From zedwiki

Note: this meeting will be held at 1300h UTC. Find the meeting time in your geographical location.

Contents 1 Scribe 2 Present 3 Regrets 4 New action items 5 Previous Action Items 6 Agenda 7 Minutes 8 Previous Minutes

Scribe

Per

Present

Boris, Dennis, James, Josh, Marisa, Markus, Sam, Per

Regrets

Kenny

New action items

• Everyone to make a decision about XForms and QTI

Previous Action Items

• James and Markus; come up with a plan for dealing with the limbo in spec prose -- possibly a switch in the spec with two alternate versions
• Dennis/Josh add the 2 examples that Boris mentioned (20090427 concall) to Wiki.
• Boris send image(s) for the examples mentioned to Dennis and Josh.
• Dennis/Josh contact QTI and ask about replacing the HTML with host language text content set.
• Markus to add spec prose on cascading level for inline styles: "For the purposes of the CSS cascade, these values are considered author-level"
• Markus to create wiki page devoted to level issue
• Markus to invite people for profile testing
• Markus to ask the DAISY list about how newspapers/periodicals markup is currently being done.

Agenda

Todays main agenda item: continued discussion on XForms/QTI and the forms feature.

• Email from James:

Markus & I were talking about QTI, XForms, and interactivity offline and I had a 
couple of thoughts about this to help put things into perspective.

I think that it is important to recognize that we are looking at this point to 
come up with a module/feature that will plug into the book profile, and  perhaps 
other, similar profiles for encoding print-based materials.  Given that, is the 
ultimate production of fully interactive electronic tests a requirement for this?  
I tend to think not.  Instead, I think that our scope is really limited more to 
identifying form-like structures within the print materials.  Having a completely 
automated, seamless translation of this to an interactive web application is probably 
reaching higher than we need to at this point.

We could, of course, have a completely different profile for interactive tests that 
would possibly use a totally different markup.  I could see it being based on QTI 
as something like a host language, for example.

Given that, for our purposes, I'm leaning towards using XForms elements to represent 
basic items, and perhaps developing some @role values based on QTI (and/or others) 
to apply testing semantics to "regular" book 
markup  (<li role="testItem">What is the capital of Assyria?</li>).

James

Minutes

Josh: There are som new examples on the wiki (examples 8 to 10) with both Xforms and QTI implementation

Josh: QTI has more options, especially for testing environments, however for ZedAI book, XForms seems to be the better choice.

James: An all purpose feature, handling all kinds of interactivity, seems to be to much to ask from a feature. Could rather be a separate profile. For the book profile, a less powerful system (Xforms) should be sufficient.

Markus: Thre seems to be two separate levels:

• An interactivity profile (hopefully to be developed in the future) to handle a variety of markup.
• A more limited interactivity feature to be made available to the book profile.

Markus: QTI seems to be well developed for handling exams, is that a target for ZedAI book?

Dennis: What kind of problem are we trying to solve. QTI very well for tests and interactivity. If you need an accessible version of a test, you would probably look for a QTI player, rather than a DAISY player.

Markus: Some people would like to see DAISY expand beyond the domain of publishing.

Markus: Nobody disagrees that the scope should be republishing of print. We seem to have the following options:

• QTI as a module. QTI is not written to be used as a module.
• Use a kind of subset of XForms.

Josh: With XForms, it is very easy to modify between editions of a book.

Markus: We could live with QTI and XForms as parallel solutions for a while. But, we could also investigate one further, without necessarily making a final decision. It would be nice to have a very early version of this feature for the current iteration.

QTI pros:

• The semantics (of the interaction) and terminology matches the requirements of a print publisher
• The large number of controls it supports
• Major publisher uses QTI for test material. They are already familiar with it.
• Very specialized with a focus on testing

QTI cons:

• Not suitable for injection in a host language
• Uses HTML as a vocabulary inside QTI
• Not so usable in a general context, such as a text book with a few questions on the page. Harder to integrate with other markup. It's more like an application on it self.
• QTI is so specialized, that it seems difficult to transform this into something similar to what we normally present to the user.

XForms pros:

• Similar to HTML forms, so familiar and easy to learn
• Can be trimmed down to only include what is needed in the profile
• Fewer controls, less complexity
• Written to be modular
• In theory, it's a more economic solution, in the sense that it may also be used in a distribution format.

XForms cons:

• Not everything may be represented with XForm, see ZedAI_XForms_QTI_Example_5, ZedAI_XForms_QTI_Example_6 and ZedAI_XForms_QTI_Example_7
• All models must live in the head of the document. With a text book with many question, the head will be very large.

James: A question in the book, doesn't have to be marked up with a language like XForms or QTI. No interactivity is needed. A @role could be enough

Markus: A @role should make it possible to transform into some interactivity elements in the dist format.

Markus: Will a few questions in a book, end up with some interactivity in the distribution format (say DAISY-book)?

Josh (or Dennis???): Yes, but it will of course depend on the type of user of the book. With that in mind, we need some format that makes it possible to create that kind of interactivity in the distribution format.

Markus: We should make a decision this week.

Previous Minutes

ZedAI_telcon_20090427