Asklemmy

I like R because most everything is vectorized out of the box, and you can subset multiple ways. You can do stuff like

a = c(1,2,3,4)
a - 1 # 0,1,2,3
a < 3 # T,T,F,F
a + a # 2,4,6,8
a[a < 3] # 1,2

Also map (“apply” functions in R) is implemented very intuitively, and R discourages you from doing anything object oriented.

Oh my favorite is Crystal. It's a statically compiled dialect of ruby.

It supports:

• Most of the ruby goodness: custom DSLs, patching classes/mixins (monkey patching instances is not supported)
• Compile time type checking (but it also uses duck typing)
• Coroutines / fibers that work across multiple threads (multi-thread support is still experimental, but from my experience works well)
• Possible to create small self-contained binaries (like go-Lang apps).

As much as I love the expressiveness of crystal, there are a few cons:

• It's slow to compile. Due to the dynamic nature of the language, the compiler needs to parse a lot of files (think C/C++) before it creates a binary.
• The number of libraries is very immature at the moment. Crystal is a young language and is missing support for things like aws.
• The library management mechism (called "shards" akin to ruby gems) is not great (in my opinion). There are helpful tools to create the scaffolding, but if you're pretty much stuck with the defined structure. For example you cannot have a single git repo that provides a library and an application that uses it.

Other than that, the type checking but with ruby-like syntax is awesome!

edit: fixed formatting

Brainfuck !

Not meant to do real things with it, it's more of a puzzle game, but a really brain hurting one

Can't believe I scrolled all the way down and didn't find Scala. It's the only language with decent traction that beautifully and elegantly combines functional programming and object oriented programming. Scala makes it such that the language does not limit you into a certain paradigm. You can translate your algorithm in your mind into code regardless of how you thought of it. Incredibly flexible where you need it to be.

matlab 😈 because I love figuring out how to vectorize code! I've used it for various art projects (see below) and it manipulates images beautifully. and the documentation and error messages are easy to follow.

I'm probably the black sheep here, but I love Kotlin. It has the best parts of strong typed, object oriented languages and functional languages. Though I feel like it being designed to be bytecode compatible with Java really limits its applications. Even though they have a scripting language version of it, it really doesn't perform well as a scripting language because you need to compile it. I find myself always using Python for scripts instead.

Kotlin: compatible with Java, JS, Wasm, Swift and even C++, ability to share code across all those platforms, way less verbose than Java, lots of useful functions in the standard library... I still don't understand why some people are saying it's just Java but worse

Currently Zig. It really is "better C", and i like C.

Otherwise it would be Erlang, but it does not suit what i want to do now.

Clojure. I not only get to use a functional language but also get to use all the libraries written for an actually popular language (Java, or JS for ClojureScript). Altough I'd choose something else if functional programming should ever catch on.

Scala. Multiparadigm. A touch of OO is nice in the functional world.

I try to stay language agnostic but if I'm honest JavaScript is my favorite because of the speed it provides. Also I like to build we based things so it's always in the stack

Perl. I can use it after awhile away without having to look up how to do things. It adapts to the best style for what I need to do.

No one else said it… I like Java, and more than the language all the tools available around it. They have been adding to the language to cut down on the traditional verboseness, and it can even natively compile now** some of the time.

The tools are also great, with Springboot for web services and jOOQ for databases, you can very quickly have a web app with strong typed database objects.

R. The Rstudio ide is awesome and the data wrangling packages are unmatched. It's also pretty fast as long as your dataset fits in your RAM.

Julia.

I haven't programmed anything for years and it was all self learned so I'll always have a soft spot for delphi.