XML Principles, Tools and Techniques

By Bob DuCharme, XML Correspondent

Although the XML 1.0 specification was not an official, frozen document until February, and the XLL linking specification and XSL stylesheet specification still have some way to go, the fast growth of XML's popularity since the announcement of the first working draft in late 1996 has inspired publishers to rush books to market in the hope of taking advantage of this popularity. Printing thousands of copies of a book about a fast-moving target is a problem, and so far publishers have taken three approaches to dealing with it:

• Sweep the problem under the rug. Pretend that the remaining changes to the specifications are small, cosmetic changes, and that the big fat "Learn XML Really, Really, Fast!" books getting shoveled onto the shelves by most major computer book publishers really do contain "Everything You Need to Know About This Exciting New Technology!" These books will be obsolete six months after they hit the stores, but by then the publishers will have their first batch of XML book money and updated versions on the way. These books are not worth anyone's money.
  
  
• If XML is a simplified version of SGML, write a simplified SGML book, introducing elements, attributes, declarations, tags, content models, DTDs, and entities, without mentioning SGML declarations and the other SGML features omitted from XML. This can provide a reasonable introduction to XML for curious HTML jockeys, but such a book wouldn't offer much to the typical <TAG> reader.
  
  
• A snapshot of where XML and its related specifications came from, where they are, and what they may lead to in the future, ideally with contributions from people directly involved with the XML development process.
  
  

XML: Principles, Tools, and Techniques, edited by W3C Architecture Domain Leader and HTML inventor Dan Connolly, is the sole entry in this third category. It's the Winter 1997 issue of the World Wide Web Journal, and it's also a bound book published by O'Reilly and Associates. Although you can find most of this book's specifications, overviews, reports, and essays on the Web, Connolly's selection and organization of the material saves you a lot of trouble, and if you're going to have someone pick out the best stuff for you, few people are as qualified. The book is a well-organized mix of light reading and detailed technical descriptions that makes great browsing (in the old-fashioned sense of the word) while commuting or whenever you want to get away from your computer. The book is honest and forthright about the incomplete state of the XML, XLL, and XSL specifications at the time of its publication, and still has much information that will remain valuable for another year or two.

After an introductory editorial and a background overview featuring extensive quotes from the Editorial Review Board members who wrote the XML spec, the book includes a draft of the XML specification and several related W3C reports. While much of this section is already outdated, it still provides a good introduction to aspects of XML development that you may have heard about but not investigated yet. For example, Lauren Wood and Jared Sorensen's "Document Object Model Requirements" was my first look at this object-oriented framework for addressing documents; this object model will give a big head start to future application developers by doing much analysis and design work for them as well as making different developers' work more compatible and hence more easily combined.

Most of the book consists of the "Technical Papers" section. Plenty of these aren't so technical; Norm Walsh's "A Guide to XML" and the other first few pieces provide ground-up introductions for the XML newcomer. Some reports, such as Peter Murray-Rust's "Chemical Markup Language: A Simple Introduction to Structured Documents" and Charles Allen's " WIDL: Application Integration with XML" describe the development and implementation of early XML applications. Others offer helpful background to developers who are planning their own implementations: Michael Leventhal's " XML: Can the Desperate Perl Hacker Do It?" uses an example to answer with a qualified "yes," and Tim Bray's "An Introduction to XML Processing with Lark" gives the kind of hand-holding introduction to his Lark Java class library that makes it a logical first choice for someone surveying the widening selection of Java class libraries available for XML processing. "Building XML Parsers for Microsoft's IE 4," co-authored by several people closely involved with Microsoft's XML efforts, has some interesting background on Internet Explorer's XML support and on the features available to developers interested in taking advantage of this support.

The book even includes an opposing viewpoint: hypertext pioneer Ted Nelson's "Embedded Markup Considered Harmful" explains his assertion that the structural information represented by SGML, HTML, and XML markup would be more appropriately stored in a separate file with pointers to the relevant content.

W3CXML Working Group chairman Jon Bosak's " XML, Java, and the Future of the Web" is more visionary than the other XML-and-Java pieces, describing not how to code XML Java applications but rather the kind of applications that will become newly possible with the architectural flexibility offered by this combination of technology. By adding the distributed processing made easier by Java to the easier transmission of multi-purpose structured information allowed by XML (as opposed to the single-purpose data of formats such as HTML, RTF, PostScript, and Acrobat) we can create unprecedented new categories of applications. And the proliferation of Java class libraries for XML development mentioned earlier won't hurt!

The book's one weakness is its lack of an index. This may not fit in with the overall style of the World Wide Web Journal, but if someone hears about an interesting aspect of XML technology and wants some quick context, a table of contents won't always find the relevant information in a book that covers so much. Any nonfiction book, especially one on technical topics, benefits from an index.

Overall, however, the book offers a broad range of information from the people who know XML best, and beginners and experts alike will learn from and enjoy it.

XML: Principles, Tools, and Techniques, published by O'Reilly & Associates, ISBN: 1-56592-349-9; 266 pages, US $29.95,

Contact Information