NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/comp/lang/c/11704

Text File | 1992-07-29 | 1.6 KB | 41 lines

Newsgroups: comp.lang.c
Path: sparky!uunet!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!malgudi.oar.net!caen!destroyer!gumby!yale!mintaka.lcs.mit.edu!hal!mjd
From: [e-mail] (Sievan Janacziewski)
Subject: Re: Pointers outside of array boundary
Message-ID: <mjd.712419017@hal>
Originator: [e-mail]
Sender: [e-mail]
Organization: MIT Laboratory for Computer Science
References: <[e-mail]>
Date: Wed, 29 Jul 1992 14:10:17 GMT
Lines: 28

[e-mail] (M. J. Saltzman) writes:

>    float *x = malloc(10*sizeof(float));
>    --x;

>which is supposed to give an array whose elements are addressed
>as x[1]..x[10].
>My question is: Can anyone name an existing machine on which this code
>could fail, or give a really convincing hypothetical argument why it
>is a bad idea?  My interlocutor claims the criticism of this form is
>legalistic, and not based on practical considerations.

I was going to make up some thing about how maybe the array gets
allocated at the top of memory, but there's a much better reason not to
do this: C arrays always start at 0, never at 1, and by resorting to
cute compiler hacks to fudge up the appearance of a 1-based array, one
will only confuse oneself and other people who have to read,
understand, maintain, and modify the code.

If 0-based arrays are unclear to your correspondent, then he or she
should practice more, so that they become more familiar, rather than
resorting to obfuscation to cover his or her inadequacies as a programmer.

This has nothing to do with legalism.
It is the most practical of all possible considerations.

Sievan Janacziewski
[e-mail]