Are functional programmers the most entitled idiots on the planet?
They say "muh abstraction"
But what they really mean is "muh refusal to interface with the way actual computer hardware works"
I think most of them read the first couple of pages of SICP and actually walked away thinking that computer science is magic, where you can just say something and expect the compiler to make it so.
It's no wonder all FP code runs like ass. Even Java is faster
Yeah it's like starting a restaurant with someone else's money even though you have no idea how to run a restaurant
>>55488147
this actually happened in one of Ramsay's Kitchen Nightmare episodes
>>55488222
wat happened in the end?
>>55488232
The episode ended
>>55488232
Ramsay fixed their shit just like he always does
http://www.realitytvrevisited.com/2011/11/us-season-5-burger-kitchen.html
here's the full thing, if you don't care for watching it
>>55486523
ah, it's you again, the anti-SICP moron.
0/10, homeboy.
>>55486523
Do you program in assembly exclusively? If you don't, you use a language where you "just say something and expect the compiler to make it so".
Also, who cares about speed? Maintainability, readability and correctness are far more important things to consider. 99% of code doesn't need to run fast. If you're working on the 1% of it that does, you'll know. Besides, except for a couple of terrible interpreted languages, you can make almost everything run fast enough (including Haskell and OCaml, for example).
What is the point of this post anyway? What real "FP code" have you run and how did it "run like ass"?
functional programming is shit
lambda calculus is better
still saging though
>>55486523
While it's true that Hasklel that trades one category of errors for another and Lisp with its muh AST is data are defective memes and the ML family lacks a good language standard and implementation, your statement about abstraction and Java shows you don't know what you are talking about, fag OP.
Just check out Elm, to see how FP language is done right, because the devs know what's important.
Or Scala, once dotty is ready and their main cum piler - maybe in a decade or two.
FP is a stale meme. All cool guys are into machine learning nowadays.
>>55488147
No, it's like going to a restaurant when you don't how how to start a restaurant.
Whatever happened to the "Haskell is deprecated" shitposter?
>>55486523
Common Lisp is pretty fast with SBCL. But it doesn't hold your hand or constrain you unlike other functional languages. Functional programming is just one thing you can do with it. Most of it revolves around mutative operations.
>>55486523
brainlet detected
>>55489160
CL under SBCL is bleedingly fast. If you know how to use the type/compiler promise system to annotate the code.
Haskell is also profoundly fast.
You could view Haskell as a declarative, functional version of high-level C++ if you wanted to.
The actual tradeoff with Haskell is not efficiency but up-front design costs in terms of time and effort. You get paid back in bug-free AND fast code but it is a cost nonetheless.
>>55488652
>hey forget about those old memes guys, check out these new meme! they're way better and I'm not at all a trend hopping faggot who'll be talking up some fresher meme a year from now
>>55489325
Well, unlike the old memes the new ones at least are thought through to some point.
>>55486523
>>55489319
>bleedingly fast
>profoundly fast
>>55489476
go away pajeet
>>55489319
>Haskell is also profoundly fast.
No, it's not. Haskell is only fast when it's used as a DSL to write C with the FFI like in the Debian benchmarks game. Idiomatic Haskell is very slow.
Haskell faggots used to say that Snap and Yesod were blazing fast until real benchmarks came out showing that they're both basically Ruby on Rails tier. They were so embarrassed that they couldn't make Haskell competitive that the Yesod authors had it removed from Techempower's benchmarks. Snap is the only Haskell framework with the balls to compete and you can see that it's right around Ruby on Rails.
https://www.techempower.com/benchmarks/#section=data-r12&hw=peak&test=query
What about Elixir?
>>55489830
meme to the extreme -tier
never use completely useless, zero adoption languages like elixer, nim, kotlin, etc
>>55489476
Hindu nuffin excuses
>>55486523
Computer science is magic. We can just say something and expect the compiler to make it so. Now if you expect computer science to make a tradeoff between theoretical purity and hard resource book keeping, academic computer scientists will tend towards theoretical purity.
>>55488566
>Haskell and OCaml
please don't talk about ocaml in the same sentence you talk about your memelanguage, it's above that. thanks
>>55489830
I like elixir mostly because of phoenix framework.
>>55486523
Were you raped by a first class anonymous function, or were you gangbanged by hygienic macros and continuations while TCO filmed it?
>>55489476
>javascript
>clean and easy to read
my face is undefined
>>55486523
Look at this and tell me Java is faster than FP.
Undereducated faggot pajeets. http://pastebin.com/2djVDiUq
>>55486523
Object-oriented code is at least as "magical" as functional code. Try implementing a full-featured object system in assembly, or to be charitable, try reimplementing C++ in C.
But the "way actual computer hardware works" is that a flip-flop is a complicated delay line (a bistable multivibrator), and nearly everything else is implemented with multiplexers. If you're even using an assembly language, you're still miles away from what hardware really does.
>>55486523
OOP is as abstract as the functions - except it's way, way more messy and error prone. It's simply easier to make mistakes in it due to the fact that it does not set any bounds to the way you program. It's the same as trying to get to a destination with and without followind a road.
Also, the fact that the imperative programming makes you tell the computer HOW to do a certain task, instead of WHAT to do makes you repeat yourself a real fucking lot of times.
Don't you feel that MAYBE, just maybe, there's a better way to do something for every element in a list than the old fuckingfor (i=0; i<array.length; i++) {
...
}
? Don't you get tired of writing that shit a thousand times a day?
Programming exists to interface a human to a computer, not to make the said human repeatedly write always the same stuff on text editor. Optimizing is the compiler's task, not the programmer's.
Also,
>FP is slow
FP is a paradygm. Haskell is a language.
Haskell is not as fast as C++, but it comes pretty close to it.
https://benchmarksgame.alioth.debian.org/u64q/compare.php?lang=ghc&lang2=gpp
And, if a language as high level as haskell dares to challenge C++, one of the hardest languages out there, in the matter of SPEED - well, allow me to say that maybe we've got some fucking problem here.
>>55490055
yeah, just use c amirite
>>55488370
>Homeboy
Fuck off nigger
>>55488930
Bloody hell m8, ever since those hipster news sites started writing about "NeURal n3tworks lerns how 2 play MARIO oMg!!!1!", an influx of retarded memefaggots have been trying to pollute the field. Fortunately, ML is very heavy on math and algorithms, so most of them give up rather fast.
