XML 2001 - December 11, 2001

From HTML to XML

Speaker:	Wendell Piez
Track:	Publishing
Day:	Tuesday December 11, 2001
Time:	4:00pm - 4:45pm

Many organizations have made significant investments in creating web sites presenting considerable amounts of information; others are planning to do so. Many developers and managers have become aware of the challenges and difficulties of creating web-based publications, as well as their promises. What may not be so familiar is the promise of newer XML-based technologies, both in giving us new tools for handling familiar problems, and in supporting operations, such as information reuse and repurposing or %22cross-media%22 publishing, that have been difficult to accomplish on the HTML-based web.A key insight is that HTML and XML are not direct competitors. The aim of HTML is straightforward: simply, to format information for display on a web page. In contrast, XML provides a format that is both rigorous enough to be processed by machines, yet also open-ended enough to allow the description and organization of electronic information for potentially *any* purpose -- not just display on the web, but also indexing, searching and cataloging publishing in other forms such as print, e-books, or audio, complex hypertext (beyond what is practical to engineer in HTML), analytic operations such as %22data mining%22, etc. The bad news is, that with every new application developed, we face new problems. In order to make sound strategic decisions, the new technologies have to be well understood, both at a high level (what are they capable of doing?) and occasionally in some detail (what can be done this month? what can be done this year or next year? what tools are available and what do they cost?). The good news is, that XML does not make the current web obsolete: we still need HTML to do what HTML does and in many cases the most rewarding course to take may be a gradual one.If XML and HTML are not direct competitors, how do they relate to each other? For us to determine how we will migrate data and what we can achieve by doing so, requires knowing what they actually are as formats. This presentation examines a range of different approaches to encoding, including varieties of both HTML and XML, as they apply to a single test case. The relative advantages of different kinds of HTML and XML can thus be seen directly, along with the challenges that are posed, and opportunities that arise, from different migration strategies. Specific notice will be made of tools and techniques that can help do the job.