Goals of the WebEQ Project

Everyone knows it is going to be important. Everyone knows it is going to change the way we work and interact. But as of today, the web is still too complicated and cumbersome to make a big difference for most of us.

As with any new technology, there is a problem:

• Until tools are created and refined, it is too hard for most people to do anything serious with it.

• Until most people do serious things with it, there is little incentive to create and refine tools.

Our goal with WebEQ is to provide a tool and develop techniques that help the scientific community use the web to exchange ideas simply and easily. We believe that many mathematical and scientific ideas are best conveyed graphically, and that scientific information often lends itself to hierarchical exposition in hypertext. In addition, electronic documents can be searched, expanded to include feedback, dynamically extend bibliographic data, and be rapidly and cheaply distributed.

WebEQ is an attempt to get the ball rolling. Stable, robust solutions for including mathematical text in web documents is underway, but further developments in browser technology and refinement of language standards are still necessary. However, in spite of the technological inadequacies we must still contend with, we believe WebEQ is a sufficiently powerful tool to enable serious researchers to create viable, interactive, visual electronic documents presenting their research.

WebEQ vs. latex2html

• Programs like math2html require users to prepare documents in latex. While tex and its relatives are widely used by researchers, they are extremely complicated programs, with a great deal of built-in flexibility. Beside the obvious consequence of being difficult to learn, experience indicates that tex documents are difficult to standardize. A document that compiles here won't compile there. This is not good for a medium like the web.

• Programs like math2html and latex2htmlare also large and complicated. They are design to work primarily on Unix workstations, and require a much higher level of computer savvy than many web users possess.

• Math notation included as graphics is difficult to edit or change. For one thing, there tend to be a large number of tiny .gif files with cryptic names. For another, it is not much easier to fix one tiny error than to regenerate a large piece of the document.

• Notation included as images won't scale to match the surrounding text and it is difficult to make it look good.

• The source code for the equations is right there in the .html document where it is easy to find and edit. Making a small change is trivial.

• Web authors can see easily see what they are getting without going through an involved conversion process.

• The level of complexity of WebEQ matches that of .html and is much more appropriate for a medium where the primary display is a computer screen where resolution and layout flexibility are limited. Languages like TeX are closely tied to the requirements of print documents.

• WebEQ requires only a text editor to create .html files. It works just as easily on a Window or a Mac machine as on a workstation. It requires no complicated installation, and it takes little disk space.

• The WebEQ applet remains active while the reader is on the page. In principle, it is simple to have WebEQ equations interact with other java applets such as graphing windows.

• Unfortunately, WebEQ is a bit ahead of its time still. The Java language and the Netscape Browser are both still suffering growing pains. There are problems with the communication between WebEQ and the browser that are major limitations until Sun and Netscape fix them (hopefully soon). In particular,
  • WebEQ cannot resize itself on a web page when the user changes font size. It issues the requisite function call, but at present, the browser ignores it. This requires a cumbersome height and width specification in the .html source.
  • Netscape won't print WebEQ equations (or any other Java or Javascript output).

WebEQ vs. other Strategies

However, many of these techniques offer a good deal of promise. A fuller discussion of some of these alternatives can be found elsewhere.

The Geometry Center Home Page

Comments to: rminer@geom.umn.edu
Created: Apr 01 1996 --- Last modified: Wed Jun 12 16:41:02 1996