Xref: utzoo comp.misc:4983 rec.humor.d:1578 rec.humor:18641
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!nrl-cmf!ukma!rutgers!njin!princeton!phoenix!pucc!BVAUGHAN
From: BVAUGHAN@pucc.Princeton.EDU (Barbara Vaughan)
Newsgroups: comp.misc,rec.humor.d,rec.humor
Subject: Re: Looking for Computer Folklore
Message-ID: <7129@pucc.Princeton.EDU>
Date: 8 Feb 89 18:11:36 GMT
References: <7143@pyr.gatech.EDU> <532@geovision.UUCP> <768@ur-cc.UUCP> <1012@ncrcce.StPaul.NCR.COM> <380@frksyv.UUCP> <7754@netnews.upenn.edu>
Reply-To: BVAUGHAN@pucc.Princeton.EDU
Organization: Princeton University, NJ
Lines: 18
Disclaimer: Author bears full responsibility for contents of this article

In 1972, I was assigned the task of writing an interactive user
interface for a statistical analysis program written in FORTRAN IV.
I was told that the users were "MBA types; not very quantitative and
with little background in statistics." ( I hope this is no longer
true of MBA's.) Anyway, writing such an interface in FORTRAN IV was
no picnic, but I tried to make it very friendly. Plain English
questions, examples of correct answers, range checks to determine
validity of responses, helpful error messages.  One of the first
users to test the program said that it kept bombing out on question
3. "Enter number of thingamabobs (Valid responses 1 to 5):". I
asked what her response had been and she said "Five". Puzzled, I
asked if I could watch her run the program. This is what I saw:
...(Valid responses 1 to 5): FIVE
That's when I realized what nonquantitative really meant. Even
though FORTRAN IV had no character string handling capability
(You had to declare your characters as INTEGER or REAL), I had to
write a routine to read all keyboard input as characters, convert
to numbers, and add a friendly message to explain what a number was.