See also: IRC log
<scribe> Scribe: dsr
see http://www.w3.org/Project/IRC/ for irc client info
<cgi-irc> test
Fabio has added a short section on handling errors in task models, see http://www.w3.org/2005/Incubator/model-based-ui/wiki/Task_Model#Describing_Errors_in_Task_Models
This was in response to comments in the last telecon.
Fabio gives a lengthy explanation
Can you model behavior in response to external input (other than from users)?
Fabio: we would use a different icon to indicate input from the system.
The use might be editing and gets an external notification that must be handled immediately.
Fabio: this can be treated in terms of concurrent behavior
Dave asks whether events at this level are similar to external notifications?
Fabio: yes, but we want to keep the descriptions at a high level so don't want to cover lower level events
Dave asks about the discussion on extensions to CTT, Jaro and Jose agreed to work on that in last week's call.
Jaro explains how that involves the relationship between CTT and statemodels
Jaro asks about the positioning of state charts (SCXML) and other work on modelling behavior
Statecharts provide a more concrete description of task models.
Dave wonders if StateCharts provide a means to describe dialog at the Abstract UI layer.
What purpose does specifying a task model serve?
Fabio explains ...
Task models can also be used to describe and hence explain the user's behavior as part of collaborative UI
Jaro notes that state charts are independent of specific implementations
Daniel: there may be some confusion when we use the term state charts.
What is missing is what are the events and what state/abstractions are being represented.
Jaro: task models involve higher level events than are involved in the abstract ui
Fabio: we have to be careful about comparing expressive power. The power needed depends on the goals.
Dave: perhaps we ought to bound this discussion and perhaps we need to apply specific use cases to show how the task model and abstract Ui support UI design
Daniel: we should list and
describe the events at the two levels
... some people think of task models more concretely, and see
CTT as at a higher level of abstraction.
The idea of business events, tasks and data models versus UI
Fabio: task models can be considered as a missing link between business models and abstract UI
Jaro talks about system tasks.
versus user tasks, What does CTT support?
Fabio: CTT caters for three kinds of tasks: user cognative, user input and system
Dave asks what people plan to do over the next 2 weeks
Jaro: will continue work on state charts
Fabio: will look work on CTT extensions
Daniel: I have a pending action to work on the use cases and I can also add a page on different kinds of abstractions
Gottfried: for my presentation, I could talk about two topics.
Fabio: I can't make the 9th could we hold the call on the 16th?
<DS> As I said, it would be difficult for me to join on the 9th, I'll be away then; I can make it on the 16th.
Gottfried, I will be in Berlin on the 16th, but could call in.
<Gottfried> ANSI/CEA-2018 Task Model Description standard
<Gottfried> Universal Remote Console framework - abstract user interfaces
<DS> Gottfried, did you see http://www.w3.org/2005/Incubator/model-based-ui/wiki/Layered_Architecture
<DS> it might be a good idea to relate it to this
Gottfried: I would prefer to take the first topic (task model description language) and defer the second topic to a future call.
Dave will email the list with the proposal to have the next call on the 16th and to feature Gottfried's talk.
We also need to pin down the next face to face. Daniel notes that TPAC is in March in Boston, but perhaps the XG is more European centred.
Daniel: suggest we try to co-locate with a relevant event
Daniel mentions a conference in Spain on Web engineering in June in San Sebastian
Dave could look into possibility of space at ERCIM in Sophia in June, but will need a specific date.
<DS> ICWE in San Seabastian June 24-26th
Daniel: San Sebastian conf Jun 24-26