Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!mailrus!tut.cis.ohio-state.edu!bloom-beacon!mit-eddie!uw-beaver!apollo!mrst!sdti!turner
From: turner@sdti.UUCP (Prescott K. Turner)
Newsgroups: comp.lang.misc
Subject: Re: The Joy of Zero-based Arrays
Message-ID: <213@sdti.UUCP>
Date: 4 Mar 88 18:07:00 GMT
References: <1012@PT.CS.CMU.EDU> <17419@think.UUCP>
Reply-To: turner%sdti@harvard.harvard.edu (Prescott K. Turner, Jr.)
Organization: Software Development Technologies, Sudbury MA
Lines: 26
Summary: One-based numbering is natural.

As someone said, people do 1-based counting.  (Sorry, I can't find the
reference to quote.)  A child learns to count by pointing at the objects 
and reciting the numbers, as "one, two, three".  The counting process
naturally labels the objects in a one-based fashion.  Are any of you 
zero-based partisans teaching your children to recite "zero, one, two"
and then deduce the cardinality to be three?  

I have seen pros and cons to zero-based arrays, and don't really mind
them in a programming language.  What I do mind is when it spills over
into user interfaces and even human discourse.  When the cursor is blinking
in the first column of my screen, my text editor should report the column 
number as 1, not 0.  (I know, my text editor is wishing it could put
the cursor just to the left of the first column, not smack on top of the
first column, but when it resorts to the latter real problems result.)

And in Dijkstra's letter in the June 1987 CACM, he numbers his points and
his references starting with 0.  I find this annoying and wrong-headed.

Since computer programs must relate to humans at some point, it is
desirable to stick with the human convention that when labeling things
by counting, you start with 1.  And this applies to arrays.

--
Prescott K. Turner, Jr.
Software Development Technologies, Inc.
375 Dutton Rd., Sudbury, MA 01776 USA        (617) 443-5779
UUCP:necntc!necis!mrst!sdti!turner