MINSE vs. LaTEX

• Programs like LaTeX2HTML and Hyperlatex require users to prepare documents in LaTEX. TEX is a large, very powerful, and complicated program, and TEX documents are difficult to standardize; often documents that compile in one place won't compile in another. This causes problems in an environment like the World-Wide Web, and makes the job of programs like LaTeX2HTML very difficult.
• LaTeX2HTML is itself quite complicated, and converting from TEX in general requires the installation of many tools. It is usually only feasible on a Unix workstation. The conversion process takes your document through many steps to produce the resulting images:
  • First your document has to be compiled by TEX.
  • Then dvi output from TEX gets converted into a PostScript file with dvips.
  • Then GhostScript, a PostScript interpreter, has to render the result to a PPM (portable pixmap) image.
  • Then the image has to be cropped to the right size.
  • Finally the image has to be converted from a PPM to a GIF using ppmtogif and an appropriate link inserted in the main document.
  You need to install all of these programs to make it work. If a problem occurs anywhere during this convoluted procedure, you may have to spend some time figuring out what went wrong -- and this process has to take place for each and every image in your entire document.
• This means the process takes a long time, and editing the document is difficult. Once an equation has been rendered as an image, the image is fixed and difficult to change; you have to edit the original document and run the conversion again (on the entire document!) to see the results.
• Moreover, the output produced by LaTeX2HTML is full of arbitrary numbers and generated link references like node19.html#SECTION000410000000000000000 and fragment identifiers like tex2html631, since it's produced by a computer that can't read and understand your text. You may end up generating hundreds of images from the equations in your document, all with incomprehensible filenames like img66.gif or _5696_tabular1030.gif.

When you use MINSE, things are quite different.

• Your document gets processed in one pass. The polymediator parses the expressions in your document and inserts the link references. It's simple, and if something is wrong with your expression you'll be shown right away what it is.
• You don't have to wait for conversion very long. You can edit the document in place and preview it. It's quick, and you don't have to install anything on your machine; all you need is a text editor, not a Unix workstation.
• The URLs that get produced are very long (because they contain all of the information that the renderer needs to make the image), but you will never have to see or work with the file or the filename.
• For a graphical rendering, the images are generated each time the document is viewed; you don't have to save dozens of tiny files along with your document. For a text rendering, the document is processed and cached. You get multiple renderings of your document for no extra work on your part.
• More people can read documents containing MINSE structured expressions, because it can be rendered to different media. Currently, expressions are viewable both as graphics and as text! Output from LaTeX2HTML will never be viewable on a text terminal.