From jwagener@trc.amoco.com Sat Sep  5 01:45:22 1992
Received: from noc.msc.edu by dkuug.dk via EUnet with SMTP (5.64+/8+bit/IDA-1.2.8)
	id AA28459; Sat, 5 Sep 92 01:45:22 +0200
Received: from uc.msc.edu by noc.msc.edu (5.65/MSC/v3.0.1(920324))
	id AA23002; Fri, 4 Sep 92 18:45:24 -0500
Received: from [129.230.11.2] by uc.msc.edu (5.65/MSC/v3.0z(901212))
	id AA12404; Fri, 4 Sep 92 18:45:31 -0500
Received: from trc.amoco.com (apctrc.trc.amoco.com) by netserv2 (4.1/SMI-4.0)
	id AA26580; Fri, 4 Sep 92 18:45:00 CDT
Received: from crmac1 by trc.amoco.com (4.1/SMI-4.1)
	id AA23637; Fri, 4 Sep 92 18:44:56 CDT
Date: Fri, 4 Sep 92 18:44:56 CDT
Message-Id: <9209042344.AA23637@trc.amoco.com>
From: Jerrold L. Wagener <jwagener@trc.amoco.com>
To: SC22WG5@dkuug.dk
Subject: a question of values
X-Charset: ASCII
X-Char-Esc: 29

To X3J3 and WG5 -

Here's a light-hearted story to wrap of this week and take us into the 
holiday (in the US) weekend.  The values here are the (sort of boring) 
Fortran kind, not of the ilk being discussed in the US presidential race.

I decide to teach a Fortran course (actually a beginning course in "Principles
of Scientific Computation" - notice I spelled "Principles" right, I think) 
this term, and the second day of classes I did a simple example of numeric 
computation involving real values - computing the areas of certain shapes.  
One of the shapes was a circle, area = PI*R*R.  I duly initialized PI with 
the value 3.1415926  (that's all the digits of pi that I remember, but it was 
only single precision anyway).  I even ran the program before hand to make 
sure there was no mistake lurking somewhere - I had all the bases covered.

So here's the question I get:  
"Doesn't Fortran know about pi?" 
        <<the premier scientific language of the 1990's, right?>>
        <<even $20 calculators know pi out to 10 digits, right?>>

gulp..  <<watch the Fortran expert backpedal like mad>> 

The first flashes through my mind were that Fortran does array processing 
and data abstraction and ...        <<but it doesn't know about pi>>

The second flashes were that of Taylor series, then 2*ASIN(1.0) or 
ACOS(-1.0), both of which I've now confirmed do indeed give 3.14159 
on my system.  Or maybe better, 2*ASIN(1D0) or ACOS(-1D0), which 
do indeed both give 3.1415926535898.  But also in a flash I decided
it best not to "enlighten" these second day students with
      PARAMETER ( PI = ACOS(-1D0) )
so my answer was a sort of wimpy "no".  As somebody commented in a
meeting I attended today, "it's the o-rings that kill you".

Y'all have a nice weekend.

Jerry