What went wrong? Why doesn't anybody use it unironically?

>>59614524
Plenty of top programmers use functional languages, especially for prototyping stuff rapidly in Lisp / Python / Haskell

When it needs to be mass produced and sucked through the normalization field, that prototype gets converted into Java by code monkeys.

>>59614567
Yeah you'll never be a code monkey, you'll show 'em!

>>59614567
Except I've never seen a single line of functional code in the real world in my entire life, and literally every mainstream piece of software you can think of is not written functionally.

>>59614594
>>59614596
>they don't prototype stuff directly on an FPGA

should have worked harder in post grad, top shit, mates

>>59614524
Why do functional languages always use garbage collection?

I mean if the whole program is just one big nested function call, can't you just allocate everything on the stack?

Better yet, since linked lists are such a huge deal in functional languages, use a linked callstack. That way it never has to overflow

>>59614647
>Why do functional languages always use garbage collection?
The real question is: how does the garbage collector manage to not collect the language itself?

>>59614524
I'll use functional languages when they:
>use the parenthetical structure of the language to their advantage to not need a garbage collector
>have some nice imperative-like syntactical sugar for certain common functions instead of just everything being a function call
>become compiled languages (since in these languages functions are closures that store dynamic context information, they can be compiled into subroutines that accept context pointers as hidden parameters, in the same way that methods have a hidden "this" parameter; if eval MUST be used (ugh) it can be implemented as an embedded compiler whose input can be provided by the already running program and whose output can be called by said program and accepts the current context as the context pointer parameter)

>>59614647
If you're on Linux, you can just map 8GiB to the stack and it won't actually get used because of overcommit.

>>59614664
Haha nice one, have my upvote ;)

>>59614647
>Garbage collection is bad meme
Welcome to 2017, where there's no problem using a garbage collector.

>>59614596
Counter anecdote: I see it all the time and where the programmer discerns that stateful operations only occur during input or output of data at during the beginning or end of the computation, it's considered good practice to program in a functional style to avoid errors.

>>59614524
Because managing just a couple layers of functions become a nightmare when working on a wide code base. Also functional languages can't easily be transformed to handle objects while you can always handle some functional code in oop languages.

>>59615125
Except almost nothing is purely functional in nature. There is always some states that need changing. Unless you want to wrap every outside interaction that can change state into another function..

>>59614664
Have my upvote too!

>>59614524
People use patterns from functional languages, mapreduce for parallel computations for example.

4chan consists of pajeet schoolkids, they want to look like a top-tier code monkeys that's why they pretend that they use specific languages.

>>59614807
http://www.sbcl.org/manual/#Compiler_002donly-Implementation
Welcome to functional languages (lisp is not pure functional tho, you can write your entire program imperatively).

If only functional languages didn't have garbage collection and a huge runtime

>>59614524
I only see functional languages in grad school, nobody here is doing anything imperative anymore in research because proving your work these days is critical to getting grant money. No verification of correctness no grant.

Functional is actually the most intuitive, easiest way to program because it just picks up where you left in highschool algebra there's almost no difference. This follows CMU's curriculum somewhat if anybody is interested https://functionalcs.github.io/curriculum/

for research purposes.
jesus you enterprise cucks need to die.

>>59614664
>heartyChuckle.png

>>59615408
what does algebra have to do with usb drivers?

>>59614647
Rust doesn't

>>59615785
Everything especially if you're formally verifying the driver's specification w/Dependent Types.

Btw Microsoft verified the majority of their drivers to prove they would not crash the kernel, same with Apple, same with everybody these days.

>>59614524
too abstract

>>59614524
It just takes a bit of mental effort to wrap your head around how to use the languages, so most people don't bother.

>>59614807
They are compiled languages since the 70s, idiot.

>>59615886
Indeed it doesn't. It's also dogshit.