Re: [Caml-list] Future of Camlp4

[Daniel de Rauglaudre <daniel.de_rauglaudre@inria.fr>]

Hi,

On Fri, Oct 11, 2002 at 01:34:39PM +0200, Kontra, Gergely wrote:

> I agree, camlp4 IS useful. (Exploring the alternative syntax)
> I just afraid of developing in ocaml, if there exists two version of the
> syntax.

No: there is no two versions of the syntax, just one. OCaml has its
syntax, and it is the official one.

Let us compare with X window and the window managers. The X server
does not give by default any window manager: when you start X, you
just have a background and a mouse, and you can move it, that's all.
X can receive orders to create windows, move them, but by default,
it does nothing.

This is like the core of OCaml: the semantics. Now, it is impossible
to have a semantics without syntax. The same way, it is impossible to
have a system of windows without window manager. Hence, there is a
syntax, a given syntax. It could be compared with, say, the window
manager KDE.

Now, you man consider that KDE has many defaults. You may be
insterested in playing with "window managers", i.e. "syntax
tools". This is the gool of Camlp4.

OCaml does not need Camlp4, and it seems that the tendancy of the
Cristal team does not include experiments and developments about
syntax.

  ----

But Camlp4 can be useful even if you want to stay inside the official
syntax: you can do your small syntax extensions, you can use
quotations, you can use extensible grammars, all of that in the
official syntax. BTW, the manual and tutorial of Camlp4 gives its
examples in the official syntax.

The revised syntax, and, the Scheme syntax are just games with Camlp4.
Games or... research! We want to proove that many things can be done
with syntax. Perhaps, latter, a good consensus can happen with one of
the syntax Camlp4 developped. For the moment, it is not the case: the
OCaml team prefers keeping its syntax, despite its drawbacks that the
Revised syntax tries to fix.

I add that having its own syntx is not a problem of communication:
Camlp4 provides a pretty printer in the official syntax. You can
therefore understand the programs of the other people. And the
Revised syntax is close to the official syntax: you can read it
directly.

> Another thing, that bothers me is the do { } syntax. It seems a bit
> silly mixture of some shell and C syntax, I think either do ... done
> or { ... } would be a good choice (or support both, this way bash
> and C programmers will be happy ;))

Ha, if you are interested in the "Revised syntax", we can talk about
its choices, indeed. For the moment, I did not found people really
interested in making a "team" about a "New Revised syntax". The main
reason is that people are not shocked by the same things! We could
not know what are the points we want to talk about.

> Ooops, so I'd like to know what is the tendecy: will the alternative
> syntax be a new standard, or users should use the old syntax, and the
> alternative syntax supporting is their problem?

IMHO, the OCaml team is very very far from adopting a new syntax. But
using alternative syntaxes cannot be considered as a "problem" thanks
to the flexibility of Camlp4: I wrote GeneWeb entirely in Revised
syntax (45000 lines of code) and I am sure that it prevents nobody
to make changes in it.

-- 
Daniel de RAUGLAUDRE
daniel.de_rauglaudre@inria.fr
http://cristal.inria.fr/~ddr/
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners