Re: [f-cpu] Winograd DCT on my seul.org account

["Christophe" <christophe.avoinne@laposte.net>]

CAML combines imperative and functionnal programming. In a certain way, LISP or
C programs can be easily ported to CAML. There is another project COQ, which
based on CAML and have the proprietary to check if an algorithms is correct
(I'm speaking about the correctness of an algorithm, something that C is unable
to check). We don't also need to use YACC and LEX with CAML because it can also
be used as a parser.

Please don't think we should use CAML :). I'm just telling you that CAML is
more than a simple language.


----- Original Message -----
From: Juergen Goeritz <goeritz@oekomm.de>
To: <f-cpu@seul.org>
Sent: Monday, April 22, 2002 1:39 PM
Subject: Re: [f-cpu] Winograd DCT on my seul.org account


> On Mon, 22 Apr 2002, Marco Al wrote:
> >Juergen Goeritz wrote:
> >
> >> What one would need is a language with a very good
> >> type checking like in Ada, Pascal, Modula a.s.o, a
> >> language that can do easy prototyping and reuse of
> >> coded parts like Forth, a language disallowing the
> >> side-effect programming styles, a language that can
> >> handle OO approaches and a language that is capable
> >> of parallel execution pathes.
> >
> >How about Occam? ;)
>
> I know occam from the time I looked at transputers. The
> language didn't really turn me on. Everything felt too
> complicated. Maybe that was one thing why the transputer
> never made its way to a real hit (beside being much too
> expensive).
>
> >As far as type-checking/encapsulation/concurrency goes it seems to
> >qualify, and functions are ensured to be side-effect free at least  (I
> >think that hits the sweet spot, imperative programming is here to stay).
>
> Yes, and you also find this in a lot of other languages...
>
> >It obviously lacks some basic features, foremost handling reference's
> >and OO ... but an implementation for zero-aliasing references was
> >introduced not so long ago, and and some interesting proposals for OO
> >extensions have been made by others (the distance between message
> >passing and prototype based OO is not long).
>
> Have these proposals made the way to implementation yet?
> Is there a PD compiler to start with?
>
> BTW I spent years on message passing and I don't share
> your opinion about that 'not long' distance to OO.
>
> JG
>
> *************************************************************
> To unsubscribe, send an e-mail to majordomo@seul.org with
> unsubscribe f-cpu       in the body. http://f-cpu.seul.org/

*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/