Do you guys recommend Haskell for learning?

Another question, how fast is Haskell?

hax my anus

>>2
Haskell my anus

>>3
anus my anus

No; not very, until you need timing-critical capabilities, at which point it becomes very fast in bursts that come at random.

>>5 WTF?

Yes, Haskell is great for learning. I wish I had started with Haskell and not with the C++ turdmound. Fuck, thinking that programming is about pointers and message passing is a memory of hell to me. Haskell will let you actually program instead of all the bookkeeping shit in those archaic languages.

As for speed, GHC Haskell is on par with Java. Can't say anything about JHC, though.
You do have to worry about laziness, though. Gotta keep track of what gets evaluated when.

You should start with a simple assembly language like brainfuck.

Yes, learn Haskell even if you don't use it for money. An in-depth experience with Haskell gives you a mind-set understanding the benefits and implications of functional purity, static typing, and arguably-less-important-but-unique lazy evaluation. (I'm unhappy with every statically typed language including Haskell. Yes, I believe static typing is superior to dynamic, yet we still haven't achieved an amazing static language.) Carrying this mindset into whatever other language you use will boost your ability to break down problems and form proper solutions. Of course it's not a sliver bullet, but it's the closest thing we got to it. At the very least, it's training wheels for your programming skills.

Off the top of my head, a few cons for everyday use:
- Monad transformer functions are essentially method-functions in every OO language. However, methods are easier to write and maintain by comparison.
- Naming collisions. The record syntax often needs to be prefixed if the data type is (over-)exposed. It's quite annoying.
- And, fact that the 'lens' library exists and is immensely popular implies that some things in the language are broken.
- Historical baggage. For example, map should be a function under Functor. fmap, lift(s), <$>, etc. should be removed.

Having said that, I find myself programming in dynamic languages most often because they're cleaner and easier syntactically and unnecessarily-flexible implying future-proof for non-invented concepts. Thinking in Haskell, let's me approach this cautiously.

>>10
and arguably-less-important-but-unique lazy evaluation.
One thing that trying to write actually useful Haskell code would teach you is that lazy evaluation by default sucks. You can learn the same lesson by abusing generators in Python, and it would be a better way because your code that uses laziness promiscuously just wouldn't work instead of working fine on small datasets.

For example, map should be a function under Functor. fmap, lift(s), <$>, etc. should be removed.
Most <#$%^#> operators should be removed. The raw typing speed or character-by-character reading speed doesn't matter for anyone except people new to the computorer. Write that shit out using human-readable names!

>>11
Yes, operators are bad! Different names for each function is good! add_float! add_int! I can just taste the tears of the goyim!

>>12
Go indent another exclamation point you subhuman slav faggot.

>>12
Yes, operators are bad! Different names for each function is good! add_float! add_int!
Don't be obtuse, goy (if you can help it).

Of course basic arithmetic operations on numbers, vectors, matrices, reified functions deserve to be operators.

The proliferation of &<#$%^>& in Haskell goes far beyond that though.

My point is one of moderation, not one of "No operator overloading is allowed" or "Overload everything you can as =<^_^>=".

Moderation is hard, I know. Especially for people with ASD. How can one consider implications of a choice outside of a rigid framework of rules that orders every weird thing of this world?

You people should wear an autist badge on the internet or something. So that we neurotypical people don't waste our time recognising your bullshit as being autistic (we can do that, sure, but it's distracting).

bump