Xref: utzoo comp.lang.modula2:670 comp.lang.misc:1123
Path: utzoo!mnetor!uunet!husc6!necntc!ima!haddock!karl
From: karl@haddock.ISC.COM (Karl Heuer)
Newsgroups: comp.lang.modula2,comp.lang.misc
Subject: Re: 0-based/1-based arrays
Message-ID: <2851@haddock.ISC.COM>
Date: 3 Mar 88 02:45:00 GMT
References: <7161@sol.ARPA> <2740@mmintl.UUCP> <4343@june.cs.washington.edu>
Reply-To: karl@haddock.ima.isc.com (Karl Heuer)
Organization: Interactive Systems, Boston
Lines: 18
Summary: rebuttal from a 0-based user

In article <4343@june.cs.washington.edu> pattis@june.cs.washington.edu (Richard Pattis) writes:
>I too detest 0-based arrays (and think 1-based arrays are better).  In trying
>to rationalize this choice, I came up the the following argument: When we
>process arrays, we often need to know the size of the array and/or the
>position of the last element.  In 0-based arrays, these quantites are off by
>one; in 1-based arrays they are the same.  Hence, I think that people will
>make fewer errors with 1-based arrays.

When dealing with 0-based arrays, the appropriate concept is a half-open
interval: the index for an N-element array comes from [0,N).  I find that this
makes things simpler than with 1-based arrays, because half-open intervals can
be abutted without having to adjust by one.  (I.e., [0,N) U [N,P) = [0,P).)

I very seldom need to know the position of the last element.  Instead, I use
the position of the first element beyond the array.  Always use postincrement
and/or predecrement to traverse it; everything works out fine.

Karl W. Z. Heuer (ima!haddock!karl or karl@haddock.isc.com), The Walking Lint