How XML is changing applications

Extensible Markup Language (XML) is the latest Web technology to be hyped as a foolproof path into the future; but just what does it mean to XML-enable an application?
After being fast-tracked to approval by the World Wide Web Consortium (W3C) in February this year, XML has been incorporated into a diverse range of products, from Inso’s DynaBase publishing system to Microsoft’s forthcoming Office 2000 suite. Research firm Ovum has predicted that XML will also cause a resurgence in electronic data interchange applications.
Interest in the technology continues to grow locally as well. “It’s gone from a community where we personally know every single person by name, to large projects being undertaken by whole teams of programmers we’ve never heard of,” said Allette Systems’ Nick Carr.
Allette Systems is helping organise the SGML/XML Asia-Pacific Conference in mid-October 1998, and expects a substantial response, with double the number of last year’s attendees. Key local users include the government-developed Business Entry Point, which uses XML for data exchange.
XML is derived from Standard Generalised Markup Language (SGML), a long-established standard for marking data structures within documents. Data is marked with a series of tags, similar to those used in HTML (except that the new tags can be defined by the user at any time).
Two key features make XML easier to deploy than SGML in a Web environment. Firstly, it removes some of the more complex and arcane aspects of SGML. Secondly, it makes the use of a document type definition (DTD) — an outline of the structure of a particular type of document — optional, allowing documents to be created and parsed on the fly by applications which understand XML structures. (For a more comprehensive overview of XML, see APC September 1997, page 113)
The practical upshot of all this is that XML allows data (whether for online delivery or not) to be easily tagged, searched and indexed. Like its distant cousin HTML, this can be used to format the output of a document; XML makes use of the develop-
ing Extensible Style Language (XSL) standard to describe how data marked with a particular tag should appear onscreen.
XML’s status as one of this year’s hot Web technologies has led to some companies jumping onto the bandwagon without an overall strategy. “The sophisticated applications are the ones coming up from SGML, rather than those that are coming down from HTML,” said Carr. He also noted that the simplicity of using XML may have become exaggerated: “They think that it’s going to be easy because it’s different.” Nonetheless, companies are beginning to use XML in more creative ways.
For Lava Systems, which develops software for enterprise document management, XML provided the ideal tool to allow users of its products to access information from a range of data sources.
The Lava approach uses an n-tier architecture with three basic layers, all of which utilise XML data. A ‘repository adapter’ is used to convert data from specific proprietary formats into an XML document. This is then passed to the Lava applications server, which renders the data for either Lava’s proprietary client or viewing in any standard Web browser, using DTDs to format information consistently.
Lava vice-president of delivery and services Charles Latucca said that while the company actually began planning the
strategy before XML emerged as a standard, it was a natural fit. “XML is a nice format for us to communicate between applications, both our own and standard applications,” he said.
One key advantage of XML’s standardisation is that Lava can develop its own repository adapters for common applications (Exchange, SAP and JD Edwards are among the early candidates), but companies with more unusual configurations can also be catered for with custom adapters.
Tools for developing XML are also becoming more common. Last week. Object Design launched its XML Object Manager, which allows RDBMS development of XML applications for both clients and servers. “Server-side XML is going to become an increasingly important topic in the coming months as more companies realise that XML is not just about document data, but also application data,” said NC.Focus president )P Morgenthal.
The XML standard has led to the usual bickering between Microsoft and Netscape over which company has a better implementation. Explorer’s XML support predates the final version of the specification, meaning it has been on the market longer but may not incorporate recent changes. Netscape has only offered XML support since midway through this year, but the cross-platform appeal of XML and its browser products may make it a more natural fit.
Standards battles are also still being waged. The XSL specification has only just reached its first public draft, and some problems remain in its implementation. However, the importance of XSL shouldn’t be overstated, said Carr. “Formatting languages (such as XSL] are fundamentally different from modelling languages (such as XML],” he told APC. With SGML, Rich Text Format (RTF) has often proved an easier solution for formatting, he said.
Microsoft and IBM have further muddied the waters with Document Content E>escription (DCD), a recent proposal to replace the document type definitions (DTDs) used by both XML and SGML with a new scheme, itself written in XML. The DCD proposal has been submitted to the W3C.