Week 4 - Markup Languages

The Virtue of Text

As mentioned in the first week's notes, text was chosen as the medium of choice on the Internet because it is a "least common denominator" which any computer can understand. (Binary formats suffer from endianness and wordsize incompatibilities between architectures.)

The Virtue of Markup Languages

In order to apply typographic features to otherwise boring text, markup languages have been invented which describe the text they will modify. Many of these have been invented over time. The following is by no means a complete list.

• troff / nroff - This form of markup was originally invented at AT&T Bell Labs by the same guys who wrote the original UNIX (Kernighan, Ritchie, & Thompson). This format is still used for man pages.
• Post Script - This format was created by Adobe. It was originally a proprietary format, but it's proliferation led to greater and greater openness. Many printers support Post Script to format text when printing.
• TeX & LaTeX - A Mathematics and Computer Science professor by the name of Donald E. Knuth set to the task of writing a textbook for his students. Along the way, he learned that he would need some kind of markup language to format his textbook. Seven years later, the TeX typesetting language emerged. Many believe that it was well worth the wait, as it is a markup language which gives authors a high level of control in the formatting of mathematical equations, and can handle very large sized documents.

• SGML - The SGML Project is funded by the Information Systems Committee of the UFC. It is a a project to create a Mother-Of-All markup language which allows you to rigorously, precisely, author a document using data type definitions (DTDs). The idea here is that you can create a document which is completely independent of the tools used to author it. In addition to authoring documents, you can also define new markup languages with SGML. This format was submitted to all major word processor makers as the format of choice for document saving, and practically none of them went for it.

• HTML - As the World Wide Web was being developed, it was apparent that a markup language was needed. SGML, however, was just too fat and didn't have built-in support for things like hyperlinks. A stripped down version of SGML was made, and lo, HTML was born.

Shortcomings of HTML

HTML worked well as the ML for the World Wide Web: It's small, fast, supports embedded documents like sounds and images, and supports hyperlinks to other sites. It is not without its disadvantages, however.

• It is not precise - This is by design, actually, and from a certain standpoint cannot be considered a shortcoming. For people who want control over their document's layout, this is a real bugaboo.

• It does not extend easily - This shortcoming was manifested most prominantly durring the "browser wars" where the two prominent browser vendors did their level best to break or harrass each other's browsers by adding non-standard extensions, all in the name of "product distinction".

• It does not allow for re-use of formatting (or styles) - Desktop publishing software has supported style setting and reuse for years. Long-time users of said software were aghast to find this feature missing from the HTML spec.

• It does not have built-in support for dynamic content - Lots of people that do programming and want to make spiffy, neat-o Websites complained about this.

The Answer: XML

The W3C took these criticisms to heart and came up with a new standard for document markup which would address these problems. XML provides authors with the following features:

• The ability to define and re-use styles - This is a feature that is present in SGML but was excluded from HTML in the interest of efficiency (remember, bandwidth was (and still is) scarce).

• A standard way to make non-standard extensions - This is also done through the "style" mechanism mentioned previously.

• Support for scripting - Most notably, the ONMOUSE-OVER/LEAVE/OUT attributes which allow you to handle mouse events.

• A graceful transition from HTML to XML - XML supports default styles for the existing HTML tags that you have already learned. This is particularly bright of them as an abrupt transition would have been rather painful.

You can look at styles and how they are used in the styles section of the HTML 4.0 documentation.

Beyond ASCII

The Shortcomings of ASCII

One last little problem with Internet communications is ASCII. As many of you already know, the 'A' in ASCII stands for 'American'. Well, as it turns out, people other than Americans use the Internet too, and a lot of them use different characters than the Roman/Latin set. What's worse, the 8-bit character format does not allow for more than 255 characters, and whoops! all of those spots are taken.

The Answer: Unicode

A new, universal (well, global anyway) character set based on a 16-bit size has been created. This allows for 64 million different characters, which is enough to hold not only the existing Roman/Latin characters (still available at their original 255 slots), but other character sets like Arabic and Hebrew, and very large character sets like Kanji.

Changelog