Re: University License fees are short sighted of Wolfram Research

From: Dave (nospam_at_nowhere.com)
Date: 03/20/05


Date: Sun, 20 Mar 2005 22:13:22 +0000

Richard Fateman wrote:
> Dave wrote:
>
>
>> <snip>
>
>
>>
>> What would happen if it was free?
>>
>> The usage would dramatically rise.
>
>
> My guess is that you are wrong here.
>
>
> This assumes a kind of elastic market that probably does
> not exist. By your reasoning every piece of free software
> would be used by every student.

No, I am *not* claiming every bit of free software would be used by
every student. How you infer that from what I wrote I do not know.

Whilst I might guess some historians studying 17th century literature
might use Mathematica, the product will only really be used in any great
number in science, engineering and maths courses, with lesser (if any)
usage on other courses.

But providing it free would likely mean it would likely to be taught in
just about every physics, engineering and maths course. I know it is
used in computer science too.

I know that several years ago a decision was made in my department not
to contribute any more toward the cost of a university site license for
Mathematica, on the grounds of cost and the number of users. That
effectively reduced the number of users to about 3 in a department of
150 or so staff. There must be 50+ staff using Matlab.

The uni still has a site license - but since we don't contribute to that
cost, we don't use it unless we buy a copy, as I have done.

To a large extent, the number of undergrad students using a product is
set by the number of lecturers/post-docs that use the product. And that
is set to a large degree by the cost of the product. If two products are
each capable of doing the required task, the cheaper or one best known
to that individual will generally be bought.

> If Mathematica dropped its
> price, it might sell about the same number. The number
> is bounded by the "math geek population" more than the
> "population with $20 to spend".

But where I work at least, it is not someone with $20 to spend, but (for
staff like myself) someone with around $360/year to spend. I know with
100% certainty that we no longer use it much due to the cost. Nobody is
too keen either on software that stops working after a certain date. No
longer being able to get support and updates is not such an issue, but
it is not too tempting to buy something that you know will stop working.

I can't say how many more users there would be, but believe it would be
significant.

And I know there is talk of not renewing the site license any more. With
around 20,000 students (many on mathematically based courses), that
would be a loss of Wolfram Research's potential future sales.

> In reality, most people in their jobs do not even need
> Matlab (whose numerical-only capabilities are much less ambitious
> than Mathematica or Maple or ...). The employer is
> much more likely to provide a license for Microsoft
> Powerpoint or Excel to each technical employee.

Matlab has a toolbox for symbolic maths, which still does add much to
the cost. Someone told me the symbolic toolbox for Matlab is a cut-down
version of Maple, but I don't know if that is true.

> It is a relatively small market, and "several million"
> is probably inflated.

I suspect that too.



Relevant Pages

  • Re: University License fees are short sighted of Wolfram Research
    ... >> would be used by every student. ... > number in science, engineering and maths courses, with lesser (if ... > to contribute any more toward the cost of a university site license ... on the grounds of cost and the number of users. ...
    (sci.math.symbolic)
  • Re: Image processing prototype software
    ... two main complaints I have against MATLAB are no user interface ... Optimas (a now defunct package that had a lot of higher level ... ImageJ has a fairly user friendly user interface, ... time and cost as compared to getting a different package, ...
    (sci.image.processing)
  • Re: MuPAD is too expensive / Axiom is free
    ... The cost to download it ... WinXP is free because of a site license. ... if you ask your students to use a particular software which runs ... Compiling maxima and axiom from sources is equally ...
    (sci.math.symbolic)
  • Re: Call a MEX function from Matlab Engine?
    ... Compiling is not the issue, ... classes and custom functions in a mex routine. ... with a separate MATLAB process via a COM link. ... If you can set your problem up so that your cost function ...
    (comp.soft-sys.matlab)
  • Re: Telecommuting
    ... The test equipment is downright cheap compared to the cost of the development software I use in my business. ... Even Matlab is comparatively cheap. ... Although he did say that he bought 3 or 4 toolboxes. ... plus the matching simulink tools. ...
    (comp.dsp)