>>42
Thank you for your explanation, but i know that already. The reason why i still don't use haskell is that monads are said to do a lot more than they really do. Especially because they are called Monads and that they are used to encapsulate side effects. Call them wrappers, containers, templates, whatnot, but the name monads in itself has no apparent meaning and should go away.
And about how they encapsulate side effects (As far as i understand). Monads either force the compiler to execute a function with side effects exactly the number of times we want it and no known (or UNKNOWN) optimization can change the order or number of times or they are broken. But i could not find any article describing which laws of mathematics would enforce monads to act how people claim they do. And we talk about laws which can be enforced in a pure functional language using only the core concepts which weren't written to explicitly support monads, because the developers of haskell claim, that they didn't need to add special support for them.
If i think about it, we get several possible cases:
1. There are some rules which make monads possible in the most efficient manner without bad effects on the other parts of haskell in terms of run speed, stability and usability.
2. it is not possible to efficiently optimize haskell code, because the used laws are to strict, else they could not support monads and hence functions with side effects.
3. Monads are broken.
4. Haskell is not a pure functional language.