Html and Xml

If you knew SGML, you probably didn't like it. The "mother" of HTML and XML, SGML was huge (a 300-page specification), it had all the charm of an alpha personality, and was complicated, difficult, uncompromising--SGML never did learn to put on a stylish face. DSSSL, the standard applying style to SGML content, never met with commercial success. Since then, you've had a long-term relationship with HTML, but are finding its superficiality no longer satisfies. Recently you've been attracted to XML, flirting with the idea of getting to know it better. Now you're ready to take the plunge. XML seems to combine the best elements of SGML and HTML. XML publishing is far more disciplined and more powerful than HTML, but it seems less demanding than SGML. What would a commitment to XML be getting you into? Just this: XML requires a commitment to serious discipline, and likely a fundamental change in the way you do business. But with everybody wooing XML, those prenuptial agreements may be getting easier to accept.

Before we put down HTML, let's not forget: HTML's simplicity and flexibility spawned one of the biggest inventions in history--the World Wide Web. HTML won't go away any time soon, and HTML (like SGML) is based on Document Type Definitions (three to be exact). HTML was the dating phase of markup languages, and it charmed browser vendors into forgiving Web pages their sloppy code. Nonetheless, HTML's teen blemishes are becoming obvious. Just look at all the plug-ins and proprietary extensions demanded by a world that wants to go beyond brochure-ware and just another pretty HTML face.

Enter XML, much leaner than SGML (only a 30-page spec), yet promising discipline and flexibility impossible in HTML. XML insists that valid documents conform to a Document Type Definition (DTD) or document model, yet retains the flexibility you want and need. Follow the 30 pages of the XML specification, and you can develop a model with practically any combination of elements, attributes, and entities you choose.

Once XML invites you home for dinner, you're surprised to see how many standards sit at the table. You recognize SGML and HTML immediately, but who are all these others? Why there's the Document Object Models (level 1 and level 2), the Multimedia cousins (HTML+Time, Synchronized Multimedia Integration Language, SMIL-Boston Integration Language), and Namespaces. There are also standards for linking, querying, pointers, and even XHTML (Extensible HTML) to help you transition from dating HTML to an engagement with XML.

Not far down the table, and obviously uncomfortable, you notice XSL, the Extensible Stylesheet Language, who introduces itself as "just like DSSSL, only simpler." The other specifications snicker that XSL is "just like DSSSL, only later; it's still just a working draft." At this you shudder, wondering if SGML's troubles will recur. It's true that XML is only about 30 pages long, but if you count up all the pages in all the other specifications, it's even bigger than SGML's 300-page count. And you don't even want to go into the next room where an even larger number of XML cousins await: Dozens of Markup Languages (MLs, really DTDs) that every industry on the planet seems to be incubating. It's going to be really tough getting acquainted with this extended family.

What about life after infatuation? You're all for discipline, but will you really have to give up your word processor and WYSIWYG Web development tools?

Some vendors promise you can keep your old, favorite word processors