W3C home > Mailing lists > Public > www-math@w3.org > July 2006

Re: Webmath

From: Mihai Sucan <mihai.sucan@gmail.com>
Date: Tue, 18 Jul 2006 21:00:35 +0300
To: "Robert Miner" <robertm@dessci.com>, "June Lester" <june@sfu.ca>, www-math@w3.org
Message-ID: <op.tcwca90vmcpsjgr0b0dp@localhost.localdomain>

Hello Robert and June!

Le Tue, 18 Jul 2006 18:02:13 +0300, Robert Miner <robertm@dessci.com> a  

> Hi June,
> Thanks for reminding us that at the end of the day, it's not about how  
> you put the math on the screen, it's what you do with it. Your  
> interactive Flash widgets for linear algebra are nice.  They remind me a  
> bit of
>    The National Library of Virtual Manipulatives  
> <http://nlvm.usu.edu/en/nav/index.html>
> and
>    ExploreLeaning Gizmos <http://www.explorelearning.com>


The Flash "widgets" linked by June reminded me very much of The Geometer's  
Sketchpad [1]. A dynamic geometry software, like many others: Kig [2]  
(from KDE Edu), and Geogebra [2], etc. For me it simply does not matter  
the fact GSP uses a proprietary binary format, as long as it provides what  
I need.

> For the manipulatives, I've seen all kinds of different technologies,  
> but the main ones are Java, Flash and DHTML.  Each has it's own  
> strengths and weaknesses, but my sense of the current trend is that  
> Flash is in the ascendent for this kind of thing.  Java is okay, but  
> client-side support for Java is not as good as that for Flash, and I  
> suspect Flash is easier to program for most people.  DHTML has a huge  
> range, but it is very hard for fancy stuff, not only because of problems  
> displaying math, but because of all kinds of browser differences --  
> versions of JavaScript, DOM differences, CSS differences, HTML layout  
> differences, etc.  It's ironic, that while there has been a huge amount  
> of effort over many years to standardize and implement all the parts of  
> the DHTML platform (MathML included) it has only been partly successful  
> (see  squabbling on this list as ample evidence of that).  Java, with  
> it's open community process, is better, but still suffers from  
> incompatibilities.  Purely proprietary Flash, on the other hand, seems  
> to be fairly solid and uniform across most platforms.

A good point, however Flash is plagued by its own flaws too. After  
extensive use of Flash, switching back to the "good-old" HTML is  
refreshing and relaxing. I do not like Flash too much, even if it's proven  
to be rather stable and solid along the years.

Flash is very good for the purpose of June. Those widgets wouldn't be  
better done in HTML+JS+CSS, nor better in Java.

However, I do not like too much the fact those are quite limited. Flash  
does not provide the appropriate way to just start inputting mathematical  
formulas, dynamic geometry, nor HTML-like content. Gladly, HTML-like  
content is something they are getting at, but it's not enough, for my  
taste. Regarding dynamic geometry, you can do that if you write your own  
ActionScripts, your own "drawing application". Flash is still not  
something any teacher (or a student for that matter) can use on a daily  
basis for teaching/learning maths and/or geometry. It's very good only if  
someone, like June, provides premade Flash documents.

I'm almost sure June can't add a new dynamic arrow, a new function plot, a  
mathematical formula and/or a calculus in less than 5 minutes. Everything  
properly integrated in the same "sketch". This is a problem Sketchpad and  
Kig-like software solves. These can also be used for teaching mathematics.

GSP provides a Java applet which allows authors to publish their works  
online. Even if I don't like Java either (I'd pick Flash instead of Java  
in a blink), this is a full-fledged programming language: mature and solid.

Flash is not oriented towards educational software/applications. It's  
oriented towards entertainment.

As for XHTML+JS+DOM+CSS+XML+XSLT+Xwhatever I personally see this best fit  
for publishing math documents. As noted, differences between  
implementations make it harder to use, and, for example, even if you stick  
with an Opera-only site it's rather hard to develop some web application  
powerful enough. In the coming years, many of the problems will be slowly  

> However, most of the folks I know still trying to do real, interactive  
> math on the web -- regardless of the technology they use for the  
> manipulative -- still complain bitterly to me about how hard it is to  
> create the halo pages, with a mix of expository math and text.  They  
> always ask me for a good authoring tool along the lines of an HTML  
> editor with an built in equation editor.  They wouldn't care if it made  
> the cruftiest HTML + images, so long as it was simple, graphical and  
> intuitive.

Yes, that's very true. Everything comes down to that. All I can say I have  
tried several math editors and Mathematica 5 is well ahead of all others  
(IMHO). MathCAD, Maple and all the others (which are not worthy of  
nomination) don't even come close. This applies to speed and editing  
features. For caclulus I've heard Mathlab is better, but that's not what  
an editor needs much.

Mathematica has features very, very useful for quickly writing  
mathematical documents. I myself find it very annoying to have to play the  
click-game in MathCAD for something simple. For me speedy editing is a  
requirement for any math editor, then comes rendering beauty (for me  
Mathematica is good enough).

Function plotting is not too easy to use. For that, and for geometery, I  
am using GSP.

Mathematica also provides export to MathML, HTML and the like. The  
generated code is not the worst nor the best I've seen.

So, these are my recommendations: Mathematica and Sketchpad (or Kig).

AFAIK, there's no open-source/free project coming close to Mathematica, or  
at least MathCAD.

What do you usually recommend?

[1] http://keypress.com/sketchpad/
[2] http://edu.kde.org/kig/
[3] http://www.geogebra.at/

ROBO Design - We bring you the future
Received on Tuesday, 18 July 2006 18:08:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:27:38 UTC