I honestly don't like Haskell. I've read a lot about the language, I understand Functors, Monads and all the misused Category Theory concepts sprinkled throughout the language I've come into contact with. I've comfortably read the online tutorials, some of the books and even a few academic papers. Conceptually, it's at least a little interesting.
Quite frankly though, "getting shit done" in Haskell is painful. I want to vomit every time I have to look up the details of something stupid like IORef. Oh, it's dead simple to understand (
http://cvs.haskell.org/Hugs/pages/libraries/base/Data-IORef.html), but it's yet another stupid hoop to jump through to do something really simple. "But the referential transparency, the purity, lOL!1" You know what? I really don't fucking care. I mean that honestly. I don't care that you've made a magical clean sandbox (built upon a framework of rotten wood and rusted metal, I might add) where doing anything harmful isn't possible, except it is, but in a really clean (read: painful) way.
1. I understand the separation of code with side effects and code without (I've done that for years anyway). I still don't care, because the language I'm looking at is a complicated mess. Slapping some mathematical names
2 on a sloppy hierarchy of mismanaged concepts is not proper language design.
What
really pisses me off, however, are all of the dropout hipsters who don't have the discipline or intelligence to do "real" theoretical mathematics who evangelize the language. "It's so pure, it makes my code like pure mathematics", they say. Guess what, you assholes, programming and mathematics aren't even philosophically the same thing
3. Please sandwich your
Foster the People EP inside your copy of
The Haskell School of Expression and shove the whole festering mess up your ass.
To me, Haskell is a lot like C++. I have a reasonable amount of knowledge about it, and I can sometimes get what I want done. However, every project involves wading through a morass of complicated bullshit that makes solving a problem unnecessarily hard because the designers of the language found one concept they liked and then took its application to an absurd extreme (but didn't bother to think about the poor sod who might have wanted to use the language for something other than a Master's thesis).
And you can't say any of this to anyone in public, since speaking out against Haskell with your name attached is grounds for immediate crucifixion by the Kool-Aid drinkers (who themselves can barely write enough Haskell to do anything useful, but are still convinced it's,
liek, the totaly smertest thing evar). Actually, the reality is that those who don't like Haskell are "dumb." It's unfathomable that anyone wouldn't want to join the strange cult that's gathered around a sterile academic language.
In short, I have learned things while playing with Haskell, and it probably has influenced my coding style at least a little. But overall, I find the claims of Haskell's practicality to be grossly exaggerated, and the language always ends up leaving a bad taste in my mouth. If you use Haskell and like it, more power to you, but leave me the fuck alone and stop yelling at me when I don't use your favorite language.
1 Don't reply reminding me that "LOL, A MONAD IS JUST AN ABSTRACTION OF A COMMON PATTERN OF COMPUTATION", because I understand that. What I dislike is how the line of reasoning used to design this abstraction has left us with a complicated, nasty mess.
2 For anyone living under the impression that Haskell's use of category theory is mathematically perfect or even sensible, I don't recommend looking too closely, or you might end up realizing how cheesy and stupid it really is.
3 Don't throw technical definitions at me after you read that, because you've only skimmed Wikipedia to find them, I already know them, and I'm talking about the process of development anyway.
Paul Graham footnote POWER FORMATTING