Xref: utzoo comp.object:3437 comp.lang.misc:7699 comp.lang.eiffel:1566
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!udel!haven.umd.edu!socrates.umd.edu!socrates!rockwell
From: rockwell@socrates.umd.edu (Raul Rockwell)
Newsgroups: comp.object,comp.lang.misc,comp.lang.eiffel
Subject: Re: A Hard Problem for Static Type Systems
Message-ID: <ROCKWELL.91May2083101@socrates.umd.edu>
Date: 2 May 91 13:31:01 GMT
References: <1991Apr20.010347.28984@leland.Stanford.EDU> <554@eiffel.UUCP>
	<1991May1.194620.1141@leland.Stanford.EDU> <566@eiffel.UUCP>
Sender: rockwell@socrates.umd.edu (Raul Rockwell)
Organization: Traveller
Lines: 31
In-Reply-To: bertrand@eiffel.UUCP's message of 2 May 91 06: 03:19 GMT

Bertrand Meyer:
> But the reverse is not true: if you have a dynamically typed language,
> you will *never* be able to write a type checker for it because it
> would lack the necessary information. 
...
> In other words, I don't think, as Mr. Chambers does, that the static
> vs. dynamic debate is a conceptual discussion at all.
> Conceptually, static wins hands down every time.

I disagree.  Completely.

If the language includes the empty function (one which can not be
closed -- any value applied results in a "domain error" or "message
not understood"), you have a very clear mechanism to express type
errors.  (any predicate which allows application of that function...)

The main issue I see between static typing and dynamic typing is that
static typing is, well.. static.  The difference between static typing
and dynamic typing is very analogous to the difference between
evaluation to a constant, and evaluation to a function.

Of course, what I call "dynamic typing" others may call "static
typing" -- I'm still pondering the question of user defined types.
[With sufficiently powerful primitive types, and sufficiently powerful
function re-write capability, what does user defined typing buy you?]

[[[ Actually, I'm struggling with some specific models for typing.
Basically, I just haven't found a way of expressing user types that is
as expressive as I'd like.  Yet. ]]]

Raul Rockwell