-------------------------------------------------------------------------------- Title: Interest statement for W3C Workshop on Web Device Independent Authoring Author: Peter Ferne, petef@atg.com Version: 0.1 Date: 2000-09-20 -------------------------------------------------------------------------------- 1. Background ATG (http://www.atg.com) is the vendor of the Dynamo platform, which includes: an open J2EE compliant application server; a personalization server; a 'scenario server' used to track users across sessions and channels; and a commerce server. I am a consultant/developer working with ATG clients and partners across Europe assisting them through all stages of the lifecycle from architecture to deployment. I have been working with wireless Internet technologies on and off for the last couple of years. Some brief notes on issues which I would like to see addressed follow. 2. Issues There is an increasing use of personalization on the Web. Users have come to expect to be able to customize their view of a web application, at least at the level of filtering and prioritising the display of content. Increasingly markup languages such as HTML and WML are being used not to structure static documents but to define the GUI to a web application. Some very successful sites, such as Slashdot, consist largely of user generated content. Skinnable applications have been popular for some time. With the advent of XUL support in Mozilla they are becoming genuinely useful and starting to blur the line between the desktop application and the web application. Inconsistencies between devices of the same type, e.g. WAP handsets, leads to the need to transform/render content on a per device basis rather than per format. The rapid proliferation of devices, formats and platforms creates a need for the ability to add support for a new device type post facto. There is a widely held belief that it is important to deliver a coherent user experience across multiple channels. There is increasingly a need to tailor the *flow* of an application, not just the rendering and filtering of content. For example, a registration form which is presented as a single page to a desktop Web browser might need to be split into two screens for presentation on digital TV and into four for presentation on a mobile handset. There is a need to do transformation in a manner which provides acceptable performance. This might indicate a requirement to be able to stream the resulting output whilst transformation is still underway. 3. Approaches Transformation is typically done on a per document basis. There is a tension between providing transformation hints within the source document, which allows greater control over the rendered output, and providing hints in an external document, which may allow for the easy addition of new formats and devices. In the field transformation logic is often embedded in application code rather than encapsulated in annotations. -------------------------------------------------------------------------------- Ends