's alright, we all start somewhere :) perhaps try doing some beginner projects first before you attempt aoc, the puzzles do get quite challenging

[glad tartan]

with some rigid body theory my Python solution runs in 60 ms (both parts, including tests)

[wooden jewel]

wow

[noble skiff]

Now I'm curious what you mean by rigid body theory

[sleek cave]

r- rigid body theory-

[drifting saddle]

Nice! What aspects of rigid body theory did you use? Is there something like a Wikipedia page that I can read, or something? 😛

[mossy stream]

[[[[0,7],4],[[7,8],[6,[6,7]]]],[1,1]] when the last 7 explode, does it actually have to increase the 1 on its right?

[cursive temple]

yeah

[mossy stream]

pain

[pliant comet]

my exploding / splitting works for some of the example inputs and times out for the rest of them 😔

[mossy stream]

are the remaining days actually going to be challenging or just painful like that

because honestly I'm not having fun on those last few puzzles haha

like the pathfinding one was really fun

here it is just adding a bunch of corner cases just to spite you

[glad tartan]

just thinking of the 12 (or more) overlapping beacons as a rigid body where all the points are connected by rods. Then provided rods are unique within each scanner you can use those to simplify things a bit 🙂

[pearl moon]

Don't think I understand 19....how can you tell if beacon readings from scanners overlap when coordinates are relative? Couldn't all the scanners be super far away from each other and have no overlap?

[jolly comet]

vbBuSOtTSQU08h/oMt6GJsXJD5pcuveUr/MD3VnEVq4St1WK7nqUMgKP7skkqi/JwUldcB+pwTzAzPyHNAO5PcLugBd/MjPX0pv77EJOvAdjV+e1QZh6iSstO637NW/vKK9/wudHTHvB/jSxV/yO6h7aK1I+oZiMSqbb9BiWpJ4=

can someone decode this

[peak dock]

it guarantees you that if two scanners see a cluster of 12 beacons at the same relative positions, they're the same 12 beacons

[peak dock]

??

[jolly comet]

dk what it is

i need it for the riddle

xd

[proud cosmos]

base64

Most likely

[pearl moon]

Hmmm....so the same positions relative to each beacon? Or the same literal position values

[peak dock]

positions of the 12 beacons relative to each other

like in the 2d example, the 3 beacons form the same triangle

[pearl moon]

Hmmm ohhh

[sleek cave]

moral of the story: ||when recursion breaks brain, use iteration instead||

5.5 hours

that took some time lmao

||i used brute force and permutations on part 2 so it's quite inefficient, don't know if there's a clever way||

[proud cosmos]

Hey don't worry still better than me

I am still not really anywhere close to a solution

:)

It appears random rotations are my enemy

I just lose interest

[pearl moon]

AoC lately
https://c.tenor.com/TMgawVld2pAAAAAM/nasty-indiana-jones.gif

[proud cosmos]

Just please let me solve this today

it's just not working

[sleek cave]

damn, day 18 really was tough for me

[noble skiff]

it was rough

especially for people who stuck with the recursive approach it seems

trees kind of stop being elegant and easy to work with once you start needing backreferences

[proud cosmos]

Okay I almost kinda have it find the first 12

That's something

Woo.

[glad tartan]

It wasn't too bad, just needed to keep the path down the tree

[sleek cave]

cough me cough

[robust heart]

Stuck at day 12

The more I wanna commit, the less i get stuff done

Gonna be in as many dead ends as the cave's paths

[prisma valley]

where can i see the puzzles of AoC?

[woven solar]

https://adventofcode.com/

[noble skiff]

Well, I guess you either keep the path or you use back references

Most people used back references it seems

You don't need both

[humble copper]

||you dont want to see what i did lmao||

[proud cosmos]

Sometimes I truly do hate node js

Error in your code?

Process exited with code 1

fk you too ig

How to debug in node.js: just paste it into firefox instead it'll tell you whats wrong

[snow cargo]

is there any way i could implement if/else logic without actually using it?(may be design pattern)

[woven solar]

How did you like the approaches you got suggested?
also, this is #advent-of-code, not #python-discussion

[humble copper]

day 19 part 1 complete!

god this took forever to do

[proud cosmos]

Yeah I'm still not done

I got close tho

The test almost worked once but it had no reason to

[proud cosmos]

Oh well maybe part 2 is eaasy then

except my part 1 now takes about 1 minute to find scanner 1 being close to scanner 0

This is so bad

And it can't even find 4 for some reason

[humble copper]

might need optimizing

||i know mine does||

[proud cosmos]

It just doens't work

idk

[humble copper]

oh

[proud cosmos]

my code is super broken somehow

It like kinda got something but the coordinates of scanners were wrong

[woven sable]

nicee!!

[proud cosmos]

so I changed it so it klnows where the sacnners are except now all the other stuff seems broken

[humble copper]

do you want to see the spaghetti used to solve it lmao

[proud cosmos]

Idk how this works

How are either my beacon coordinates in the example right

or my scanner coordinates

what is going on

That makes no sense

[woven sable]

sure

let's post in #advent-of-code-spoilers-archive

just so we don't need to spoilers everything

[low condor]

Completed Day 19!

[woven sable]

Nicee!

[low condor]

||you didn't use recursion ?! You must be an imposter||

[proud cosmos]

It finally finds all the scanners at the right places but it finds way too few beacons

[sleek cave]

lmao i still used ||recursion for other parts of the tree, just not the parent traversal for explosion part||

[proud cosmos]

79
O:::

pls work on actual inputtt

thanaksksss

If this doesn't work my spirit will be broken

Especially since I'm estimateing this has to run for like half an hour to get anything

this is so bad

[low condor]

||ditto - I used recursion to make the entire thing flat before doing anything else||

[low condor]

rip

[proud cosmos]

How long does your part 1 run

[woven solar]

print intermediate results so you can see progress - I print ||every pair being considered||, that kind of thing

[proud cosmos]

I mean I have one idea how to optimize it actually

I print ||scanner coordinates and the queue||

[woven solar]

my day19 takes like 30-60s? part2 is instant-ish once you do part1

[proud cosmos]

I see

I hope it is instant for me

[low condor]

||I'm at 0.5 secs for part 1||

[proud cosmos]

otherwise I'll cry

[sleek cave]

from what I've seen, there're three factions: ||the tree purists, the tree flatteners, and the we-dont-talk-about-them cough trig solving it with strings cough||

[proud cosmos]

I mena my code is just so bad but it should work theoretically

[woven solar]

||trees|| are poggers

[proud cosmos]

Please work for the love of god

[low condor]

|| The best solution is flat, just like the earth||

[proud cosmos]

I'll start trying to implement some optimization cause this is gonna take awhile

[sharp panther]

||the earth is not flat smh||

[proud cosmos]

It's a doughnut

[sharp panther]

||if only it was a tree||

[sleek cave]

oh shit- the earth is flat????? I thought it was a slice of pizza this whole time

[proud cosmos]

A doughnut is extremely accurate if you think about it

Red filling, ball shaped, stuff on top

[low condor]

Depends on the donut

[woven solar]

I think the topology is wrong for that lol

[low condor]

I was thinking ring donuts and was confused lol

[proud cosmos]

Only if you think about the wrong donut

[woven solar]

what's the right one

[sleek cave]

Lol, in the olden days a donut was actually the ball

[sleek cave]

oh

[woven solar]

hmm

[low condor]

If you reheat a jam doughnut in a microwave, the jam inside is as hot as the centre of the earth

[proud cosmos]

Sounds about right

[sleek cave]

petition: rename your stereotypical ring-shaped donuts to toruses

that's what they're supposed to be called smh

[proud cosmos]

Yes

Might as well rename mugs as well

Still running btw

[low condor]

ouch

[proud cosmos]

Yea

My code is made really badly and I know it

THE DOPAMINE

aaaaaaaaaaaaaaaaaaa

NO I HAVE TO RUN IT AGAIN

FK

I'm so happy

wait nvm I don't I got that already

[robust heart]

ok I think I'm getting the grip at day 12 (FINALLY)

[molten crow]

This is my first AoC, and I've done alright I think for 2021, skipped or got stuck on 2 or 3 puzzles using C#, my first language with 20 years strong experience. Now I hope to complete the 2019 AoC in Python, my second favourite language with 2 years self-teaching

[devout tusk]

is there a competition like AOC that runs for an entire year? Only 1/2 questions each day with varying difficulty?
That would be fun to do!

[proud cosmos]

YESSSS

its doooooooone

i somehow did ||manhatten|| wrong

im so good at thisss

aaaaaaaaaaaaaaaa

That took like

10 hours since I first saw the problem

Though to be fair I did not programm 10 hours long

I did take breaks

It's not even that bad

[robust heart]

leetcode?

[devout tusk]

that is still quite employment oriented.

Basically puzzles like AOC or Project Euler with a constant leaderboard (which is optimization oriented instead of speed of solution) that have new problems everyday

[proud cosmos]

That sounds close to impossible to create though

making puzzles is pretty hard

[woven solar]

which is optimization oriented instead of speed of solution
that'd be codingame's optimization or bot programming puzzles, which I love

but they are far from one per day lol

[devout tusk]

but what if there's an entire team behind it?
I guess this is a good idea for a DAO.

[potent pumice]

somewhere I read it takes 6 months or something

[sharp panther]

Eric Wastl said that it takes him 4-5 months to come up with all these, yeah

The puzzles are most of the work; preparing a new calendar and a new set of
puzzles each year takes all of my free time for 4-5 months. A lot of effort
went into building this thing - I hope you're enjoying playing it as much as I
enjoyed making it for you!
view-source:https://adventofcode.com/

[sonic silo]

are there bot programming puzzles?

i wanted to build one

[pallid quail]

So I've been self teaching myself python - I am about 1 month into my journey so far, and half way through the online bootcamp I've been following. Decided to take a break, and start on advent of code to see if I could do these challenges.

I've cleared day 1-3 so far. I'd be curious to know if anyone here would be willing to have a look at my solutions and offer feedback on whether my approaches are considered good, bad, over simple, over complex, or what.

Being that I am self teaching - I've got no one to give me feedback on my code really.

[woven sable]

Feel free to share your code, just remember to put it in spoilers since some people haven't completed thoes days yet

those*

[pallid quail]

I was going to just link to my github.

My day 1 and 2 solutions are on here - so of course - spoiler warning

https://github.com/max-is-loud/adventofcode21

I just solved day 3, but I am cleaning up the code a bit before I commit it.

Changed me nick - forgot to update it in discord.

[low condor]

@pallid quail without going too deeply into it, your code looks pretty clean so that's a plus

[pallid quail]

Thanks - glad to hear that. Been putting effort into that - not sure if I am following all of the proper python style conventions though.

[low condor]

If you are ever unsure, there's a handy song for that :)

https://youtu.be/hgI0p1zf31k

[pallid quail]

haha nice

I'm especially proud of day 2, where I wrote my first recursive function - did that again on the day 3 solutions.

[low condor]

This year is partly Advent of Recursion

I've written a lot more recursive solutions than I've done before

[pallid quail]

yeah, as I was noodling that, I sort of accidentally stumbled into recursion - the online course I've been doing hadn't even introduced that yet.

I just sort of wondered - can a function call itself, googled it and thats how I learned recursion.

[low condor]

Ah nice!

AOC is pretty good for that

It's not too common to write recursive solutions in Python

Most of the time you use loops/iteration

AOC is good for utilising idioms like that more often

[pallid quail]

Honestly, I've learned a TON just in the last 3 days.

That, and I've also I think just by the extra practice have further strengthened the new python neural pathways in my brain.

Things come a bit easier now, instead of having to sit and try to remember syntax and stuff.

I can focus on just writing code to try and solve whatever it is I am working on.

[proud cosmos]

Nice!

[wooden jewel]

I just have my editor setup to format with black on save, and mostly stopped worrying about formatting other than that

[pallid quail]

I've no idea what that means.

[sharp panther]

black is an automatic python code formatter. See https://black.rtfd.io and #black-formatter.

[pallid quail]

neat - I'll have to look into that.

[glad tartan]

I think i've used recursion just twice so far

[woven sable]

Right, I've learned a lot about recursion with AOC

[pallid quail]

I've posted my day 3 solution

https://github.com/max-is-loud/adventofcode21

I think its my best code so far!

[sleek cave]

lmao didn't know readthedocs.io had rtfd

[sharp panther]

took me a while too haha

I only saw it in the channel description for black

[limber needle]

yeah really easy to use

just do black <filename>.py

[sleek cave]

oMg hOW DarE yOu SAy rTfd ThAt'S TOxIC!!!111

wait mods get pinged when regex detects r t f m as a single word right

[sharp panther]

oh wait, does that mean the same thing as "rtfm"?

woops

[sleek cave]

read the fucking documentation, yeah i think that's what it means

[limber needle]

debatable

[sleek cave]

pings the entire mod team

congrats

[limber needle]

i've learned it as Refer to the following documentation

[sleek cave]

oh?

hm, debatable indeed

i'm not sure then

[limber needle]

and that's why people send a link along with "rtfm" or "rtfd"

[sleek cave]

pretty sure rtfm is not refer to the following manual

[pallid quail]

lol

[limber needle]

contextual i guess

[sleek cave]

https://www.dictionary.com/browse/rtfm

[limber needle]

if i'm not being passive aggressive i think it's fine

[wooden jewel]

literally 1984

[sleek cave]

uh oh sorry mods for the ping lmao

[sleek cave]

and sorry for this, too

lol

[sharp panther]

we did not get pinged :)

[sleek cave]

what. kat, you lied to me

i'm pretty sure kat said there was a filter for rtfm

[sharp panther]

there may have been at one point

[sleek cave]

yep, DATA got filtered

[sharp panther]

now that one did ping, thanks

[sleek cave]

oops sorry!

[sharp panther]

all good haha

[sleek cave]

i may have gotten the two mixed up - DATA is the one being filtered and pinged

[minor cave]

rtfm in a link won't ping iirc, I don't remember if regular rtfm pings

[sleek cave]

rtfm

does it ping

[sharp panther]

it does not, although I should probably suggest that

[minor cave]

Sometimes rtfm is referred to in actual docs, so I think we did have it at one point but it was too noisy

[sleek cave]

what is the regex for DATA? the whole thing or just d ontasktoask.com (added space so no potential ping)

[sleek cave]

ah, okay

.src filtering

[minor cave]

It's in Python, let me pull the filter for you

Although it's added via command so, but iirc we added it as a domain filter

[proud cosmos]

If only that website was written in a nicer way

The meaning really isn't that wrong in many cases

It's worded so mean though

[wooden jewel]

yeah it's genuinely useful advice in a vacuum

[minor cave]

So the website content itself is honestly fine, we don't filter it because of the website itself

Rather, it's how others here were using it

[sharp panther]

I think the website is written fine, it's the dismissive manner in which it is dropped that becomes a problem

[limber needle]

does XY problem get flagged

[minor cave]

!xy

[marsh current]

xy-problem

The XY problem can be summarised as asking about your attempted solution, rather than your actual problem.

Often programmers will get distracted with a potential solution they've come up with, and will try asking for help getting it to work. However, it's possible this solution either wouldn't work as they expect, or there's a much better solution instead.

For more information and examples, see http://xyproblem.info/

[minor cave]

I hope not =P

[limber needle]

anyway

[minor cave]

But someone new to the server would ask "hey, can anyone help me with something" because that's the polite thing to do in non-internet spaces. Someone would just drop the don't ask to ask link and say nothing else, making it a very bad experience

[limber needle]

yeah pretty sure it's stated to not do that in the coc

doing AOC has helped me garner a lot of useful tricks and modules in my Python armoury

especially collections

and re

pythonifying my severely grep'd regex has been my fear until now

[wooden jewel]

Yeah, I really like the fact that everyone is mostly on the same day, makes discussion for that day much more productive

[sleek cave]

ah yes, i do agree with that. i haven't considered this case until now, where it's common sense to ask to ask irl

so someone who's just starting to be on the internet could have a bad experience

[proud cosmos]

AOC has teached me a lot too

like the fact i am really bad with arbitrarily rotated data

and I guess also some positive things

Recursive regex is pretty cool was quite happy to use that

[sleek cave]

lol, aoc taught everyone they can't read http://adventofrealizingicantread.com

[sleek cave]

r- recursive regex??????

[proud cosmos]

Yea

[sleek cave]

what.

i've never heard of that

[wooden jewel]

https://adventofiaintreadingallthatshit.com/2021/day/16

[proud cosmos]

r"\[((?R)|\d+),((?R)|\d+)\]"

Here is some

[limber needle]

ooo interesting

[proud cosmos]

||Used it for day 18 to parse the snailnumber||

[low condor]

I used recursive regex in ||2020||

[proud cosmos]

||It basically either matches a number on each side of the , or itself which is another snailnumber, so it works quite well to capture the outer part||

[limber needle]

[sleek cave]

||... i just literal_eval'd||

[limber needle]

||same, ast.literal_eval for me||

[sleek cave]

lmao

[proud cosmos]

Honestly maybe smarter but eval bad

And regex is pretty cool

[sleek cave]

b- bUt iT's LitErALeVal!!!111

[wooden jewel]

||json.loads anyone?||

[sleek cave]

that is actually so smart

i should learn recursive regex

[limber needle]

functions exist so that we can use them

[proud cosmos]

It's cool but I also just noticed that loading it as ||JSON or literal_eval|| is like 100% easier

Still like it though

[sleek cave]

ohhhhh smart

wait, ||json can load lists?||

[wooden jewel]

||yeah json has lists||

[sleek cave]

agreed, don't reinvent the wheel

okay ima go do d15 now to get my missing stars

bye

[placid lake]

yessir

it's time to.. refactor my horrific solution for today

[sleek cave]

okay there is no way im brute forcing d15

||but i don't know how to implement a* 😔||

time to learn it

[limber needle]

||Check outheapq module||

[minor cave]

<@&831776746206265384>

[topaz summit]

!shhh

[marsh current]

✅ silenced current channel for 6 minute(s).

[iron shadow]

!unshh

[marsh current]

✅ unsilenced current channel.

[limber needle]

that was scary

anyway

||for me doing day 15 it was all about heapq||

i really want an AOC day where i can load stuff to the GPU

[proud cosmos]

You can prob dop that for many days

not sure if you want to though

[wooden jewel]

today might be a good candidate?