Path: utzoo!attcan!uunet!snorkelwacker!tut.cis.ohio-state.edu!pt.cs.cmu.edu!n.sp.cs.cmu.edu!flc
From: flc@n.sp.cs.cmu.edu (Fred Christianson)
Newsgroups: comp.lang.c
Subject: Re: C strongly typed?
Message-ID: <8356@pt.cs.cmu.edu>
Date: 9 Mar 90 15:33:47 GMT
References: <259@eiffel.UUCP> <1990Mar1.172526.28683@utzoo.uucp> <849@enea.se> <2963@goanna.oz.au> <4401@hydra.Helsinki.FI>
Organization: Carnegie-Mellon University, CS/RI
Lines: 36

In article <4401@hydra.Helsinki.FI> grano@cs.Helsinki.FI (Kari Gran|) writes:
>a strictly (strongly? - not well defined terms anyway) typed language. 

Here's ONE POSSIBLE definition.  I'm not arguing that "strongly typed" is
well defined. I'm just giving one definition.

From Aho, Sethi and Ullman's _Compilers,_Principles,_Techniques,_and_Tools_:

	A language is strongly typed if its compiler can guarantee that 
	the programs it accepts will execute without type errors ...

	For example if we first declare
		
		table: array[0..255] of char;
		i: integer
	
	and then compute table[i], a compiler cannot in general guarantee
	that during execution, the value of i will lie in the range 0 to 255.

The C equivalent example

	char	table[256];
	int	i;

shows that C is not "strongly typed" according to Aho, Sethi, and Ullman.

My other compiler/language books don't even use strong in their discussion
of type checking.


MY OPINION: For any reasonable definition of "strongly typed" 
	    C is not "strongly typed".

----
Fred
p.s. Let's not argue about a reasonable definition for "reasonable".