ZedDist Interactivity Telecon20091028
From zedwiki
Wednesday 28 Oct 2100 UTC
Present: Keith, Dennis, Josh, Boris, Marisa, Kenny
Contents |
Agenda
From our last call, our actions items were:
@Marisa: math navigation example
@Josh: summarize xforms/QTI work
Status of Interactivity
Our three definite in-scope items are:
- Forms
- MathML
- End-user Annotations
We were also considering a generic navigation model that could enable author-extensible interactivity.
Important comment from several group members: The harder-to-specify features should not get in the way of the easier ones (such as forms, a must-have feature). We recommend breaking up interactivity into its components before beginning specification work.
Forms
For forms, we need:
- concrete examples of test/quiz/form content that we want to represent in DAISY 4 format
- a list of the types of controls that we want. this should also include form features such as submission.
Dennis: we can reference QTI/XForms zed ai work
@Keith: will send excerpts from "test ready" series by the end of this week
@Boris: will submit comments on QTI/XForms zed ai work by the end of this week
@All: list the types of controls required to represent the content examples. prepare this via email/wiki by the end of next week.
Math
Marisa presented list of math navigation actions
Boris: user might want a choice of how to hear the math (compact or verbose)
Dennis: render as braille or other tactile rendering
Marisa: these sound like user agent options. what is the requirement of interactivity here? maybe just that the specification contain enough information to be rendered in different ways by user agents or user agent plugins.
Keith: doesn't math already exist in DAISY?
Marisa: existing math-in-daisy is more of a specification/grammar solution. does not talk about how to interact with the content.
Boris: why can we manipulate math and not text/images/audio of the book?
Dennis: other activities might involve manipulation. overarching: need to idenfity significant chunks of the interactive content.
Marisa: an example of another manipulation activity: word sort.
Keith: we should recruit a mathematician! might be able to recruit some people from APH's network.
Keith: what are other ways (than ID ref) to refer to content chunks? xpointer?
Dennis: additional math features - import or export pieces to communicate with vendor extensions
Boris: math and forms must interact nicely with each other. user might need to enter mathematical expressions.
@Marisa: add relevant features to the math use case list
End-user annotations
We didn't have time to talk about this.
@All: please review the annotation use cases. Nothing new there though.
Navigation and author-extensible interactivity
Marisa: there are two options for extensible interactivity:
- interactive components must expose navigation and action options (aka "operations"). exposed items are then consumable by user agents.
- don't get so specific. say that any interactivity beyond what is built into Daisy 4 must follow certain accessibility guidelines.
Keith: my idea of navigation is based on virtual reality (go somewhere, have options, go somewhere else, have more options)
Boris: this is different from linear Daisy navigation
Dennis: hard to itemize all operations; maybe we could come up with a few basic ones (identify current location). also: use introspection to allow objects to say what operations they permit.
Marisa: too hard to require widget to expose all operations. the model of requiring API-like compliance (even via introspection) is too rigid.
Marisa: author-extensible interactivity would have to comply to some content guidelines (WCAG?) and also say what are the compatible target devices. e.g. a javascript animation to supplement a chemistry text is not going to be playable on a handheld audio device.
Keith: maybe this feature doesn't belong in DAISY. maybe is just an accessible application but not Daisy.
Marisa: maybe instead think about how to put DAISY inside other accessibility contexts (embed in web pages)
We haven't solved whether to allow author-extensible interactivity. Will bring our analysis to the whole group.
