What languages are you guys thinking this year?
Last year I used mainly crystal. This year I'm thinking pharo smalltalk, if I can pick it up in time
I also want to do visualizations, not sure how possible that is with smalltalk.
I always use Rust, because I cannot use it at work and I am still bad with it.
New Years resolution the past 5 years: I will get better with Rust.
...and I do get better but somehow it always feels like it's not enough. Like, I'm still an imposter.
I can program an entire embedded USB keyboard/mouse firmware from scratch that can do all sorts of things no keyboard has ever done before yet I still feel like a newbie somehow. Like there's all these people that talk about traits and mutli-threaring with async and GPU and AI stuff and I'm like, "I wrote an embedded_hal crate that lets you use both 8 and 16-channel multiplexers simultaneously!" or, "I wrote an interface that let's you use the extra space in your RP2040 flash memory as a filesystem!"
Yet everything I ever write in Rust always just uses the most basic and simple features because I still have trouble with complex lifetimes (passing them around quickly gets too confusing for me) and traits that work with non-basic types (because in the world of embedded 'static is king).
Good news: if you’re writing #Rust and only using basic features of the language, you’re doing it right.
People who use the advanced stuff either have unique, interesting challenges, or they’re over-engineering. Since the former are overrepresented in the blogosphere, you’re probably comparing yourself to them. But just because their problems are interesting doesn’t mean yours are not! Nor does it mean you have to use the same solutions.
If you can solve interesting problems (it sounds like you can!) and keep the code simple, more power to you!
This year I'm thinking of a real challenge and writing brainfuck with butterflies.
Or Rust. Rust is the way.
Real programmers don't mess time messing around with butterflies or physically interacting with the world. They just intimidate the program into acting as they want through sheer fucking will
I'm going to try with an awful language so I can better criticise it in future.
I will try to learn rust while doing it, seems like an up and coming language with some interesting features.
I never get past the first 10 days anyway due to how long they start to take and how many things are going on in December.
I hate the ones that turn out to require some niche mathematical knowledge too....
I like to use lisp. It is about the only time I get to use it, and I get a little better each year.
Last year i did rust and lisp both, lisp most of the time was easier.
I skipped 2023, but in 2022 I got decently far using newer Excel functions like LET and LAMBDA. And then if I could I would golf the solution into a single cell formula. Years before 2022 I used Python. I think I had more fun with Excel. Will I be up for it this year?
Going with c# again. I know the language super well but don't often have a chance to get really deep into it with the stuff at work. These often present very non typical problems that require lesser used features.
C# is like a black hole for me. Ever since I started using it, I have been unable to really get into anything else.
If I would have learned it later, I could have learned five other languages instead ( really learned, I have still played with a bunch ).
I'd advise to pick up at least a second language for your non-critical projects. At first it feels so uncomfortable, but as you gain proficiency, you'll see it's a real ego boost. You realise that there really isn't anything you can't learn.
Whitespace
whitespace
Go. I actually I haven't written a single line of Go since last year's AoC so it's a good excuse to go again (pun intended).
Haskell! Because it fits the way I think nicely, and I don't want to write in anything else :)
Probably start with Rust again this year, although it definitely makes some of the days a lot harder. I might switch to something better for quick code if I fall too far behind.
I might even try PHP - I coded it professionally at the start of my career but haven't touched it for a decade and I'm curious to know if its improvements make it pleasant to use.
Portuguese and Gleam.
Might finally do something with Elixir. Plenty of ideas for using it with Phoenix and while I've seen a couple of tutorials for simple stuff like a live chat, I've done fuck-all thus far.
Might try V one way or another as well. Super small compiler and very small executables make me happy.
Thinking of using nim again like last year
I have previously done it in Rust, but have toyed with the idea of taking this year as a reason for looking into OCaml.
Really curious about how'd unsupervised AutoGPT handle it. Will probably require making some kind of agent that handles input and feedback loop, and it will definitely be a disaster, but could be interesting exercise.
Scala 3 again. More inline. More better.
Python. I don't have much fun in that language and I figure maybe doing AoC will help delete the mental barrier of "Python make sad" and also get me to learn more about it, so if it still makes me unhappy at least it won't be an "I have very little understanding of what I am doing" unhappy.
I don't entirely know why, but it's difficult to enjoy programming in python, even if you enjoy the end product. Something about it's design, which usually forces you to do things imperatively or reach for a library, it's position as mainly a glue language, it's philosophy of "there should only be one way to do it". They're practical philosophies, but not very fun
You might have put into a words a suspicion I've had for a long time. I don't know if this is completely the case for me. But yes, Python most of the times doesn't spark joy
Go, out of the languages I use at work, it is the one I learned most recently and have the least experience with. I am not planning to get on the leader board (or even comple more than the first week of challenges), but it is an excuse to get more comfortable with the standard library.
I want to use ArkScript this year, a small lisp like language written in C++. Might be fun, might find ways to improve the standard library!
Hm. Want to try to use the Relational Model, but am starting to hate SQL. Anyone know of any decent alternatives?
Typically when people want to avoid sql they use an orm, but um, I'm not sure why you want to use a database for aoc
I don't code enough to justify investing time and energy into becoming fluent in a new language, andd habits die hard, so... perl. It does what I need it to do, and it is so thoroughly embedded in the back of my head that it's extremely rare that I have to look something up when writing.
I want to learn rust but I don't think I have the socks for it
Been doing Ruby the last few years, since it's what I use most at work.
Might try for some C# as well though.
You should try crystal, if you haven't already
I have tried it, but it's never really clicked when I've used it.
I always use R! Sometimes I sprinkle in some Rcpp (c++ with extensions for interacting with R objects).
6502 machine code
On real hardware?
It'll definitely be running on hardware
Rust again, more than likely. I enjoyed using it last time, and want more practice
I'm boring and going to try kotlin and rust. Perhaps some other language too if something comes up till the first.
Some Mixture of Haskell and Python and maybe Rust if my friends get me to join them.
TypeScript for me. I've got an interview for a primarily TypeScript position coming up this week so would like to brush up on its quirks.
If the interview doesn't go well, then I'll probably switch to Go, though. I've been really enjoying Go for hobby programming.
Seriously? This idea of always wanting to use the latest and best technology is a nightmare for those that come after you. I used to run an IT department where I literally had more languages/technologies that I had people to support them. You need two people to support each, for vacations etc so you do the math. It is a nightmare. Spare a thought for the poor support people and stick to Python or Node or whatever - even Perl...
Indeed, you make an excellent point that Advent of Code solutions need to be maintained for years to come.
even COBOL...