Growing Ideas at XML 2001
December 19, 2001
The XML 2001 show floor displayed a wide variety of different offerings from over 50 vendors. Much of the show featured improved versions of what had come before, but there were some new ideas sprouting as well as different takes on older ideas.
There was a much broader range of XSL Formatting Object implementations, with RenderX, AntennaHouse, ArborText, and Advent 3B2 showing off their XSL-FO support. ActiveState's Visual XSLT Debugger tools gave many developers hope of managing to debug their XSLT stylesheets. SVG was also a common topic of discussion on the floor, as the notion of XML for graphics seemed to be catching on.
Like last year, the show also offered a special area for new technologies from developers eager to bring their new creations out into the world. The "Incubator" area had a wide variety of new tools solving new problems, or solving old problems in new ways.
Topologi Shows Off a Different Kind of Editor
While many vendors offer editors which hide the markup or work as trees, Topologi is developing an editor that takes a very different approach. The Topologi markup editor keeps all of the markup accessible to writers while providing support for a wide variety of common tasks and teamwork.
Aimed at customers who already intensively use XML or SGML, the Topologi editor is designed for people "who want more exposure to angle brackets. They want to be close to the markup," says CTO Rick Jelliffe. Topologi watched teams creating and editing marked-up documents. They examined the flow between team members and looked at both how to help them do their jobs and how to get out of their way.
Ideally, "If someone's running another text editor, we want them to be up and running with ours in one minute," said Jelliffe. The rest of the features, which include peer-to-peer collaboration, extensive support for a wide variety of character encodings, packaging through dzip, support for documents which may not even be well-formed XML, search-and-tag, paste-as-blank-markup, and a feature for creating markup through highlighting, are designed so as not to interfere with the basic operations of text and markup editing.
An initial beta of the editor will be available in January, with a smarter and more schema-aware version arriving a month later. True to Jelliffe's work in creating Schematron and a variety of other schema languages and tools, the schema tools will support a wide variety of schema languages.
GooseWorks Produces Open Source Topic Map Processor
GooseWorks was demonstrating an Open Source Topic Map toolkit implementing TMPM4 (Topic Map Processing Model 4). The toolkit is written in C and has front-ends for Python, Ruby, and other languages. Taking a SAX-like approach to Topic Map processing, the engine parses Topic Map documents (and some other similar formats) and generates a sequence of events that together form the graph described by the Map (or Maps).
Recalling the many times when he'd found James Clark's nsgmls tool to be incredibly useful, Sam Hunting noted, "We're following the nsgmls model in that there's lots of detailed output." Developers can load, merge, query, and process Topic Maps.
GooseWorks unabashedly focuses on graphs, not syntax. Hunting noted, "The quest for perfection is not realized in angle brackets," and the toolkit's ability to also work with NewsML, NITF, RDF, and XHTML suggests general applicability beyond the initial Topic Map focus.
YAWC Pro Simplifies Word-to-XML Conversion
YAWC Pro's demonstration was right next to Topologi's, but it went in a very different direction. Instead of giving users direct access to markup, it lets users create documents in Word 97 or higher, export them to clean XML, and then convert that XML to whatever final format they need.
YAWC Pro uses Word styles as a foundation for its conversion, and can also capture metadata in Dublin Core. Its built-in accessibility functions make it possible to create much more accessible HTML files than the standard Word output, while its configuration tools make it easy to create a variety of XML outputs from the same XML document. YAWC Pro also supports XSLT post-processing, to automate conversion of YAWC Pro's XML to whatever XML form is needed.
YAWC Pro and the Topologi Markup Editor may actually fit very well together, with authors creating initial drafts in Word, converting them to XML using YAWC Pro, and passing them to markup specialists using the Topologi tool.
Barbadosoft Battles XML
To judge by the fliers in the Barbadosoft booth, they don't think XML is a very good idea. Headlines like "'XML impossible to model,' say experts," "Top reasons for not using XML," and "Third-generation XML-based applications 'worse than Cobol'" suggest a large set of problems.
Identifying a set of problems seems to be the first step toward developing a set of solutions for Barbadosoft, however. The recognition that XML offers its own set of challenges lies at the heart of their new CorteXML product, which attempts to "enable the application of sound software development principles to XML by managing the conceptual level of design and providing unlimited extensibility."
Versioning and change management, two aspects of XML development which frequently bedevil organizations, are at the core of the Barbadosoft toolkit. Managing applications when data structures change has long been a complex task, especially when structures are built directly into code. The Barbadosoft suite, while insisting on schemas and validation, uses those tools to buffer applications from change rather than building in brittleness which makes change difficult.
Barbadosoft is aiming its efforts at large organizations with difficult modeling problems, and its framework is built on the Java J2EE platform and Java throughout.
Other Highlights from the Incubator
The Incubator area had a number of other projects which were getting started, on the verge of release, or presenting themselves to the XML world for the first time.
The folks from Curl, an Internet development environment most notable for stepping outside of the traditional HTML-centric browser, was showing off its newly added XML support. Based on SAX2, it simplifies the integration of Curl applications with XML information flows. Curl has also added SOAP 1.1 support for Web Services integration.
SIL International was showing off its tools for working with non-Roman scripts, including its upcoming (and open source) Graphite technology for rendering complex scripts, as well as its WorldPad text editor. Graphite supports TrueType fonts but also permits rule-based extension of their rendering.
At Bind Systems, business-centric modeling was the order of the day. By modeling business processes, rather than specific XML flows, Bind Systems hopes to make it easier for different levels of developers to work together. The initial modeling creates a reusable foundation that other developers can fill with detail, while the management software allows developers to examine transactions in various levels from an overview to document-by-document. BindPartner, a J2EE-based system, supports the ebXML process model, SOAP, and WSDL.
Finally, Media Fusion USA was demonstrating its high-speed native XML database, Sekaiju 1.0 (also known as Yggdrasill). Initially built for Windows, Sekaiju offers high-speed querying of large stores of XML documents using a custom-built query language called XBath. Sekaiju supports both valid and well-formed XML documents, permitting its use in a wide range of development environments.