Why Wolfram Tech Isn’t Open Source

[Fredrik Johansson]

https://blog.wolfram.com/2019/04/02/why-wolfram-tech-isnt-open-source-a-dozen-reasons/

[rjf]

On Wednesday, April 3, 2019 at 5:36:58 AM UTC-7, Fredrik Johansson wrote:

https://blog.wolfram.com/2019/04/02/why-wolfram-tech-isnt-open-source-a-dozen-reasons/

There are various good points made in the blog, but the executive summary is

this..

Much of this blog could be summed up with two answers: (1) free, open-source software can be very good, but it isn’t good at doing what we are trying to do; with a large fraction of the reason being (2) open source distributes design over small, self-assembling groups who individually tackle parts of an overall task, but large-scale, unified design needs centralized control and sustained effort.

.............

Some of the issues raised are addressed by Eric Raymond

https://en.wikipedia.org/wiki/The_Cathedral_and_the_Bazaar

but that book is really about 2 variants of open-source.

I think the guy writing this blog is missing the main reason

Mathematica is not open source. That is, that Stephen

Wolfram believed that he would make more money that

way.

It also bypasses the reality that the central unified design

(led by one or a few people including Stephen Wolfram)

is no guarantee of success.  SW totally failed with SMP.

I have argued that his second bite of the CAS apple,

Mathematica, is quite flawed.  The Wolfram language

in particular is what we used to call "a dog's breakfast".

 Making parts of Mahematica

open source might help, but why would anyone

pick it up and enhance it free,  vs. consulting

for Wolfram and getting paid?

The counter-example of everything open source

that comes to mind for this group discussion is

of course Sage.

Sage probably illustrates many of the things that

can go wrong with a large open-source project

that requires many hands and many (expert)

minds in many application area.   Including

mysterious interactions, poor designs. Maybe

neglect of essential but boring parts. And

difficulty funding.  As mentioned in the blog.

One argument --  before the success of RedHat--

(I failed to find the quote online

but it goes something like this...)

.. If I believe in open-source programming,

does that mean I have to deliver pizzas at

night so I can pay my rent?

We have, of course, a possibility to

see how this open source model works for

SageMath financially. And compare it to

its stated targets  (Mathematica, Maple)

and its unstated targets (Maxima and

other free systems that are in actuality

competing in the same space).

RJF