Path: utzoo!attcan!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!julius.cs.uiuc.edu!psuvax1!psuvm!uh2
From: UH2@psuvm.psu.edu (Lee Sailer)
Newsgroups: comp.sys.amiga
Subject: Re: Scientific programs on Amiga
Message-ID: <90273.105607UH2@psuvm.psu.edu>
Date: 30 Sep 90 14:56:07 GMT
References: <31796@nigel.ee.udel.edu>
 <1990Sep29.021658.26896@evax.arl.utexas.edu>
Organization: Penn State University
Lines: 55

In article <1990Sep29.021658.26896@evax.arl.utexas.edu>,
hill@evax.arl.utexas.edu (Col. Ames and Pixel) says:

>    My wish list:


> 1) Plotting program. Ability to plot xy, scatter, polar, parametric. Plot as
>      wire frame and ability to rotate (ala Calligari, and Lightwave with a
>      cube, rectangle etc...) around all axes. Ability to figure volumes of
>       closed solids, limited symbolic math, (find intersection of planes in
>       y=mx+b format or others, min/max of f(x)'s.


Isn't there a freely distributable package that does some of this stuff?
It writes HPGL output, and includes a way to send HPGL to any preferences
printer.  I think it is called MultiPlot, or PLPLOT:, or some such, or
maybe that is two different packages altogether.

Perhaps the real problem is that it isn't commercial enough!  There needs
to at least be that little ad in the back of Amiga World every month to
help people find it.  Better yet, the ad should be in Science.


> 2) Vector Math - Simple 3D static systems. Normal finding etc...

MatLab, from some gov't agency or other, is there.  Again, maybe the
problem is that it isn't commoercial *enough*!  Most people don't know
it exists.

Also, Maple is available from U. Waterloo.  Perhaps it isn't as sexy
as Mathematica, but it flies OK.


My wish list includes

a)  A statistical analysis program like Systat or SAS.

b)  A lot of small niche special purpose stuff that my social scientist friends
    have written over the years for cluster analysis, blockmodeling, multi-
    dimensional scaling, chorpleth mapping (spelling?), and content analysis.
    Almost of all of this software has been written for msdos, because
    that's what most social scientists have.

Question:  Is there a "market" for this stuff?  I don't mean a million copies
per year, just enough to make it worth while?  Would you buy a typical
Nonmetric MDS program for $79?  Or a collection of half a dozen typical
Cluster Analysis programs for $89.50?  It would be easy to find lots of
such stuff and port it to the Amiga, but a giant headache to deal with
customers and so on.

Amigifying it all would be a big project, since it is almost all written
in a sort of 1960's style batch oriented Fortran.   No indentation, no
interaction, fixed width input, no comments, 8-(

                                                     lee