commented: Glad they’re getting rid of the leaderboard. AoC is more valuable as a way to explore algorithms and novel languages than as a leetcode contest. commented: AoC is more valuable as a way to explore [...] novel languages This is by far my favorite part of it. I've done at least a few days every year since 2019 and all of them have been languages that were new to me (except 2023 when I did Rust for a second time). I think it's a really nice way to naturally explore new languages. commented: I think the decision to cut down on the number of puzzles may also reflect the way that people engage with AoC. It certainly does mine: I usually start on the 1st and then begin lagging behind when the problems get more complicated, and by the time it's the 25th I have usually completely given up on finishing it. commented: Even if they cut it to 20 it would make a big difference to me. Those last few days are distinct from the rest of the month in a lot of people's personal lives. commented: Yeah if only because end of year stuff starts happening. Will try much harder this year to 100% it commented: This makes me reconsider my decision to skip this year... commented: trying to decide if I hate myself enough to do it in TS this year. commented: I think this is great. Specifically getting rid of the global leaderboard: it’s basically impossible to get on it if you don’t live in the right time zone, and LLMs make cheating on it very easy (last year I remember seeing several people who just gave the challenge to Copilot and got in the leaderboard without even trying for the first few days), which just demotivates the people actually trying to get on it by actually doing the challenges. I love AoC, so I am glad it’s not going away. commented: I hope the challenges this year are less tedious. It's fun to solve the problem, it's less fun to glue and bodge a bunch of transformations on the data before you actually solve the problem with it. commented: Can't wait (as always)! Still trying to decide which language I'm going to try it in this year. I'm leaning towards either Gleam or Elixir but haven't looked too much yet. Going back to 2019 I've done it in Kotlin, C, Rust, Nim, Rust, and then Racket. Kotlin, the first time doing Rust, Nim, and Racket were completely new to me and all 4 times I learned programming paradigms that were pretty different from what I was used to. I don't use any of those languages day-to-day but maybe once or twice a year I end up going back and looking at my repos from all the years and admire the way the languages do certain things. commented: Try F# with .NET 10 this year. The language feels amazing to use and it's quite a good experience in vscode with the ionide.ionide-fsharp extension. commented: I see no real downsides to this, and looking forward to that feeling of being finished and yet having more time left. (I have never gotten all stars, so it's really just a feeling :P) commented: I very rarely made it to the later puzzles (by the time the last ones come around, it's holiday season and my development machine has been offline for days) and it might be nicer to play without the leaderboards showing just how far beyond the top 0.01% I am, so this seems like it's probably an improvement. commented: I loved AoC and used to do it in a different language every year. Maybe this'll make me join again. Twelve days is pretty much about as much energy as I have in any given December. commented: Getting rid of the global leaderboard is much appreciated, though I'm a bit sad about cutting half of the puzzles. The later days were always the most entertaining to me (bonus points for having to reverse engineer the input for p2) and it just sounds like there will be fewer of those, in addition to having a shorter "ramp-up". Oh well, life always takes priority, I guess, and after ten years one probably feels like something has to change to keep things interesting and/or sustainable commented: AoC has been an amazing project so far, and I'm glad the creator made some sensible decisions to scale it down instead of abandoning it entirely. commented: No more leader board? I might actually do this year’s AoC this year. commented: Usually I don't care about Advent of Code, maybe only doing a few days. A couple of years I have done the challenges during my summer vacation instead. These two changes actually make me excited to consider doing it this year. 12 challenges feel like enough. commented: This is very reasonable. AoC puzzles—except maybe the first few years—are seriously good. It must take a long time to design them in a way that encourage such enjoyable, intuitive solve paths. Having the event conclude mid-December is a great change as well. As much as I love AoC, Christmas celebrations take priority for me. commented: I wanted to do this in my own programming language this year and use the programs as both demos and tests at the same time, but then I saw that they don't allow publishing problem inputs, which means I can't use the code as tests, which makes them less useful. So I'll probably pass. I kinda understand that they don't want you to copy puzzle text, but I don't understand what the problem is with uploading the inputs to a git repo. commented: I'm not saying it's fine to upload the inputs but many people do it anyways because it's very easy to miss that part. In fact I had already uploaded some years when I first read that. And many people will still miss it because it's not something easy to notice on the website AoC has always been copyrighted and Eric has said that he doesn't want people sharing the inputs online because there's a set of inputs for each puzzle and having all of them online would make the job easier for someone making a clone. Honestly I think if someone wants to make a clone it's still easy without the inputs online but that's what he says. Other platforms go even further, for example Project Euler asks you to not upload the solving code for any problem except the first ones. commented: You can use a different input with a different result, right? commented: Are you really suggesting generating inputs for 12 problems like these? commented: I wanted to do this in my own programming language this year Nice! I just thought I could do that with mine, depending on how far along I get. It'd be a good test of what I want it to look like. .