My experiences with MathML -- notes from a non-professional.

Hello:
 
Some time ago, I decided it might be an interesting thing to do to study
mathematics independently and seriously.  With the help of a university
library and a computerized library catalog I was able to make no small
amount of progress over the period of several years.  I collected a
bibliography and managed to reach some interesting conclusions.  I
decided that it might be rewarding to others to share what it was I
found.  The Internet pointed to a way I might do that.
 
Moving the collection of books that I had written down to my desktop
computer was easy using the bibliographic software ProCite.  Making my
bibliography available on line has obstacles, but none of them
insurmountable.  Actually trying to write down my observations proved to
be another kind of problem altogether.  After about a year of trying to
publish simple text with equations to the internet, I have decided that
it is at this point completely impractical using the available
technology.  It is the gravest disappointment to me that the technology
established by the World Wide Web Consortium for doing precisely this
has failed me in my goals.
 
Here are my efforts so far in publishing mathematical discussion to the
internet:
 
http://galileo.spaceports.com/~symbiota/HIR/HIR.html
 
And especially,
 
http://galileo.spaceports.com/~symbiota/HIR/HIRDifferentDiffForm.htm#Dif
ferential
 
 
Microsoft Word
 
My first experience was with Microsoft Word which has the capability of
creating HTML documents and even saving the equations that you use in
the document along with them as GIF files.  The results you get doing
this is very disappointing.  It is impossible, as near as I can tell, to
apply changes to simple things like background color and expect any
changes to the GIF file you have created.  Microsoft Word is where I had
my first experience with MathType, the Word plug-in that allows you to
create "Math Pages" capable of being served to the internet.  I
experimented with this and found that you could successfully create
documents that could include equations and also serve to a wide variety
of browsers.  This was encouraging, but the technology depended heavily
on scripts and also, once created, the HTML document could not be edited
by Word.  It could only be edited using an original Word document and
then recreated to the HTML document.  This is a real nuisance if all you
want to do is correct a few typographical errors or change some simple
formatting like background color by editing the HTML document directly.
If you have to edit or add an equation later, you find yourself
functioning with two different documents, the HTML document and the Word
document.
 
LaTeX
 
I heard that LaTeX had some potential for adding equations and also
creating documents to be served to the internet.  I looked into this
possibility and was disappointed to discover that despite several days
of exploration, investigation and research that I was completely and
utterly unable to even perform so simple a task as creating a readable
file of some kind -- any kind -- using LaTeX.  This, despite having had
considerable success with command line editors like vi.  Despite this
frustration, I investigated *TeX again when I heard about tbook while
hoping that XML might open the way to applying equations to HTML
documents only to be disappointed again.
 
XML
 
When Netscape and Mozilla finally began to be making browsers that could
interpret equations using MathML, I decided that this really was my last
chance for serving HTML documents with equations in them.  I had first
discovered this potential in using Amaya, the browser and editor that is
supposed to be able to create math legible documents.  But I was not
able to accumulate any real success using this software.  For some
reason, documents made with Amaya cannot be read using Mozilla/Netscape
and I despaired of asking my readers to install Amaya just to read my
work.  Asking them to use Mozilla or Netscape seemed more reasonable.  I
was also pleased to find that somebody had created a Math equation
editor in Mozilla, but was disappointed to find that I can't seem to
make it work.  You can save to clipboard and paste from there, but it
still is rather limited in how you can apply it.  It was experimental,
but I still had hopes.  There is also the EzMath Editor that will save a
MathML equation to the clipboard to be pasted into an HTML document.
This has been my latest hope and inspired me to explore XML since this
is the source of MathML as a technology.  But even when you can manage
to paste an equation into an XHTML document, there doesn't appear yet to
be any way of actually creating equations with EzMath Editor
 
I really like XML, I have purchased the Academic version of Visual
Studio .NET and really like the way that it handles XML.  I had hoped
that this software might provide some insight into applying MathML, but
I have not been able to utilize its XML functionality to explore MathML
despite successfully downloading an  XML Test Suite page and its XSL
pages from www.w3.org/Math and successfully serving them to myself in
Visual Studio .NET's browser (MathPlayer enabled).  Visual Studio .NET
cannot seem to interpret the MathML elements as data.  I really think
that XML is the future of the internet.  I would rather use XML to
publish text to the internet then HTML.  There is a bit of a learning
curve, but I am convinced that there is a very short bridge between
learning HTML for simple text applications and using XML for the same
purposes.  That bridge is style sheets.
 
Another XML programming tool that I have tried is XMLSpy.  This is a
very satisfying tool to use, it looks as though it might actually have
an advantage over Visual Studio .NET and there is a home version
available for about $100 US, so it is much more accessible then Visual
Studio.
 
Dreamweaver by Macromedia is the first web designing software that I
have seen that will create XHTML documents in WYSIWYG environment.
There may be some hope here, but it is certain that the environment does
not accommodate MathML at this time.
 
Scientific Notebook by MacKichen software is another math capable editor
that will save to HTML, but will not edit the HTML directly once
created.  Also, I had trouble with my browsers not reading the equations
correctly. 
 
Somewhere in the bowels of XML is the secret of MathML and applying it
using XML editing concepts.  I could imagine struggling to add equations
using XML elements alone if I had some confidence that what I was
applying yielded legible or even XML valid results.  I have attempted to
do this using various samples cut and pasted from every MathML editor I
could find and have yet to deliver up some kind of useful document.
 
The thing I see in common between XMLSpy and Visual Studio .NET is their
dependence on schema for creating some kind of XML project.  But, as
near as I can tell, MathML at this point depends completely on DTDs.
Now I'm not exactly sure what the necessary relationship is between DTDs
and an XML document, but I do know that Visual Studio .NET can actually
create schemas based on an XML document.  I've tried this, but Visual
Studio has problems doing it.  Would it be worth exploring the errors?
I've seen at the W3 consortium site that there is some kind of effort to
create a MathML schema and so it they have not been able to come up with
something, I doubt that I could hack my way to one using Visual Studio
alone.
 
If I had a good MathML schema could I start crafting MathML documents
using XMLSpy alone?  I'm not really sure; I actually have nothing to
suggest that I could.  This is how desperate I am, I am grasping at
straws.  I am convinced, however, that the key to a good, easily
available WYSIWYG MathML editor is not text editors like Word, but an
XML programming editor like XMLSpy.  I don't think such a thing is
beyond the grasp of your average high school teacher.  I am convinced
there is a good elementary pedagogy for making XML useful to the casual
user.  Generally, there is a need for a good WYSIWYG editor for creating
XML pages that are controlled by XSL style sheets and schema.  This can
already be done in many editors with HTML and CSS sheets - Visual Studio
.NET, for example.  Once XML pages can be edited as transparently as
HTML is now, then MathML can be easily integrated along with many other
excellent XML based technologies like SVG.
 
Editors and Proprietary Interpreters
 
I would hope that software developers abandon the idea of creating
proprietary interpreter/editor combinations for serving up mathematics
on the internet.  Even in an isolated environment like a university, it
is simply too much to ask to expect readers to install software to read
documents that can only be created by one editor.  This contradicts the
Internet and its vast potential.  But why is it that the hopes of a
universal markup language for mathematics should be so completely
impenetrable to the casual user?  Why should school teachers be obliged
to purchase or explore university grade software in order to post study
material to the internet for their students?  What is missing?  I can't
imagine that it is an editor alone that is the obstacle.  Again, I am
certain that there is an easy pedagogical approach to creating MathML
documents that is as transparent as the simplest "Hello World!" HTML
document.  Anyone can program HTML and, eventually, anyone will be able
to program the same simple HTML documents in XML too.  Where does MathML
fit into that simple potential and why hasn't it been filled already?
 
I thank you for your attention to my experience in this and thank you
for your efforts so far.  Despite my frustration, it has been a most
rewarding effort.
 
Paul Pikowsky, pksky1@hotmail.com, pksky@finestplanet.com