[Caml-list] "Modules Matter Most"

[Caml-list] "Modules Matter Most"

[oliver]

Hello,


this article I like.
I also have experienced, that thinking in types
(or: thinking in signatures) give a big advantage
on planning a program, and later implementing it.

Types make things very clear: you know what is input
and what output and how things need to fit tigether,
and what kind of signature is needed to maybe fill
a gap between an input and an output module... just look
at the missing link as a signature.

Thie article also mentiones modules as a key point
of OCaml, and I fully agree here:

  Modules Matter Most
    http://existentialtype.wordpress.com/2011/04/16/modules-matter-most/


Ciao,
   Oliver

Re: [Caml-list] "Modules Matter Most"

[AUGER Cédric]

Le Mon, 27 Feb 2012 15:38:09 +0100,
oliver <oliver@first.in-berlin.de> a écrit :

> Hello,
> 
> 
> this article I like.
> I also have experienced, that thinking in types
> (or: thinking in signatures) give a big advantage
> on planning a program, and later implementing it.
> 
> Types make things very clear: you know what is input
> and what output and how things need to fit tigether,
> and what kind of signature is needed to maybe fill
> a gap between an input and an output module... just look
> at the missing link as a signature.
> 
> Thie article also mentiones modules as a key point
> of OCaml, and I fully agree here:
> 
>   Modules Matter Most
>     http://existentialtype.wordpress.com/2011/04/16/modules-matter-most/
> 
> 
> Ciao,
>    Oliver
> 

I find the posts on this blogs too much Ocaml biased. Ok, haskell has
not modules and it is somewhat missing, but using TypeFamilies, you can
emulate modules (and even 1st order modules which are available in Ocaml
only since 3.11 or 3.12), and in the other hand there is still not yet
any convenient way to have type classes in Ocaml. And it is not very
often that I would like to implement a module where a simple type class
wouldn't fit.

Are type classes planned in the future (and if not, what are the
reasons not to have it)?.

Re: [Caml-list] "Modules Matter Most"

[Abdallah Saffidine]

[-- Attachment #1: Type: text/plain, Size: 809 bytes --]

2012/2/28 AUGER Cédric <sedrikov@gmail.com>

>
> I find the posts on this blogs too much Ocaml biased. Ok, haskell has
> not modules and it is somewhat missing, but using TypeFamilies, you can
> emulate modules (and even 1st order modules which are available in Ocaml
> only since 3.11 or 3.12), and in the other hand there is still not yet
> any convenient way to have type classes in Ocaml. And it is not very
> often that I would like to implement a module where a simple type class
> wouldn't fit.
>
> Are type classes planned in the future (and if not, what are the
> reasons not to have it)?.
>

I guess you might be interested in the following article ML Modules and
Haskell Type Classes: A Constructive
Comparison<http://www.springerlink.com/content/m7v201w152h03670/>

Abdallah

[-- Attachment #2: Type: text/html, Size: 1091 bytes --]

Re: [Caml-list] "Modules Matter Most"

[Milan Stanojević]

> I guess you might be interested in the following article ML Modules and
> Haskell Type Classes: A Constructive Comparison

free version:
http://www.cse.unsw.edu.au/~chak/papers/modules-classes.pdf

Re: [Caml-list] "Modules Matter Most"

[AUGER Cédric]

Le Sun, 4 Mar 2012 21:25:21 -0500,
Milan Stanojević <milanst@gmail.com> a écrit :

> > I guess you might be interested in the following article ML Modules
> > and Haskell Type Classes: A Constructive Comparison
> 
> free version:
> http://www.cse.unsw.edu.au/~chak/papers/modules-classes.pdf

Thanks, although in fact I already read it (it was on the basis of that
article, that I sent my previous post).