MathGroup Archive 2005

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

Search the Archive

Re: How to View Mathematica and Hardcopy Books

  • To: mathgroup at smc.vnet.net
  • Subject: [mg62415] Re: How to View Mathematica and Hardcopy Books
  • From: "Steven T. Hatton" <hattons at globalsymmetry.com>
  • Date: Wed, 23 Nov 2005 01:13:20 -0500 (EST)
  • References: <NDBBJGNHKLMPLILOIPPOGEFAEMAA.djmp@earthlink.net>
  • Sender: owner-wri-mathgroup at wolfram.com

David,

On Tuesday 22 November 2005 12:32 am, David Park wrote:
> Stephen,
>
> Mathematica would be perfectly suitable for calculational needs of many
> data problems in the lab and field. 

I tried using Mathematica to write up some labs several years ago.  
Unfortunately a combination of some very nasty bugs, and my having neither a 
sufficient grasp of how Mathematica works nor access to someone who could 
help me in realtime, made the effort fruitless.

> It would be far quicker to develop and 
> this would also be an area where palettes and GUI interfaces might be more
> appropiate than my favorite textbook style document. 

I belive something along the lines of a cookiecutter lab book provided for 
students could be valuable for both introducing Mathematica, and providing a 
tool that enhances the original objective of say a physics or chemistry 
course.

The same might apply to certain professional environments.

> I've been using Mathematica fairly intensely for
> about 8 years. There are still many areas of it that I don't know very
> well, such as how to use the statistics routines or the ins-and-outs of
> differential equation solving. Still I find that my limitations are not in
> understanding Mathematica, but in understanding the math in the fields I am
> studying. Once I understand what is happening with the math I have little
> trouble in applying Mathematica to the problems.

The way I have attempted to use Mathematica in the past has been to try 
applying it to problems on the edge of my understanding.  That has often been 
fairly non-productive.  Furthermore, I have a stunning knack for encountering 
bugs and gotchas that 99% of people will not encounter.  See:
news://Message-ID: <cpuh5b$h24$1 at smc.vnet.net>
for an example.  

I received an email from someone at WRI explaining what was happening, but by 
the time I received the message, I was off doing other things.  I have since 
lost the message, so I have never learned what was actually happening.

> Certainly the developers of Mathematica used principles of computer science
> in its development. And it certainly may be that computer science could
> make significant improvements in Mathematica, or help to develop a much
> better CAS. If that is so, I would be glad to see it. But elegant computer
> science does not always lead to a useful product. There are a lot of
> products with elegant theoretical design that totally fail to comprehend
> what the customer wants and needs.

This is a list of Stroustrup's guiding principles in designing C++:

        * C++'s evolution must be driven by real problems.
        * Don't get involved in a sterile quest for perfection.
        * C++ must be useful now.
        * Every feature must have a reasonably obvious implementation.
        * Always provide a transition path.
        * C++ is a language, not a complete system.
        * Provide comprehensive support for each supported style.
        * Don't try to force people. 

You can find more elaboration on these at this site:
http://www.ibiblio.org/javafaq/newsletter/1.1.html

For even more on this topic you could read Stroustrup's very interesting 
_Design and Evolution of C++_

What I find lacking in Mathematica is far more pragmatic than any kind of 
theoretical formalism.  One aspect of computer science deals with very arcane 
and formal ideas.  Another deals with the essential usability of a 
programming language.  It is that latter which is most relevant to me 
regarding Mathematica.  

> In the meantime, users who want to get the most use out of Mathematica
> should forget computer science, use Mathematica the way it works, and try
> to learn to think mathematically.

Actually, that has been a problem for me when trying to learn Mathematica.  
Figuring out how to translate the mathematics I am familiar with into 
Mathematica expressions did not work well for me.

> You have to actually try it on many problems
> of the type users would tackle and see that it actually works, is
> convenient and is intuitive. A single fancy idea is not enough.

There are a number of issues with creating and using packages which are not 
obvious without a fairly good understanding of how Mathematica works.  As for 
usability, I believe it depends on whether you are trying to create a package 
for your own use or for the use of a wide audience.

> Right now I think it is learning how to use Mathematica properly that is
> limiting its potential. Many uses do not get the most effective use from it
> because they don't understand the capabilities it already has. There are
> ideas that might be useful, such as live Mathematica web pages that you
> could collaborate with someone else in real time, but I don't have the
> equipment for it or have any idea how I could actually do it.

If you are responding to my comments regarding the similarity between 
Mathematica and a web browser, I believe you are missing my point.  My point 
is that the Mathematica front end is a lot like a web browser in how it 
presents a Mathematica notebook locally.  The W3C DOM is about much more than 
simply creating web pages, or web browsers.

I will also add that understanding the correlation between, for example, 
MathML, and the displayed result of rendering it in a web browser can cast a 
lot of light on how mathematical expressions actually communicate ideas.
Steven


  • Prev by Date: MATHwire: Information for Mathematica Users
  • Next by Date: Timing of looping operators
  • Previous by thread: Re: How to View Mathematica and Hardcopy Books
  • Next by thread: Re: How to View Mathematica and Hardcopy Books