Re: Historical CAS question

From: Koch-Beuttenmueller (Heike.Koch-Beuttenmueller_at_kiz.uni-ulm.de)
Date: 01/25/05 

Date: Tue, 25 Jan 2005 10:20:57 +0100

Richard Fateman wrote:
> r.e.s. wrote:
>
>> In 1980, in "The Mathematical Experience", Davis & Hersch wrote ...
>>
>> "It has been thought that a program like FORMAC
>> or MACSYMA would be an invaluable aid to the
>> applied mathematician. But this has not yet been the case, for
>> reasons which are not clear."
>>
>> With 25 yrs of hindsight, what were the reasons?
>>
>> --r.e.s.
>>
> In 1980...
> 1. Hostility of most mathematicians to computers.
> 2. Inconvenience/cost of the computing experience except for people
> deeply embedded in computing. (Formac was a batch program, mostly).
> 3. Inadequacy of the programming environment. Until 1978, Macsyma
> on a PDP-10 had 1.2megabytes of memory, and a speed of maybe 1megahertz.
We used first Reduce, then Macsyma and later other CAS-systems since
1976, not the mathematicians but physicists and chemists. The greatest
problems at that time were the lack of memory and speed of the computers
and the cost of the computers. The first nice interface for Macsyma we
had on the symbolics computers, but the price of such a workstation was
extremely high compared to now.

> 4. Poor user interface, documentation, bugs, and an incompatible
> world view between computer scientists and most potential users.
> 5. Lack of education in symbolic or even numeric computing
> in college curricula for mathematicians.
>
> Maybe others will have some suggestions?
>
> The question in 2005 is
> "It would seem that programs like Mathematica or Maple (or ...) would
> be an invaluable aid to the applied mathematician. But this has not
> yet been the case. Why?"
I think engeneers, physicists and chemists use CAS even more then the
mathematicians. In my opinion, this was the greatest problem for Axiom.
You need such a lot of mathematical knowledge and understanding to be
able to use it efficiently. And the pure mathematicians prefer pencil
and paper.
Heike
>
> I would answer 1,2,5, and to a lesser extent 4.
> regarding #5, my own institution, in its scientific computing
> course for engineers, still doesn't have any time for symbolic computing,
> so that Mathematica / Maple / etc appear only in optional
> calculus labs (or in my own graduate class).
>
>

Relevant Pages

  • Re: People are stupid
    ... >> Who says mathematicians need to know how to do arithmetic? ... Computers would more likely be designed for computing ... The Statistics Department is highly concerned ... are those of the Statistics Department or of Purdue University. ...
    (sci.math)
  • Re: Sieve distinction, prime counting
    ... Establishing the primality of X by computing the ... > prime versus letting the mathematical function find out by recursion. ... > Now in any other method that you will see mathematicians talking about ... in "Prime Numbers and Computer Methods ...
    (sci.math)
  • Re: People are stupid
    ... >>> Who says mathematicians need to know how to do arithmetic? ... the emphasis on computing throughout the educational ... are those of the Statistics Department or of Purdue University. ... Herman Rubin, Department of Statistics, Purdue University ...
    (sci.math)
  • Re: People are stupid
    ... > Who says mathematicians need to know how to do arithmetic? ... the emphasis on computing throughout the educational ... > of Purdue University. ...
    (sci.math)