Date: May 14, 2013 3:15 AM Author: David Park Subject: Re: Work on Basic Mathematica Stephen! I don't believe the "training wheels" model can really achieve the

objective. Are these people preschoolers? The situation is rather something

like this. Suppose all education was conducted orally until the junior year

in college. No or little writing. Then the students are to get into some

serious writing work. They are to write a critique of Marcel Proust's

Remembrance of Things Past. They do get some powerful tools to do this: a

computer with Microsoft Word, which has spell checking and some passable

syntax and grammar checking; a dictionary and French/English lexicon, and

French and English versions of Proust's work and maybe a couple grammars.

And since the students have never written much before, Word has been

augmented with training wheels. A little button always appears at the start

of a new paragraph with choices 1) Would you like to type a new sentence? 2)

Would you like to enter a sentence in free spoken form? 3) Would you like

Literary|Alpha to search for ideas on some topic? 4) Would you like to start

a new Section? At the end of a paragraph a Suggestions box will appear with

something like "Would you like us to add a sentence on Marcel's health

problems in relation to the topic?" Gee, I never thought about health

problems. So nice to remind me of that. Several weeks of this and they

should be up to speed.

Do you believe that method would achieve some worthwhile objective? Isn't it

rather that it usually takes years for a student to become really good at

expressing and manipulating ideas in written form? It's not surprising that

some students, without much experience, would become terrified at a blank

sheet of paper. Now add in mathematics and all of the new active and

dynamical possibilities for expressing and manipulating mathematical ideas

and don't we have a considerably greater learned skill? You can't replace

extended education and practice by software. It's the failure to get

Mathematica into early education that is the problem and getting it there is

the remedy.

As far as the development of Applications goes I don't see why WRI couldn't

buy or commission good ones and incorporate them as standard add-ons.

Remember that a good application might have many specialized convenience or

display routines for the subject matter and it would really be somewhat of a

burden to have all of these in the core Mathematica, both for WRI and for

users who would have thousands and thousands more routines thrown in their

faces. If WRI tries to ameliorate the situation by a truncated set of

routines then they end up with unpolished and inconvenient capabilities.

Most users might use only one or a few Applications.

The major problem with Mathematica Applications is that they are a new

medium and we are still learning how to write them. Unfortunately,

mathematical skill and skill at writing Applications seldom go together. I

won't even begin to explain all the quirky things that happen! Yet a good

Application needs both good mathematics and good design.

David Park

djmpark@comcast.net

http://home.comcast.net/~djmpark/index.html

From: Murray Eisenberg [mailto:murray@math.umass.edu]

Re "clean" notebook interface:

An experienced Mathematica user might well prefer a totally clean notebook

as starting point for some work. But a new user, or potential new customer,

might well panic at an essentially blank window. (The faint horizontal line

with it's "+" icon at the top of a new notebook window is at least a

starting point.

Similarly for once the new user has typed and possibly evaluated some input:

what should go there? what's the correct form? what can I do with it? So WRI

has attempted to provide some guidance directly in the notebook, outside the

Documentation Center. (Whether it has the intended effect is a separate

question.)

For example, suppose the new user, or somebody just trying out Mathematica,

successfully types and evaluates:

Plot[Exp[-x] Cos[x], {x, -Pi/2, Pi/2}]

And possibly (probably?) the user wants to enhance the graph. How do that?

Well, the Next-computation Suggestions Bar provides an immediate and obvious

way to approach it -- without having to look up Plot, wade through the long

list of Options.

I write the above as somebody who has helped hundreds of university students

learn Mathematica and who realizes how much more efficient the learning

would have been over the years had such front end doo-dads been available.

They're like bicycle training wheels: they can help you get started, but you

can get rid of them when they get in the way.

Re specialized applications:

One of the things I hated about another technical computing product that's

very popular, especially in engineering circles, was that to do so many

things -- at least at one point, even to do symbolic manipulations -- you

had to purchase separate add-on products. By contrast, one of the things

I've appreciated about Mathematica from the start was the inclusion, in

fact, integration, of graphics and symbolics with numerics; and

increasingly, the migration of add-on packages into the kernel. (Surely part

of the reason is that software and hardware advances permit doing so.)

On the other hand, there are areas of applicability where WRI can expect to

target specialized products that are, I presume, more expensive than plain

Mathematica. I see no reason why they should not leverage the Mathematica

platform to produce such products -- especially if the revenue thereby

generated can help support maintenance and further development of

Mathematica itself.

Re maintaining & improving Mathematica:

Nothing I wrote above is meant to minimize the importance of other things

that need to be done, from fixing bugs to maintaining a stable notebook

appearance -- e.g., not willfully changing the default font family or font

color for Section, Text, etc., cells. And by all means, make it MUCH easier

to produce and deploy packages, including their documentation.