[plt-scheme] Re: Scheme sources readability

2008/9/8  <hendrik at topoi.pooq.com>:
> On Mon, Sep 08, 2008 at 07:56:39AM +0300, kbohdan at mail.ru wrote:
>> Richard Cleis wrote:
>> <snip/>
>> >>Without pattern catalogue i can see the only way to become good scheme
>> >>programmer : traverse tons of web links and articles without any idea
>> >>what is good in practice and what is just mathematic experiment.
>> >>And that is what i'm currenly doing :)
>> >
>> >What have you found that is not good in practice?  How was it merely a
>> >mathematical experiment?
>> >
>> >RAC
>>
>> Many things.
>>  For example famous "amb" from SICP. It looks great, but i haven't seen
>> something like this is used in practice.
>>  Other example are monads which look great and promising, but people
>> say that in scheme they are "not big fun".
>
> Monads were invented so that you could do imperative stuff (at least
> top-level) within a purely functional language, and the theoretical
> combinator-calculus coding was to justify that it was, in some
> theoretical sense at least, still functional.
>
> Scheme is not purely functional, it has side-affecting
> operatins, and has no need for the monads.

There are other types of monads in Haskell for doing different things.
See the 'Maybe' monad for an example.

    Henk