Awesome! Joined!
MicroPython sounds interesting, I targeted embedded as well with my first one - but just used C 
I’d like to try at least a few in Go and Rust this year
1 Like
One hour to the next reveal! I’m going to try to do it fast 
1 Like
I believe in you! Youre crushing our two-person leaderboard 
1 Like
I just completed “Rucksack Reorganization” - Day 3 - Advent of Code 2022 #AdventOfCode
Whew, that was a hard one to wrap my head around. Was happy to finish in ~15 min (~8 for part 1) but not nearly fast enough to get on the leaderboard 
1 Like
Did it this morning! Not nearly that fast and I feel like this time I made substantial changes between part 1 and part 2 which ive been trying to minimize…
Did you see someone was able to get the AI to solve it in 10s?? Crazy world we live in!
Edit: heres a blog post, but i dont think thats the person at the top of the leaderboard who seems to be automating the submissions… Ai solves Advent of Code 2022
2 Likes
Just a few minutes to go for the next one! I’m going to try abandoning all principles of software development and just doing everything directly at the Ruby command line.
1 Like
That seemed a lot faster but I got tripped up on all the levels of nested brackets and parentheses and wasted a few minutes untangling it. 8:24 was slow enough that almost 2000 people finished before me!
Terrible snippet of Ruby I used
eval(p.sub('-','..')).to_a
Idea for tomorrow: set up guard to run my script and output the value on every change, so I can just hit Ctrl+S to see the result in another window and keep editing.
Prototype of such a script
runme = proc do |_, _, files|
files.each do |f|
p eval(File.read(f))
end
end
guard :yield, run_on_modifications: runme do
watch(%r{\.rb$})
end
By the way, looks like our leaderboard has a new member! Who are you?
If anyone else wants to join, you can still use code 2516370-4dd8a512. I’ve adjusted the leaderboard to sort by total stars, so there’s no penalty for being slow.
1 Like
First Rust solution! And this time the there was just a ~10 character diff between part 1 and part 2!
code (I am just going to abuse `unwrap` 😅)
use lazy_static::lazy_static;
use regex::Regex;
use std::fs::File;
use std::io::BufRead;
use std::io::BufReader;
lazy_static! {
static ref RE: Regex = Regex::new(r"([\d]+)-([\d]+),([\d]+)-([\d]+)").unwrap();
}
fn one_contains_other(s: String) -> i32 {
let caps = RE.captures(&s).unwrap();
let val_at = |i| caps.get(i).unwrap().as_str().parse::<i32>().unwrap();
let a_start = val_at(1);
let a_end = val_at(2);
let b_start = val_at(3);
let b_end = val_at(4);
if a_start <= b_start && a_end >= b_start {
1
} else if b_start <= a_start && b_end >= a_start {
1
} else {
0
}
}
fn main() {
let f = File::open("input4.txt").unwrap();
let reader = BufReader::new(f);
let total: i32 = reader
.lines()
.map(Result::unwrap)
.map(one_contains_other)
.sum();
println!("Total: {}", total);
}
guard sounds like a great idea, fast feedback is so good. It looks like there’s a Cargo package that does the same, I’ll have to try it out!
lololol at exec, but that is actually very creative!
1 Like
I think exec might be the programming equivalent of an empty triangle, almost always a terrible idea. Actually I like your idea of splitting into all four parts first but it always ends up being clumsy when I try to do that at the command prompt.
Okay, for the record here are my Ruby one-liners (line breaks added afterward):
solution
File.read('input.txt').split
.map { |l| (l1,l2) = l.split(',').map { |p| eval(p.sub('-','..')) } }
.map { |a, b| (a.to_a - b.to_a == [] || b.to_a - a.to_a == []) ? 1 : 0 }
.sum
File.read('input.txt').split
.map { |l| (l1,l2) = l.split(',').map { |p| eval(p.sub('-','..')) } }
.map { |a, b| (a.to_a & b.to_a != []) ? 1 : 0 }
.sum
And here’s a prettier way to do the comparisons that of course I wouldn’t have come up with on the spot:
solution
File.read('input.txt').split
.map { |l| (l1,l2) = l.split(',').map { |p| eval(p.sub('-','..')).to_set } }
.count { |a, b| a >= b || b >= a }
File.read('input.txt').split
.map { |l| (l1,l2) = l.split(',').map { |p| eval(p.sub('-','..')).to_set } }
.count { |a, b| !a.disjoint?(b) }
Okay, I played around with ChatGPT a bit:
2 Likes
Easy one today. I was 3x slower than the leaderboard.
I tried pasting the problem into ChatGPT later and it generated a fine program with one bug, so I told it about the bug and it fixed it correctly.
3 Likes
I agree! So #4 i tried Go, and it took me hours just wrestling with the syntax. (Not Go’s problem, just a new language for me). Today’s I went back to comfy python and only took a few minutes 
I did it in interactive mode and didnt even use file io!
1 Like
I’m having trouble understanding a problem. Does anyone understand what this problem means: #675 $2^{\omega(n)}$ - Project Euler
Specifically:
What does Summation d|n mean? First time I see d|n and also a summation sign without lim on top
Usually that’d be sum of the divisors of n.
So, the sum of all d such that n=d•x for some natural number x. It fits with the example that’s given, since the divisors of 6 are 1, 2, 3 and 6.
1 Like
oooo tyvm! Yes, exactly! Searching for | was hard haha 
The big-numbered questions (500+) are really ridiculous! Just blatantly use numbers in the 1e20+ as starting point.
These are not problems, they’re just wealth-check!
I tried so hard solving a problem and when I get into the comment section I see people casually solving it with basic 0(n^4) loops.
The comment section is, like, “Ooh I just bought a new $90000 quantum computer but it’s not to my liking, it takes a whole 10 seconds to finish running an infinite loop.”
2 Likes
I’d consider those people failing to solve the question. If it doesn’t take much time because of expensive hardware, you haven’t really written efficient code.
I always thought the problems are designed such that with the right mathematical insight, they are solvable on an average laptop in less than a minute. But perhaps they gave up on that along the way…
4 Likes
I think the 1 minute rule still works, but I haven’t checked the latest problems.
I think there’s a caveat here, because it also depends on the programming language. I use Python because I don’t really know any other language it’s easy, but the excecution has a ton of overhead.
Quite a few times I’ve concocted a solution that takes, I don’t know, 30 minutes, go into the forums expecting a better solution, but all I find is actually the same approach on Java or C++ and it takes 30 seconds or something.
2 Likes
Please my friend, trying doing the same thing in Java****-ingScript 
The thing can’t even handle number bigger than 1e7!
I had to learn how to make C# console app because it’s the only other language I know.
But on the bright side, C# is great. Got a neat console app framework, and not to mention multi threading:
FP wasn’t that hot while writing JS, but when I switched to C# and get a taste of 10 threads in parallel it’s dangerously addictive.
1 Like
That’s the gateway drug. Soon, you’ll be harvesting the power of your GPU for fast convolutions 
1 Like
But I still try to keep in the 1 minute rule, by ensuring total time of all threads doesn’t exceed 2 mins.
But just think about waiting only ~15 seconds for an answer that normally take you 1 minute is already a great incentive 