Al Gilman <email@example.com>
We need actions that change the web content to be classified based on their _effect_ and not just their triggering events.
Users need to be able to allocate these classes into control level classes of things that can be done
automatically without notice [automatically with notice] after confirmation [ with default yes | with default no ] on no condition (JustSayNo) ( syntax: (comment or synonym) [option] [ choice | choice ] )
All action taking by scripts against document objects need to be governed by this user policy.
Classifying the actions.
[Legislating the user supervisory policy role in action taking]
In addition to dynamic HTML, Frames and Tables are areas where, in the field people still have problems. I think that WAI next steps on these topics should ideally be taken up in non-PF groups but the work needs to continue. It is more important for the work to continue than what WG we say it is in.
I think that there needs to be more work done to understand what causes the maintenance or breakdown of orientation in [especially audio] web browsing. This relates particularly to Frames and multimedia. It is related to work on tables because both are areas where the visual medium is heavily involved in maintaining orientation, and sufficient alternatives to the visual techniques need to be defined and provided for. We had a good discussion on the WAI-IG on frames. It showed that there is room for more and better knowledge to be beneficial. But I think that it will take experimentation with both document and browser at the same time to explore the candidate solutions right.
For tables, I think that we want to try to create suitable conditions so that the solutions adopted by "the people who come to the Future of HTML workshop with a primary interest in database publishing" are better than the half-baked solution we have in HTML 4.0 at the moment. To do this, we need to create a critical mass with database publishers, X-generation format and tool developers, audio browse tool vendors, and blind audio browser users all represented adequately. The answer may not need changes to HTML 4 but it does need more shared conventions between the data publishers and the browser builders. Even if these conventions are in interpreted media such as RDF schema or CSS stylesheet.