Re: [Caml-list] LLC book [was: Questions]

["Damien Guichard" <alphablock@orange.fr>]

[-- Attachment #1: Type: text/plain, Size: 2389 bytes --]


It's far more than just about human ressources.

What you describe is some sort of "OCaml Bible".
First you have to decide what book you want.
If you do an OCaml Bible then probably some beginners will lack an introductory text.
If you do an introductory text then probably more experienced people will lack a reference book.

As i remember it Le Langage Caml was largely about compilation techniques.
That's yet another choice:
* do you want to introduce OCaml where it shines best, as an rewriting tool ?
* or do you want to be more general and application-agnostic ?

Certainly at some point you have to speak about semantic, so :
* do you want to be practical and omit the boring theory ?
* or do you want to speak about lambda-calculus, type inference and value restriction ?

Do you want a full exposure of the module language ?
Or just the bits that allow usage of the standard modules ? 
Do you want to present advanced usage of polymorphic variants ?
Or do you want to just point the right articles and let the reader perfect its knowledge ?

Certainly you also want to discuss different programming styles.
Moreover it's much better, especially for library designers, if you introduce some functional niceties
(maps, folds, continuations, lazyness, monads...).

That's many conflicting questions.

The reason why you overlook them is because you greatly overestimate the importance of libraries. 
Even from the library point of view, it's better if the langage comes first, it encourages a much more elaborated design.

-- damien





Damien Guichard
2009-04-02



En réponse au message
de : Alp Mestan
du : 2009-04-02 15:40:45
À : Jon Harrop; caml-list@yquem.inria.fr
CC : 
Sujet : Re: [Caml-list] Re: LLC book [was: Questions]

Indeed, as I said above, a book on "today's OCaml" should cover a wide variety of topics (syntax extensions with camlp4/5, general purpose libraries, specific libraries like ocamlnet, GTK+ and OpenGL binding, etc). To write such a book, there would be the need for many authors with time and knowledge to produce a good learning and practice material for OCaml !

I think many of us would enjoy writing some paragraphs for such a project... But would there be enough people to achieve the writing of an entire (and good) book ?

-- 
Alp Mestan
In charge of the C++ section on Developpez.com.

[-- Attachment #2: Type: text/html, Size: 6685 bytes --]