Strategies for Math on the Web

There are many different kinds of math and science communication, from research articles to interactive distance learning material. Different technologies have different strengths and weaknesses, so picking the right technology for your task is important.

Here, we look at the strategies and technologies for approaching five different kinds of math and science communication: 

• Creating Web and Paper Versions of the Same Document
• Electronic Distribution of Paper Documents
• Interactive Math in Web Pages
• Interactive Math and Science Web Sites
• Miscellaneous Strategies

Creating Web and Paper Versions of the Same Document

The most important, and probably most common, communication challenge authors face is producing Web and paper versions of the same document. The power of the Web to put people and information together is well-known. At the same time, the superiority of printed documents for portability and ability to convey dense information requiring intense concentration has never really been challenged by Web documents. So the motivation to create both Web and paper versions of documents is high. The problem is that few authors have time and energy to present and maintain the same information twice in two completely different ways.

Some special effort will always be required to take full advantage of any media. For example, designing navigation bars is a task unique to Web documents, while page formatting is primarily a problem for printed documents. Thus, it isn't realistic to think of creating both Web and print versions of a single document if you want to use special kinds of formatting specific to either media type. However, the good news is that for documents consisting mostly of text , a little care and good tools can usually produce Web and print versions of a document with minimal extra effort and minimal compromises in quality.

On the other hand, you shouldn't prepare dual versions of a document just because it isn't much extra effort. Before beginning a project you should ask yourself whether dual versions really make sense in your situation. Dual version documents make the most sense in situations where full online access to information will be sufficient for most readers, but some substantial fraction of readers will want to continue working offline. This is frequently the case with training or educational material, or reference material. Dual versions also make sense when reaching a large and diverse audience is important -- for which searching and indexing of Web versions are important -- but the material itself is substantial enough that many people will want to work with it offline.

The Technology Options

Best Choice: MS Word and MathType

For authors in most situations, MS Word and MathType is the only viable solution. TeX authors have several acceptable conversion options, but the cost/benefit analysis of learning TeX just to be able to use an HTML converter makes no sense for average authors.  See MathPage: From Word to the Web for more information.

Electronic Distribution of Paper Documents

A second extremely common technical communication task is distributing paper documents on the Web that would otherwise have to be faxed or mailed. Apart from the expense of other distribution methods, electronic distribution has the advantage that it reaches large numbers of people very rapidly. With the Web, it is also much easier to set things up so the exchange takes place when the recipient is looking for the information, rather than when the sender is thinking about disseminating his or her work.

At the same time, electronic distribution is a fairly limited use of the potential of the Web. It usually works best in homogenous author/reader communities where there is already a culture of exchanging paper documents (e.g. research mathematicians), or where online reading is unimportant compared to printing (e.g. tax forms).

As a general rule, strategies for electronic distribution involve putting documents online in some format intended for printing. Because these are usually binary files, they can't be easily previewed, or searched and indexed by Web search engines. With some formats, special search engines can be used on a specific Web site, but a more common work around is to prepare HTML abstracts that link to the downloadable binary files.

The Technology Options

Best Choices: PDF or MS Word and MathType

PDF is the successor of Adobe's popular PostScript printer language and document format. As you would expect, one of its strengths is that it prints well on almost any printer. Adobe has also worked hard to make its Acrobat Reader plug-in easily available, and it comes pre-installed in browsers on many new computers. Creating PDF files is quite easy with Adobe's Distiller program, and TeX/LaTeX authors can take advantage of the free pdftex package to create PDF files. Because of Adobe's corporate reach and strong support of PDF, it is the best choice strategy for electronic distribution of paper documents in many situations, especially for highly designed documents with extensive graphics.

Some readers find online viewing of PDF files a little more awkward than reading Web pages, but most people find it acceptable. Other drawbacks are that PDF files can only be partly integrated in with other Web pages, readers may have to install the Acrobat Reader plug-in, and the files can be pretty big. But provided you don't expect it to take over the role of regular Web pages, it is an excellent solution to a common problem.

In the specific case of of MS Word documents, however, MathType's MathPage technology is also a strong contender. Print quality rivals that of PDF, and as the output is HTML, integration with the rest of the Web is much better than it is for PDF documents. The main downside is that you no longer have total control over the print quality, since page breaks and margins depend on readers' browser settings. So in situations where it is important to guarantee that readers will be printing an exact copy of the original document, PDF is still preferable.

Interactive Math in Web Pages

Part of the incentive to go online is the appeal of engaging Web pages that take full advantage of the medium. In these situations, interactivity, graphic design and multimedia become more important, and paper hard copy may not be a requirement. 

Most interactive math content falls into one of two categories: client-side interactivity contained within a single page, or interactivity shared over a collection of pages with scripts on a server coordinating between them. As a general rule, interactivity on self-contained pages is better suited to "pre-programmed" scenarios, such as stepping through the solution to an exercise. Interactive math projects involving scripts on a server are more flexible and powerful, but they require much more web programming expertise. We consider techniques better suited to interactive math "sites," considered in the next section.

Modern browsers do now have support for displaying and manipulating equations. However, many do not have math support, so a key factor distinguishing strategies for interactive math is the extent to which older browsers must be supported.

The Technology Options

Best Choice: WebEQ Controls

It is difficult to come up with a single best choice when it comes to strategies for interactive math pages, since specialized solutions often work very well in some situations and not at all in others. However, the approach embodied by WebEQ Developers Suite has a lot to recommend it.

It is always tempting to try to use standard HTML techniques without resorting to special tools and technologies, since that is what most authors know, and what most readers are guaranteed to be able to view.  However, for technical material that is frequently just too awkward and limited. While going this route maximizes accessibility to your pages, it is tedious and painstaking to build pages this way, and the result is all too frequently lacking. 

At the other extreme, full page plug-ins such as LiveMath offer slick user interfaces and integrated authoring tools. But they break so sharply from the Web mainstream that they really only work well in situations where authors know their reader community quite well, and can insure they will really install the plug-in. Moreover, with this route, authors are at the mercy of their technology provider, especially after going through the learning curve with the authoring tools.

The WebEQ Developers Suite approach is to strike a middle ground. By using applets instead of plug-ins, authors can reach a wider audience. By using math-capable components within a standard HTML page, authors can draw from the whole gamut of HTML techniques and media types that modern browsers provide. Finally, since WebEQ  components are based on standard MathML, authors can be assured an upgrade path for their documents, as browsers and other software begin to add WebEQ-like functionality.

The main drawback to the WebEQ approach is that integrating components with HTML and JavaScript means that authors must to some extent be programmers as well.  To minimize the level of technical expertise requires to start building interactive math pages, the WebEQ Developers Suite introduces a Solutions Library of templates implementing common interactive math tasks. 

Interactive Math and Science Web Sites

Larger math and science Web sites require programming logic to coordinate the readers experience of the site as a whole as well as interactive math within pages. In general, dynamic Web sites require serious Web development experience. Situations vary so widely that the concept of a best approach doesn't really make sense. However, we can list some important techniques and technologies that contribute to Dynamic Web sites.

The Technologies Options

Comments

With graphics and computation, Web developers have a huge array of options to chose from. By contrast, when it comes to getting math input from readers, there are very few choices. The WebEQ Input Control is virtually unique in providing Web developers a means of accomplishing this important task.  

The most basic approach for getting math input from readers is to use normal text input fields for entering some kind of math markup into an HTML form. Typically, input will either be some vaguely TeX-like language or a computer algebra input language. Because entering math this way is quite error-prone, Web applications often try to give readers a graphic preview of what they are entering in some way. WebEQ Developers Suite contains a powerful Equation Server that processes math markup into a form that can be displayed to the reader.

A much more natural and sophisticated alternative is to embed the WebEQ Input Control in the Web page. The Input Control functions like a mini-equation editor within the page. Using JavaScript, the Input Control can be easily interfaced with server-side scripts, and other scriptable elements in the page. 

Miscellaneous Strategies

There are a few approaches that deserve mention but do not fit easily into other categories. These include the use of specialty browser/editors such as Amaya which can both display and edit Web pages with MathML, and interfacing to computer algebra systems, e.g. webMathematica and MapleNet. Although currently limited in usefulness to situations where all the interested authors and readers have access, these tools have a role to play in bridging the gap until MathML support is widespread in browsers.