My Notes from the Hong Kong W3C/WAP public workshop

These are my personal reflections on the meeting, and nobody else's opinion. To get a less idiosyncratic idea of what happened, consult the minutes.

The workshop was 5-6 September. There was a lot of participation from groups who are members of both consortia, and a few from groups who are only members of one or the other. A major focus for many particpants was Interactive Voice Response systems - at the moment typically automated services provided by telephone.

Major discussion topics were

• the integration of WAP (small screen, text and basic images, expected to have spec-based style implementations, but variation in properties available) content architecture with voice content architecture (voice only).
• Single authoring (write once, read on any kind of device) - what it means, is it possible, and how to do it.
• Multiple modality - having text, graphics, sound, and other forms available in diffferent systems.
• The need for use cases and usability research and data to drive development in the right direction for users.
• Where the work should be done (agreed that it is more important to get it done).

My perspective on the results were

• that single authoring is seen as a goal, but there is a perception among many of the participants (who are large organisations in general) that it was only going to be avialable to large organisations who could afford to build special purpose tools for their own needs. It is agreed that the ability to provide extra presentation control for specific media is important for large content providers
• Architecturally it is possible to converge the WAP and Voice areas, but the typical use cases for each are different - for example voice output needs to be short, whereas it is easy for users to provide a lot of input in one transaction. For mobile devices it is possible to provide more output in a usable format, but it is difficult to do anything but very simple input transactions. So it is not clear whether the goal is a single language that can be directly used, a convergence between WML and VXML so that they can be used in both milieux, or a different kind of source and use negotiation (for example CC/PP) to select an appropriate transformation (eg XSLT) for the target medium / user requirements. (My preference is strongly for the latter approach, since that is scalable into general accessibility and device independence. It puts some load on the author to learn to produce better content, and a lot of load on the tool developers to make authoring tools with better support for device independence, but that is what I am trying to get anyway...)
• Multi-modality is seen, by many of the participants, as essentially new. However there has been work in the area done in fields such as education and accessibility for people with disabilities for a long time.

So... WAI has quite a lot to offer - in the area of authoring for multi-modality, and usage scenarios / usability. We should follow this work, because mobile devices are important to people who are not housebound, and for people who are deaf or blind, for example, provide opportunities that were never available when people relied on a fixed public telephone. From a wider perspective it is important to make sure this stuff fits well with the architecture we know well, and to look at what we can learn from it to apply to that architecture.

Charles McCN, $Date: 2000/09/12 10:50:44 $