[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
▼ Settings Mobile Home
4chan
/g/ - Technology

Name
Options
Comment
Options


Gimmicks


PEE text
My Bans
Verification
File

    08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
    05/04/17New trial board added: /bant/ - International/Random
    10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
    [Hide] [Show All]




    >https://everybody.codes/events

    ETA 29 hours

    Discuss what programming language you will use

    NO AI ALLOWED
    >>
    >>110160986
    >Discuss what programming language you will use
    I use the Claude Opus 5.5 programming language
    >>
    You're unemployed bro, just get real.
    >>
    >>110160986
    gpt astra ultra mode
    >>
    File: 1792905371282.jpg (270 KB, 1811x1744)
    270 KB JPG
    >NO AI ALLOWED
    >>
    > This site requires JavaScript to function properly. Please enable JavaScript in your browser settings.
    No I don't think I will
    >>
    Probably Clojure until it gets too annoying to use.
    >>
    >>110160986
    Didn't it start on November 1st last year? Why so early?
    >>
    /g/derboard
    4151e976-7b9f-4fda-bb33-fa7c70fdceea
    >>
    >>110160986
    Kimi K3.1, but I might give GLM a try too, I'll see how it goes.
    >>
    Haven't been been back here much since AOC ended, you kids still use computers, right?

    PL: Mathematica
    >>
    Thread is dying, perhaps Nobody codes?
    >>
    >>
    >>110160986
    possibly c# because i've taken up unity game modding again and need to dust off my OOPoop abilities a little
    alternatively zig if i get bored
    >>
    >>110161355
    >Filtered by step 0
    You hate to see it.
    >>
    I thought this was in November? I'm going to be too busy through the rest of October.
    >>
    >>110165620
    >I'm going to be too busy
    Filtered.
    >>
    >>110160986
    I will check this out. I never heard of it. I hope it's better than aoc
    >>
    >>110165678
    You can start now on the previous two years. Last year, EC was significantly better than AOC.
    >>
    >>110160986
    >29 hours
    is this the month long one or just one of their one-offs?
    >>
    Common lisp and I might use ai... like everyone else lol
    >>
    File: ec2026.png (131 KB, 1688x941)
    131 KB PNG
    >>110165826
    4 weeks long, but not on weekends.
    >>
    >>110165857
    >I might use ai
    Filtered
    >>
    >>110165678
    Lots of EC puzzles last year had hidden images when you visualized the input, which is more effort than AoC usually puts in.
    Especially compared to AoC 2025 where Eric really phoned it in.
    >>
    Will we even get an Advent of Code this year? Still nothing announced.
    >>
    File: d4orqnaz4kvh1.jpg (44 KB, 577x721)
    44 KB JPG
    >>110166016
    I'm not looking for a McJob nigger, fuck your shit geeeg
    >>
    >>110166792
    Didn't they say it would be fewer puzzles from 2025 on, implying there would be an event?
    >>
    >>110166928
    But last yes the AoC was already announced by now. I am worried that it first was cut and then dropped.
    >>
    >>110166008
    Why start at such an awkward time?
    >>
    >>110167010
    It's midnight in CET, Amerishart. That makes it literally the afternoon for you. This is a lot less awkward than AoC, and it's still a lot more favorable to you than europoors.
    >>
    >>110165036
    Actually based choice. I haven't used that language in years.
    >>
    File: rust.png (11 KB, 280x280)
    11 KB PNG
    I'll be using idiomatic Rust
    >>
    >>110168859
    Make sure to implement a reflection module yourself or this post will age like sunlit butter
    >>
    >>110167010
    Those are coundown timers. The first challenge will start in 3 hours 30 mins.
    >>
    >>110166945
    It either needs to be dropped, or Eric should pass the torch to others who have more enthusiasm. There's no point in doing it if he just phones it in like last year.
    >>
    File: golang.png (2.43 MB, 1729x910)
    2.43 MB PNG
    >everybody codes
    lets go!!
    >>
    File: 100%.png (58 KB, 437x567)
    58 KB PNG
    post 'em
    >>
    >>110167343
    probably means the day, last week of october feels random
    >>
    >haven't written any code at my job in like 6 months
    actually interested to see how dogshit i'll be at this compared to last year
    >>
    5 MINUTES
    >>
    >>110172516
    FUCK
    >>
    Mathbros win again.
    >>
    I forgot the input changes for each part, AAAAAACCK.
    >>
    >works on test but not real input
    filtered on day one, RIP
    >>
    part 4
    ultra pinecone magic uses a spiral
    only odd length squares again

    3 x 3   5 x 5   7 x 7
    .## #.... .######
    .#. #.### .#.....
    ... #.#.# .#.###.
    #...# .#.#.#.
    ##### .#...#.
    .#####.
    .......
    >>
    3 x 3     4 x 4      5 x 5  ...

    ### #### #####
    #.# #..# #...#
    ### #..# #...#
    #### #...#
    #####


    >The first outpost had 3 enchanted pine cones, so it had to collect 5 more to build a 3 x 3 frame.
    >The second outpost had 11 pine cones and needed 1 more to complete a 4 x 4 frame.

    this makes no sense to me. can someone explain why 3 requires a 5x5 and has a remainder of 5?
    >>
    >>110160986
    I forgot how to code without AI.
    >>
    >>110172796
    oh I'm an idiot were only counting the border. so its 8 - 3 because the border is 8 characters
    >>
    >>110172796
    you don't want to waste any pine cones, so unless you have exactly the amount of pinecones for a square border, you want to know how many you need to gather for the next biggest border. on part two, you need an empty space in the middle so a 3x3 is the smallest size possible.

    so if you have 3 pine cones, you're under the minimum for a 3x3 and have to gather enough to make a 3x3 (8 total since there is no center)

    if you have 11 pinecones, that's more than you need for a 3x3, so you look at a 4x4, which needs 12 pine cones, so you need one more

    yes it's a bit convoluted, kind of giving AoC a run for its money on puzzle descriptions here
    >>
    >>110172753
    >neovim was truncating the part3 input when i pasted it in
    FUCK ME
    >>
    File: 2026-1.png (766 KB, 515x793)
    766 KB PNG
    today I learned I still have terrible reading comprehension. not too bad though. I'm too lazy to optimize the inner loops. did anyone do it all in 1 inner loop?
    >>
    >>110173238
    >loops
    Anon, you don't need any loops. You can math it.
    >>
    >>110173352
    lol damn thanks for letting me know. I tried like the square root + 1 squared and junk but couldn't figure it out. I'll give it a second shot.
    >>
    >>110172802
    i dont know how to code even with ai.

    i wrote this program yesterday from a tutorial, and i couldn't even print an integer without using printf (not allowed according to the tutorial, to build up skills), just the write function.

    i feel like a noskills vibecoder ;_;

    #include <unistd.h>
    #include <stdio.h>

    int ft_strlen(char *str){
    int i = 0;
    while (str[i] != '\0'){
    i++;
    }
    return i;

    }

    int main(void){
    int i = ft_strlen("the size of the string is: \n");
    write(1, &i , sizeof(i)); // <<--- right here

    return 0;
    }
    [\code]
    >>
    File: 2026_quest1.png (136 KB, 1673x1490)
    136 KB PNG
    >>110173424
    For this problem it's probably not worth doing it the mathfag way anyways, and certainly not during the competition.
    >>
    >>110174180
    ohh nice! okay now I don't feel so bad. except for that first one! missed the ciel! the others I'd never get.
    >>
    File: yikes.png (14 KB, 943x66)
    14 KB PNG
    Someone get reddit on the case. This has compromised my safety and I just cannot even.
    >>
    File: 1.png (76 KB, 1989x540)
    76 KB PNG
    idiomatic Rust solution
    >>
    Imagine wasting your lifespan on this. Hours you will never get back.
    >>
    >>110174689
    Filtered
    >>
    >input changes for every part
    Why would they do this?
    >>
    I use whatever the AI is doing when I tell it to do stuff for me
    >>
    >>110174689
    It's literally the equivalent to something like a crossword puzzle or sudoku or picross or any of the millions of puzzles humans have done for millennia. The poisoned techbro mind cannot comprehend doing something for fun for its own sake, it must always be tied to some valuation.
    >>
    >>110174157
    Stay strong vibecat, ask the AI what you did wrong
    >>
    >>110174990
    enjoy your snailcat sudoku while we create 5D spiderman animal crossing hyperchess
    >>
    >>110175004
    I will thank you :)
    >>
    >>110175015
    based, total snailcat vibecat friendship
    >>
    >>110175015
    Dang this is adorable, not looking to sponsor pure land buddhism or anything – you might have triggered a dream memory; thus won.
    >>
    >>110174157
    Don't worry, once you've learned C++ every other programming language will become far clearer. I learned from this guy, oh shit, a whole decade ago: https://www.youtube.com/watch?v=tvC1WCdV1XU&list=PLAE85DE8440AA6B83.
    >>
    >>110160986
    Oh shit thank for the heads up. I'll be looking forward to this once I have some time after university exams starting tomorrow.

    Anyone else using C++ or Ocaml? I'm planning on using C++ for most challenges and maybe re-writing them in Ocaml to compare how nice code I can write.
    >>
    >>110172420
    >>110172802
    Me too, I felt like a retarded babbling monkey while solving it. Felt good tho.

    >>110175015
    Holy based.
    >>
    Which language should I use if I'm optimizing for maximum (You)'s?
    Serious answers only
    >>
    >>110175321
    Malbolge
    >>
    i just dont know math so i had to ask the clanker for the math formulas and then i solved it
    >>
    >>110175321
    Rust
    >>
    File: 1769641569314.png (38 KB, 679x897)
    38 KB PNG
    idiotic c# solution
    >>
    File: BASIC09.png (563 KB, 1200x670)
    563 KB PNG
    >>110160986
    BASIC09
    >>
    File: q1.png (34 KB, 1099x673)
    34 KB PNG
    >>110175321
    Excel.
    >>
    >inb4 seething poorfags and tinkertrannies

    I'm looking forward to doing this year's programming puzzles together bros.
    At least AI hasn't taken this away from us yet.
    >>
    >>110175777
    You can just post your Swift solution without advertising, you know.
    >>
    >parts have different inputs
    >>
    >>110176080
    A staple of everybodycodes.
    >>
    >>110176080
    Gone get yourself a download input submit solution download next input script from Claude. Works a treat
    >>
    not bad
    won't bother with math faggotry unless bigboy
    >>
    Clojure continues to frustrate me, but so far so good.

    >>110175777
    What is ECKit?
    >>
    >>110176673
    >bigboy
    every integer from 0 to 1,000,000 inclusive

    Part 1: 666166500
    Part 2: 1500024
    Part 3: 1334836000
    48.703ms
    >>
    >>110178973
    solution for part 1 is 666166501 though
    32 ms with >>110174664 fixed to handle 0 cones
    >>
    >>110179130
    I'm getting 666166500
    >>
    >>110178973
    >>110179130
    >>110179216
    >The smallest valid formation was therefore a 1 x 1 square
    Anon's right, it should be 666166501
    >>
    >struggled with basic math
    i have gone completely fucking braindead over the past year. it wasn't even this bad when i first started programming after not touching any form of math for 10+ years
    just deleted my slopbot account
    >>
    >Log in with your favorite big tech account
    No thanks
    >>
    >>110178973
    >>110179130
    Also part 3 should be 1334836008

    Part 1: 666166501
    Part 2: 1500024
    Part 3: 1334836008
    >>
    >>110179448
    Just create a new Github account with an alt email account.
    >>
    File: Day01.png (96 KB, 785x1077)
    96 KB PNG
    Fuck I didn't know this was kicking off so early, almost filtered before I had even started. Have some hurried CShart.
    >>
    LESS THAN ONE HOUR
    >>
    i'll miss puzzle drop in 20 minutes, but take a pity bump so this shit doesn't die on day two
    >>
    5 MINUTES
    >>
    Topsort your way to the top.
    >>
    im getting filtered by part 3.. uh oh
    >>
    Okay I was pretty certain part 3 was going to be that, but I still took the easy way out for part 1.
    Luckily part 2 funneled me into the answer, if it was just part 1 and part 3 I would've taken a lot longer. But what I had to do for part 2 gave me a big hint for how to solve part 3.
    >>
    File: 1769402013526.png (165 KB, 700x1500)
    165 KB PNG
    idiomatic c# solution

    why are these harder than last year tho?
    >>
    Part 1 took me so long and made me feel stupid.
    Part 2 and 3 were nearly instantly solvable because I did part 1 well and made me feel smart :)
    >>
    >what programming language you will use
    perl because why not lol
    >>
    >>110181453
    What's the easy way out?
    >>
    What's the smart way to solve it?
    The logic I think is necessary is:
    Every alloy ID should have a list of alloy IDs that are weaker than it.
    For every pair of alloys N and M, where N is stronger than M,
    update list N to include alloy M,
    update any list K that includes N to also include M,
    update any list J that includes K to also include M,
    repeat as necessary.

    Is there a simpler way to do it?
    >>
    >>110181904
    Topologically sorting a directed graph.
    >>
    File: file.png (80 KB, 524x899)
    80 KB PNG
    >not really sure how to do part 1
    >arbitrarily come up with some bullshit
    >probably O(n^2) or some shit I don't know
    >try it
    >it works
    >get to part 2
    >try part 1 code verbatim
    >it works
    >get to part 3
    >try part 1 code again
    >it still works
    weird
    >>
    semi washed idiotic rust
    >>
    >remembering everybody codes as a november thing.
    >visit randomly
    >missed a day
    Fuck
    >>
    >>110178658
    Well here we are on day 2 and I have already decided fuck clojure and your immutable data structures. Trying to think in functional hurts my brain too much.
    >>
    using ai enough makes you complacent to ass :')
    >>
    >>110181972
    lisp is literally a bad meme. I know it seems crazy but if it were actually good, people would actually use it for useful software. It's been around for decades and has mostly faded into obscurity. Enthusiasts can't get over it because it's good in theory, in practice it fucking sucks.
    >>
    >no AI
    so no search engines or documentation that’s all written by AI (except man pages)
    >>
    >>110182082
    It's not even the lisp style that bothers me anymore. I was having fun with fennel lately. It's the combination of lispy syntax, functional style, and lack of mutable variables that makes something simple like building a graph feel painful and tedious.
    >>
    >>110182102
    sorry but the only sanctioned way to participate is to only use knowledge you already had as of approximately ~2020
    >>
    >>110178973
    <1s w/o changes
    >>110179130 and >>110179485 are right
    >>
    >part 2 solution works on part 3 first try
    >>
    File: 1779309396961271.png (336 KB, 1748x1950)
    336 KB PNG
    >>110182110
    wasn't planning to clojure this time around, but here's one possible solution
    pretty much a direct translation of >>110181944
    >>
    >>110181904
    mine solution was basically:

    >make a dict where the key is the allow and the value is a list of alloys known to be weaker than it
    >make a list of all alloys
    >find the alloy that isn't in any other alloy's "weaker than" list
    >that's the strongest
    >now iterate through the list again, ignoring the strongest alloy, to find the other alloy not in anyone's list
    >that's the second strongest alloy
    >keep iterating til you've assigned them all

    probably a faster way of doing it, but it's pretty simple conceptually
    >>
    >>110182412
    Yeah that's way simpler than what I was doing. Faster, too.
    I think I was overthinking how much information could be implied with the simple x>y statements.
    Thanks, Anon.
    >>
    File: ec-2026-02.png (1.06 MB, 599x1482)
    1.06 MB PNG
    haha funny day. late start, but finished up the topological sort... and it worked for all 3 parts??

    real cool B)
    >>
    >>110181811
    >easy way out
    Part 1 has every alloy interaction exactly once, so if there are five alloys, the alloy that is stronger than four other alloys is the strongest, the alloy that is stronger than three other alloys is the second-strongest, and so on. Counting how many times an alloy appears as the stronger one in a pair gives you the order, and the weaker alloy isn't relevant.
    >>
    File: 2.png (209 KB, 2136x1452)
    209 KB PNG
    idiomatic Rust solution
    >>
    can your code solve this?
    10>5

    you should get "105"
    >>
    File: 1782483209937155.png (1.35 MB, 1700x1100)
    1.35 MB PNG
    >>110182412
    haha mad props! you pretty much invented topo sort on your own. the only difference is you recursively go into all elements in your "weaker than" list then append them after recursing... actually it might be backwards, but still!
    >>
    What do you all listen to while solving the problems?
    >>
    >>110183582
    Just the constant intrusive homoerotic thoughts in my head
    >>
    accidentally solved all three parts at once
    fun day today
    >>
    >>110183614
    Do you give in sometimes?
    >>
    File: q2excel.png (234 KB, 3260x1727)
    234 KB PNG
    ... but why?
    >>
    File: img_1_1766594260305.jpg (95 KB, 1070x1119)
    95 KB JPG
    >>110184185
    >excel
    >>
    >>110165036
    >embarcadero
    jesus fucking christ
    >>
    >>110183614
    Glad I'm not alone, Anon.
    >>110183737
    I wish I had someone to give into them with.
    >>
    Do you guys use auto complete?

    My internet went out recently. So I was trying to code for fun and realized I'm a retard without an LSP. So now I've just turned autocomplete off completely so I can just memorize everything. Or thats my plan.
    >>
    >>110186369
    >LSP requires internet
    dumb retard
    >>
    >>110186395
    :-0 Oh did you write your own?
    >>
    >>110186998
    What the FUCK are you talking about? Do you even know what LSP is? Holy shit, just ask Claude or Grok or whatever if you're too retarded to use an internet search engine.
    >>
    >>110187027
    >pasting puzzles into grok
    kek filtered
    >>
    >>110187979
    >reading comprehension
    >>
    File: Day02.png (475 KB, 1324x3437)
    475 KB PNG
    Today was an absolute mess. Ended up writing 3 differing solutions for the 3 problems.
    >>
    >>110188336
    i dont understand your code at all, mine was more readable atleast even tho i kinda hardcoded part 1 and 2
    >>
    >>110186369
    I literally can't code without autocomplete. For one, it's a humiliation ritual, and two, I don't know all the interfaces and syntax by heart. I don't think I would be a much better programmer if I spent my energy memorizing stuff I don't use often.
    >>
    >>110188398
    I can hardly blame you, it was very much a "throw shit at the wall, see what sticks" kind of day. Probably doesn't help that I use unreadable tuple types for parts 1 and 2.
    >>
    File: ss.png (149 KB, 789x1327)
    149 KB PNG
    Trying to learn Zig.

    This stuff is a little bit harder than AoC I think. I'm not really an algorithm guy. Just trying to work the problem and if I solve it I solve it. I feel like I was close to a good solution.
    >>
    >topological sort
    Brother, just keep comparing all elements in the list with each other and swap if they're out of order.
    >>
    just randomize the list until it fulfills all the rules
    >>
    >>110189502
    Algorithmlets will be filtered after week 1.
    >>
    LESS THAN 10 MINUTES
    >>
    5 MINUTES
    >>
    LESS THAN 5 MINUTES
    >>
    LESS THAN 10 MINUTES IS STILL TRUE THOUGH
    >>
    IT'S OUT EARLY
    >>
    LESS THAN 24 HOURS
    >>
    >part 3
    Uh oh
    >>
    >>110190260
    Okay never mind, I made an assumption based off a pattern I noticed in the examples, and it turned out to be true for my real input too. I don't know why.
    >>
    i don't know what this is i've been sick for a week and i don't know how to code but i want to do stuff because i have depression and i need to be distracted

    python i guess?
    >>
    >>110190328
    nah stick to being depressed. you can get out of depression. but once you python global interpreter lock youll be stuck in python forever.
    >>
    >>110190260
    I was able to ultra-naive broot it in python btw
    >>
    >>110190328
    Well, start by solving sudoku first
    >>
    So there's definitely a way to solve Part 3 without a loop, right? I doubt there's a pure math way to get the smallest value, but maybe some kind of greedy recursive search?
    >>
    I thought of a kino solution to part 3 that Im too lazy to write myself. Imagine ISO2-xyzw ijkl = n. We do i*x^3 + j*y^2... = 137, express i as a function of j k l, so we can find if the room to minimize it is sufficient. If it is we search for a way to minimize it and if we find it vote invalid if not (in either of the two), we consider i fixed, so we express everything in terms of j instead and so on, till the second to last var, at which point we can validate the vote
    >>
    >>110190674
    >If it is we search for a way to minimize it
    To ammend myself, this shouldnt even be that hard since j k l are linear thus we can express the search space as a union of sets and reason that the number we search to achieve with i j k can fit with the step theese give us if Im making any sense
    >>
    >this part three
    oh for fuck's sake
    >>
    File: IMG_7279.jpg (27 KB, 567x567)
    27 KB JPG
    >>110190674
    It gets even better than that. Imagine we have `ISO-a b` and a function f(a, b) that computes the result. Now all we have to do is f(a, i) for i in range [0,b), and if any of these return the same result as f(a,b), we know that row is invalid and we don't count it.
    >>
    > Note that the pattern is repeated from right to left
    fuuuuuu
    >>
    >>110190735
    oh nevermind, this is very brootable. FUCK MATH
    >>
    I'm so tired I kept mixing up the standard and the value, and then mixing up base/factor/exponent. But I brute forced my way to success. Eventually.
    >>
    >>110190884
    I wasted like 10 minutes over accidentally writing =+ instead of +=. Python should definitely raise a warning or something because what would =+ even do for two ints
    >>
    >try to do DP
    >get filtered
    >try broot
    >finishes in 1s
    >>
    >>110190939
    google says that python interprets it as unary + (positive sign)... which i guess makes it basically do nothing in most cases where you'd make that mistake
    >>
    File: ec-2026-03.png (1.27 MB, 725x1184)
    1.27 MB PNG
    go go brute bros :0
    >>
    man I feel like no one is playing this. aoc gets thousands of people. for this thing I'm in the top 100 even if I start late a couple hours late and take an hour to finish. programming is dead boys pack er up
    >>
    >>110191154
    AoC is losing popularity every year, no surprise that EC is kinda dead too given it was never as popular. doesn't help that i had it in my mind this was a november thing and probably would've missed this if i hadn't seen the threads. i bet we'll see a few people next week saying they missed the first week because of this lel
    >>
    >>110191154
    >programming is dead boys pack er up
    We won't know for sure until we see the AoC numbers. Like said >>110191193, starting an event like this on October 26 makes it so only die hards are even aware it's happening
    >>
    File: elves.jpg (8 KB, 320x167)
    8 KB JPG
    >Gimlock stared at the papers for a moment before quietly announcing that he had discovered an entirely new reason to dislike elves.
    >>
    >>110190658
    >I doubt there's a pure math way to get the smallest value
    No, but it's very easy to turn into a standard minimization problem which you could feed into something like z3. The math fag terms are "Diophantine equations" or in this case it's a restricted Frobenius equation where digits are less than 9.

    Real mathchads just bruteforce.
    >>
    >>110191455
    >Real mathchads just bruteforce.
    only math i need is to be able to calculate if my BROOT will finish before the heat death of the universe
    >>
    File: 3.png (152 KB, 2177x926)
    152 KB PNG
    idiomatic Rust solution
    >>
    >>110191974
    Is it idiomatic for rust to be extremely hard to read, update and debug?
    What a fascinating language.
    >>
    >>110191154
    It's weird how little participation we get considering there's basically no meaningful quality difference between AoC and EC.
    Last year I remember the /g/derboard with 30-40 users making up like a quarter of all participants toward the end of the event.
    Need more fa/g/gots to join us for their puzzle fix now that AoC is in zombie mode with only 12 puzzles/year.
    >>
    >>110160986
    I'm using ocaml
    >>
    >>110192005
    skill issue
    >>
    >>110192023
    It has nothing to do with the puzzle quality. AOC is about getting in the holiday spirit with elves and saving Christmas etc. EC could in theory be about Thanksgiving or even Halloween (which would make more sense and be more fun) but it doesn't lean into that at all
    >>
    >>110192023
    Going by the leader board, we have only 17 active users. It's early days yet, perhaps more will trickle in. Also the name doesn't help, people might see it and think it's a program for newfags.
    >>
    File: q3excel.png (248 KB, 3213x1723)
    248 KB PNG
    Can't stop, won't stop.
    >>
    >>110191154
    this is my first year trying it and the puzzles are kinda soulless compared to aoc
    >>
    File: Day03.png (148 KB, 1100x1082)
    148 KB PNG
    Quick lazy broot before work.
    >>
    >>110192382
    AOC itself is soulless now, especially the 12 day variant. Stick with it, I'm sure there will be some fun challenges.
    >>
    Speaking of AOC, Eric returned:

    >Every December since 2015, I've created a bunch of free programming puzzles. It is my honor and privilege to announce that starting this December 1st, I will be releasing..... more free programming puzzles!

    >i heard you like puzzles <3

    >edit: 12 days of puzzles, like last year
    >>
    AoC 2026 place holder page is now up.
    >>
    >>110190750
    Thats how I solved it lol but it took 300ms instead of 1ms so I was trying to come up with something faster
    >>
    File: 1769009369398325.png (146 KB, 500x314)
    146 KB PNG
    >>110160986
    >NO AI ALLOWED
    >>
    >>110161357
    I did that a couple of years ago for AoC.
    Reckon I lasted 3 days.
    >>
    >>110192566
    >aoc is soulless
    agree after 2023 it moved in that direction when he started compensating for widespread chatgpt use. I think mostly I’m nostalgic for a lost era
    >>
    >>110192621
    Eric started punishing us when too many retards complained about his IntCode machine, that was the turning point. EC last year had some pretty cool puzzles, off the top of my head I remember the 'string art' and a volcano problem.
    >>
    >>110192301
    can you explain how this handles the part 3 logic? how do you know if there's a smaller number than the one provided in the iso2 that leads to an equal result?
    >>
    File: quest03.gif (3.34 MB, 1421x2437)
    3.34 MB GIF
    >>
    >>110192572
    >>edit: 12 days of puzzles, like last year
    disappointing
    >>
    >>110192390
    thanks, i didnt know how to bruteforce part3
    >>
    >>110193132
    It brutes through all the numbers from 1 to the number we currently have (since you're looking for something smaller). These numbers are zero padded to the same width, and then I use a matrix multiplication with the weights to get the potentially smaller results. Then it's just a matter of picking out the smallest, and seeing if it were equal to what you were given.
    >>
    File: doit.jpg (127 KB, 1200x846)
    127 KB JPG
    Eric seems to really resent doing it for free. Seriously, stop whining and just put the puzzles in the bag.
    >>
    >>110195938
    not trying to simp for him but making puzzles is tedious as fuck
    >>
    why do you care about these slop contests that barely have problems harder than cs101 shit
    >>
    >>110195938
    I literally give him 50 bucks every year. he needs to do it.
    >>
    >>110196508
    Filtered.
    >>110196517
    I cut my donation in half last year, because he only supplied half of the days. Instead of 0, he got 0/2.
    >>
    LESS THAN 10 MINUTES. Where is everybody who codes?
    >>
    5 MINUTES
    >>
    1 2 7 3
    >>
    dead general
    >>
    >part 3
    I don't feel like doing it
    >>
    >this part 3
    this is giving me the same feeling as the elves vs ogres rpg aoc day
    >>
    >>110198065
    Same.. but you just have to.
    >>
    FINALLY done.
    I had a stupid edge case, thankfully the mega example helped debug it.
    Brutally long problem today, and it's only day 4. What are the later days going to look like??
    >>
    File: ss.png (184 KB, 951x1322)
    184 KB PNG
    quest3 solved. I'm surprised that this is only 60 lines of Zig code. I thought that low level programs will require far more code (I am a dumb Python webslopper by trade)
    >>
    This one really required you to lock in. Final checksum is what got me, I was forgetting to skip ranks in the case of ties.
    >>
    File: 1773103072398534.jpg (41 KB, 409x389)
    41 KB JPG
    I was double counting so part 3 got part 1 + (part 1 + part 2) instead of part 1 + part 2 added to its times
    >>
    >me_running_my_brute_solution.gif
    see you guys in an hour!
    >>
    I'm too drunk to deal with these fucking edge cases. I'll finish part 3 tomorrow.
    >>
    >>110199451
    >brute solution
    What is there to brute?
    >>
    >>110199507
    instead of caching previous player scores for each part i'm just n**2 looping through them. desu it might just be broken it's been running a while.
    >>
    This one felt like work
    >>
    >>110199925
    Pretty cool, these threads need some originality.
    >>
    File: 4.png (432 KB, 2195x2462)
    432 KB PNG
    idiomatic Rust solution
    >>
    >>110200258
    I saw that.
    >>
    >>110198065
    same
    when's it gonna get fun
    >>
    File: tard.png (2.77 MB, 1080x1516)
    2.77 MB PNG
    Unwashed ass, certified 2 AM tart solution

    >>110199925
    This looks really cool, how did you made it?

    >>110199666
    >It might be broken
    >Trips of Satan
    >>
    >>110201213
    >class Warrior

    SOVL
    >>
    >dead thread
    >>
    >nobody codes
    >>
    Many of the few on /g/ would could code, no longer code. Programmers went from Gods to an annoyance no one really wants to think about. The common man is no longer impressed by these feats, because he, and even the dumbest person he knows, can summon such trivialities on command. It's hard to believe anyone even earned a living doing it, it's akin to how you people viewed manual calculation in the age of the electronic calculator.
    >>
    File: q4p1-2.png (102 KB, 3522x598)
    102 KB PNG
    Part Q4 P3 in excel will have to wait until the weekend.
    >>
    File: Day04.png (176 KB, 1295x1225)
    176 KB PNG
    Wasted far too much time wrangling everything into linq pipelines.
    >>
    Managed to phonecode parts 1 and 2 while out drinking. Unfortunately died before I could finish part 3 tho :(
    >>
    File: IMG_3495.png (241 KB, 1275x2907)
    241 KB PNG
    man debugging this one was rough. one thing i'll give the duck guys though is they definitely give you enough examples to figure it out. still though lots of edge cases.
    >>
    3 MINUTES
    >>
    Today is a filter day.
    >>
    uhh I'm getting filtered on part 1
    my code passes the example but not my input
    >>
    benis
    >>
    Visualization of my Part 3 input.
    >>
    File: 2026_q5p3.png (27 KB, 598x326)
    27 KB PNG
    >>110207448
    Nice. I'll add another (rescaled 2x)
    >>
    >my caveman-tier python broot takes 20 seconds for part 3

    I like how friendly EC is to brooters so far. The input is always sized such that the brute force solution takes just long enough for you to have to sit in shame for a bit, but not enough to actually filter you.

    Also, I know the nice queue-based solution, it just seemed a little riskier to implement. Probably wasn't though, in retrospect.
    >>
    >>110160986
    If it has those idiotic pyramid printing tasks I'm just clicking X.
    >>
    It's Friday night, why isn't /g/ programming at home?
    >>
    File: 5.png (641 KB, 2156x3278)
    641 KB PNG
    idiomatic Rust solution
    >>
    File: Day05.png (234 KB, 1179x1856)
    234 KB PNG
    >>
    I forgot there's no quests on weekends...
    >>
    >>110206068
    RIP anon shouldn't have tried driving home.
    >>
    >forgot the input changes between parts again
    i may be rarted
    >>
    File: Day05-1.png (31 KB, 655x220)
    31 KB PNG
    >Elves sending us unsolicited dick stamps
    >>
    File: file.png (39 KB, 755x576)
    39 KB PNG
    Day 1. Learning elisp, but not sure how many of these I'll do. Shitty p3 but I already spent too much time on it.
    >>
    File: file.png (88 KB, 637x1393)
    88 KB PNG
    Day 2 unwashed ass
    >>
    why is there a pause on weekends?
    >>
    >>110213860
    to give people a break i guess. i start getting burnt out on aoc by like day 10 so it's kind of nice, but that's mostly because aoc drops at like 11 PM for me
    >>
    just go around to day 4 and 5, both were pretty easy but a bit tedious (day 4 part 3 especially). day 5 is trivial if you're just tracking everything by coordinates in sets, which i assume everyone who has done something like this or aoc before just does by default.
    >>
    >>110203461
    >>110204404
    >>110204661
    EC wasn't very popular here last year, not shocking that it hasn't caught on now that programming by hand is a meme
    >>
    File: ec-2026-05.png (3.48 MB, 793x4350)
    3.48 MB PNG
    unwashed go. oh man. I'm missing them elves big time.
    >>
    F-FIVE MINUTES??
    >>
    File: animation.gif (3.34 MB, 449x245)
    3.34 MB GIF
    man I love netpbm what a good file format. so easy to just output pixel data.

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

    convert -delay 0 -loop 0 map_*.ppm -filter point -resize 150% animation.gif


    func write(grid [][]byte, count *int) {
    *count++
    m, n := len(grid), len(grid[0])
    data := []byte{}

    height, width := strconv.Itoa(m), strconv.Itoa(n)
    data = append(data, "P3\n"...)
    data = append(data, width+" "+height+"\n"...)
    data = append(data, "255\n"...)

    for y := 0; y < m; y++ {
    for x := 0; x < n; x++ {
    color := ""
    switch grid[y][x] {
    case '0': // water
    color = "25 166 188"
    break
    case '1': // water
    color = "21 144 162"
    break
    case '#': // land
    color = "21 162 31"
    break
    case '=': // sand
    color = "188 167 34"
    break
    case '@': // tree
    color = "85 162 29"
    break
    }
    data = append(data, color+"\n"...)
    }
    }

    num := strconv.Itoa(*count)
    if len(num) == 1 {
    num = "00" + num
    }
    if len(num) == 2 {
    num = "0" + num
    }
    err := os.WriteFile("map_"+num+".ppm", data, 0644)
    if err != nil {
    panic("couldn't write file")
    }
    }
    >>
    I fucking hate grids
    >>
    >>110215349
    Grids are easy. How could you hate them?
    >>
    File: file.png (160 KB, 1387x1178)
    160 KB PNG
    Day 3 unwashed
    Takes a couple seconds to finish even with native compilation
    >>
    >>110216834
    >(car parts)
    BTC
    bc1qtm4rep7kyp2p28hlkfktpnd0j7uew47vnt2mev
    LTC
    ltc1q9xlaysse0mpfwf025esd69kt9zu3md22gh23du
    ETH
    0x6abA18D203520c8F90ca63eE7426a7bC345D98e6
    SOL
    E5ekoiauCR8yiNp451dmuPzhUxo84P5x3m9WJBK5kxZe
    XRP
    r3jYXNdMk2Vz5hXhKCKE9re8uz93JQFG1r
    DOGE
    DSBq5jRWkiwmEh1G7zBwmQDgRkvRr61fmb
    XMR
    85CTwq5wGjWZw5GgzgwZPrHyZxpLXMK5oZSSSk5j4x9w1bpqzisaWAUZYaAPfh6rP2MZJcaFqF1XmEzkw6YCqJSS2Txu6Z8




    Delete Post: [File Only] Style:
    [Disable Mobile View / Use Desktop Site]

    [Enable Mobile View / Use Mobile Site]

    All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.