[Mirrored from: http://www.webweek.com/97Mar24/software/split.html]
XML Could Sidestep HTML SplitBy Liora Alschuler
HTML is the little markup language that could. Picked as an interim step toward a more powerful language by Web inventor Tim Berners-Lee, HTML has morphed gradually through an increasingly complex set of language extensions toward an all-purpose solution.
Particularly galling to the World Wide Web Consortium (W3C) is the growing disregard for HTML's original purpose. Comcomitant with that disregard has come the split in browser-specific HTML that makes it difficult to develop sophisticated, widely viewable Web pages.
At the heart of the HTML conflict are the contradictory needs of designers, who want tags that will make their jobs easier, and the standards bodies, which want to preserve the general application independence that makes HTML portable and powerful.
But now XML (eXtensible Markup Language), a draft standard from the W3C, offers some hope of a rationalized system for complementing HTML without creating the kinds of browser incompatibilities that currently threaten to divide the Web into Netscape and Microsoft camps. XML is an effort to recognize the desire to write new markup tags, but it lays down some simple rules for doing so, and turns user-defined markup into a force for stability, interoperability, and a powerful new breed of client-side processing applications.
XML is essentially a slimmed-down, Web-enabled version of the Standard Generalized Markup Language (SGML), the International Organization for Standards' "meta-language" from which the original HTML was crafted. But SGML's complexity has been a barrier to its widespread adoption. XML was specifically created as a way to simplify the language and create an alternative to HTML.
Major Web players such as Microsoft, Sun, Novell, Hewlett-Packard, and IBM sit on the XML Editorial Review Board (ERB), which launched its effort to define the standard last November. Noticeably absent from the effort has been Netscape, whose ultimate support will be crucial if the standard is to make it.
In the meantime, Microsoft has emerged as the most powerful proponent of the effort.
"All Microsoft understands that HTML has an extensibility problem," said Adam Bosworth, general manager for Internet Explorer, answering questions at last week's XML conference in San Diego. While he stopped short of announcing that XML will be in the next version of Internet Explorer (IE5), the emerging standard clearly has Microsoft support.
Both the Open Financial Exchange and Channel Definition Format--initiatives introduced by Microsoft and its partners, such as Intuit, PointCast, and others--are based on XML and will be supported by IE4. The W3C is expected to endorse an XML flavor of math markup, indicating growing support for the prototype standard.
Conference speakers from Microsoft to IBM to Sun to SGML vendor Omnimark described how XML can not only stop the proliferation of HTML flavors, but can also sharpen the results of search-engine queries, alleviate bandwidth constraints by reducing the number of round-trips between clients and servers, and provide personalized data for client-side processing.
Since more intelligence is in the downloaded data, not in the application, bigger pieces can be downloaded and crunched locally.
Web developers see the combination of Java and XML, combining portable applications with content that instructs for richer programmatic functionality, as particularly interesting.
One example is the Java Universal Markup BrOwser (JUMBO) built for the Chemical Markup Language (CML).
CML gives context to chemical data by adding tags for molecules <MOL>, atoms <ATOMS>, bonds <BONDS>, and formulas <FORMULA>. Such special-purpose tags would never be added by browser makers, but using the dynamic extensibility of XML, CML data can "teach" a browser how to display the new datatypes.
The XML advantage with perhaps the most far-reaching consequences for providers is the ability to personalize and automatically distribute mission-critical content. This point was emphasized by both Eric Seversen of IBM Global Services and Peter Lamb, a senior partner at Andersen Consulting. IBM has calculated that getting the right information to field engineers, a process enhanced by XML on the client, can save an engineer as much as an hour a day. Eric Skinner of Omnimark Technologies described how XML delivers more than 600 types of personalized installation instructions based on input of a username and a password.
But the most important aspect is that XML data remains stable as applications are upgraded. Since XML -aware applications "know" how to process the document type definitions that describe tags, documents are always displayable, even as browsers go through new upgrades.
Marketplace acceptance of XML will require more than the support of Microsoft, SoftQuad, and current SGML users. At SGML '96, James Clark, ERB member and technical lead for the standard, predicted that if neither Navigator nor IE integrated support for XML, the initiative would fail. Microsoft said it will support SoftQuad's implementation of XML, but Netscape is nowhere to be found on the issue.
Marc Andreessen, Netscape's vice president for technology, was unaware of XML when asked for comments at an Internet World press conference. He shuddered at the mention of SGML, however, indicating a typical industry reaction to SGML's complexity. Whether the simpler XML can overcome SGML's past remains to be seen.
Reprinted from Web Week, Volume 3, Issue 7, March 24, 1997 © Mecklermedia Corp. All rights reserved. Keywords: html standards Date: 19970324