Menu

Netscape Puts XML Support in Mozilla

April 1, 1998

Liora Alschuler and Mark Walter

Updated from The Bulletin: Seybold News and Views on Electronic Publishing, Vol. 3, No 26

New version supports XML document markup as well as a data and metadata syntax

By Liora Alschuler and Mark Walter
April 1, 1998

On Tuesday, Netscape Communications released on its Web site (http://www.mozilla.org) the source code for Mozilla, which, in days past, would have been known as Communicator 5.0. The release makes good on Netscape's dramatic promise in February to put its browser into the public domain.

But Mozilla also contains the surprise inclusion of XML functionality we had not expected until later this year.

Showing Mozilla to a group of developers at GCA's XML '98 conference last week in Seattle, R.V. Guha, one of the principal brains behind Netscape's XML development, drew resounding applause when he displayed an XML document in Mozilla, with XML tags clearly visible in view source mode. Guha affirmed Netscape's strong support for XML as a data and metadata syntax, and for XML as a rich form of document markup for the Web.

XML support in Mozilla takes the form of an XML parser (currently one written by Netscape, but soon to be replaced by James Clark's Expat), application support for Xlink (the XML linking specification in draft status at the W3C), and support for XML namespaces, a new W3C effort released in draft form on March 27 (http://www.w3c.org/TR/1998/WD-xml-names-19980327).

On the display side, Mozilla makes use of cascading style sheets to format text encoded in XML tags. In an interview, Guha stressed that Netscape is now emphasizing the use of XML to describe documents, not just metadata, as it did last year. Microsoft, which put a full XML parser and document object model support into Internet Explorer last fall, lately has been backing off on its XML promises. It appears that the next release of Office, for example, will save only the header information in XML; the documents themselves will be saved in native formats or HTML. In its browser, it is using XSL to transform XML into HTML, rather than using CSS to format XML directly. Though Microsoft makes the XML tags available at the API level, there's no question that Netscape's approach of keeping them in the source is preferable from both display and data mining perspectives

Another example of XML functionality is Mozilla's support for transclusions. In XML parlance, these are tags that mark an object to be included by reference. The adoption of a standard way of handling transclusions would standardize today's practice of server-specific server-side include statements, and makes possible client-side inclusions, which were previously unavailable to browsers.

Let's go!
Judging from the phenomenal number of downloads Netscape is delivering, the release of Mozilla source code is being enthusiastically welcomed by developers worldwide. If it does succeed, we expect publishers to be able to ride on the coattails of those developing useful additions to the product.

One early example is the response from the SGML community, which is known for its public debates and makes frequent use of noncommercial software. Guha said he expects Clark's Expat, a pun on Clark's Thailand habitat, to be compiled within Mozilla within the week. Clark's XML parser toolkit (http://www.jclark.com/xml/) checks for well-formed documents and parses external entities.

Wait a minute.
Still unresolved, however, is how Microsoft will respond. Netscape says it is willing to work with the W3C in developing a standard way of defining transclusions; we hope Microsoft will agree that this is a positive step forward and not pursue an alternate path.