MathGroup Archive 2011

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: How to write a "proper" math document

  • To: mathgroup at
  • Subject: [mg120048] Re: How to write a "proper" math document
  • From: "McHale, Paul" <Paul.McHale at>
  • Date: Thu, 7 Jul 2011 07:29:18 -0400 (EDT)
  • References: <> <iuukk8$epi$>

>But if Mathematica had emerged from, and was continually further
>developed in, an open, academic, competitive environment, with
>publications, technical meetings, peer review, student involvement, and
>all the other advantages that this environment can provide, rather than
>a closed, quite secretive, and commercially driven enterprise, it would
>be even richer than it is today.

I believe there are other market segmented pricing structures that could increase proliferation of Mathematica.  I also think WRI needs an app store to encourage developers.  This would have to include a Mathematica based package management system.

Sal Mangano tweeted he would love to write more Mathematic books but knew the market was too small to make it worthwhile.  Growing the market will bring developers and drive the cost down.  Apple clearly leads the way with this model.

>Ask yourselves where many of the most important software tools of today
>emerged from? (and whether they are free and open tools, or closed and
>excessively expensive tools?)  Just to cite two of these:
>*  Unix:  The essentially open world of the original Bell Labs.
>*  TeX and LaTeX:  Donald Knuth and his academic students and disciples.

Comparing WRI to open source is very difficult.  Apple is the only company on earth to mainstream Unix.  Linux is to apple what ham radio is to cell phones.  Steve Jobs proved if you want to harness development and provide a cohesive mature product, it takes a strong unifying leader to provide direction. 

Look at all the flavors of Linux.  Even the package manager is inconsistent.  Develop for one distro and doesn't work on another without at least repackaging.  In the best case, Mathematica would appear cobbled together by people who haven't met (like another Ma* symbolic package).  In the worse case, it would become a fragmented effort that was impossible to usefully peer review.  I can't afford to use a product with a similar trajectory and fragmentation of Linux.  I use open source (subversion, Mantis).  I like it when it works.  When it doesn't, it gets pretty cold.

As far as LaTex, it might be a nice free academic tool.  It is not used in any of the commercial or defense companies I have worked at.  Regardless of fantastic numbers these examples allege, they have yet to enjoy the corporate desktop space.  Steve Wolfram isn't going to attract key developers to reliably contribute without using $$$.  I'm pretty sure corporate penetration is critical. 

As far as peer review... I'm sure WRI could provide free copies to open initiative regression testing groups that would follow the standard software development model of black box regression testing.  The software community has shown peer review of low level code is only useful in conjunction with ed output regression testing.  (many a slip between the cup and the lip)

Just my opinions,

  • Prev by Date: Re: How can I get better solution for this...?
  • Next by Date: Re: Bug 1+4/10
  • Previous by thread: Re: How to write a "proper" math document
  • Next by thread: Re: How to write a "proper" math document