Markup Languages

Concepts

Word processors such as Microsoft Word are based on the principle of “what you see is what you get” (WYSIWYG): What the author creates is essentially what the reader sees. Unlike an unformatted ASCII text file, a document produced in Word can be instantly and attractively formatted with different fonts and effects. Word processors achieve this by embedding control characters (strings of normally invisible characters that control formatting effects) in and around the main text. WYSIWYG word processing became possible only in the 1980s, with the invention of graphical user interfaces (GUIs) and high-resolution, bitmapped computer screens that could display a range of different fonts. Before then, word processing meant text editing—text was simply typed into and moved around in ASCII files before being printed out, often on a crude dot-matrix printer. That was fine for utility bills and simple letters, but not for more sophisticated documents that needed a range of different formatting effects.

Markup language offers a way of embedding structural codes called tags into a basic ASCII text file. To mark the start of a paragraph, one might use <P>. To make text into a main heading, one might put <H1> (meaning start a heading 1) in front of it and </H1> (end heading 1) after it. The crucial difference between a word processor and a markup language is that, while the former involves specifying exactly how different bits of text will appear, and not what they are, the latter involves using tags to specify exactly what elements of text are, and not how they will appear.

This difference is clearly illustrated by HTML Web pages, whose various elements are identified with tags such as <UL> for a bulleted (unordered) list, <P> for paragraphs, and so on. Using basic HTML, it is impossible to control exactly how a Web page will ultimately appear, because that depends on how the user's Web browser is configured to process the tags. While on one browser, <H1> might produce a large, bold Times font, on another it might produce a medium-sized, italic Courier font. The HTML conveys no formatting information of this kind.

To someone accustomed to Microsoft Word, markup initially seems confusing and perverse. What is the logic in using a text-processing system that does not allow its authors to control the ultimate appearance of their documents? Isn't markup an extraordinary waste of time? Why type hundreds of extra control characters when one can simply click on the appropriate words and immediately make them into a heading in 14-point, boldface Times?

Advantages and Disadvantages

The huge advantage that markup offers over word processing is the way it separates the process of authoring text (adding markup) from the process of formatting text (turning the markup into a printed or online document). This can bring tremendous benefits to publishers. IBM's policy of using markup languages rather than desktop publishing to produce its reams of documentation—the company is reputedly the world's second-largest publisher—reaped huge dividends in the mid-1990s. Thousands of computer manuals tagged in IBM's internal markup language BookMaster, and originally designed for printed output, were almost instantly converted for publication on CD-ROM, not by changing the markup, but by formatting the markup in a different way to produce online hypertext files instead of printed output. With the same markup, IBM could simultaneously produce output for different printers, numerous different electronic book systems on CD-ROM, and HTML files for the Web. Had the original files been prepared with a word-processing program, it would have been necessary to convert them laboriously for each different type of output.

...