MBUI Telecon 2012 October 11
Chaired by: Gerrit, scribed by Dave
Present
- Gerrit Meixner, DFKI
- Dave Raggett, W3C
- Cristina Cachon, CTIC
- Javier Rodriguez, CTIC
- Ignacio Marin, CTIC
- Heiko Braun, Red Hat
- Jean Vanderdonckt
- Vivian Motti, UCL
- Pascal Beaujeant, UCL
- Nesrine (UCL)
- Gaelle Calvary, LIG
- Sebastian Feuerstack, USC
- Javoslav Pullman, Fraunhofer
- Fabio Paterno, ISTI
- Carmen Santoro, ISTI
- Davide Spano, ISTI
- Paolo Bottini, University of Rome
Regrets
- Joëlle Coutaz, LIG
Planned Agenda
- Current status of the Abstract UI document / recent email by Heiko Braun /
Paolo's introduction
- Current status of the Working group note / Use Case refinement
- Current efforts of the OMG concerning the Interaction Flow Modeling
Language (IFML) / cooperation?
Current status of the Abstract UI document
Jean: new definition of the generic requirement just before the section on the definition of the user. Completeness, consistency, extension of concerns, and extensibility
Thanks to Sebastian for providing new section on modelling. The consolidation is ongoing. Any questions?
none raised
Thanks also to Sebastian for the list of questions at the end of the document which we can discuss in detail in Lyon.
Gerrit: sounds very good, and if there are no objections we can proceed as you suggest.
Jean: if you have any further open questions please add them to the end of the document, or to make other proposals in the document.
Sebastian: could we try to prepare a stable version prior to the Lyon F2F?
Jean: perhaps on Thursday 25 October?
several voices of approval
Jean: please feel free to comment on the requirements as there is a completely new section there.
Current status of the Working group note
Gerrit: any questions concerning the introductory document?
Jaroslav: there was a discussion on new graphics for the simplified Cameleon reference framework.
Gaelle: in favour of simplified version, but want to keep the notion of "context".
Jean: if might be interesting to have separate sections for users, context, etc. as these are mixed in the current version.
Gaelle: if it helps I can provide a proposal.
Paolo: are we adopting the Cameleon reference framework in this document or is it just one approach?
Jaroslav: we have a section on user models
Gaelle: what about the context?
Jaroslav: that is harder as it is application domain specific. That favours a black box approach as you suggest, where we leave the modelling of the context to other work.
Fabio: that isn't a particularly important aspect of this document. We should focus on the benefits of the model based approach.
Jaroslav: this could be useful for other working groups, e.g. for accessibility
Jean: we could create a minimal model for the user and reference the literature for richer models.
Jaroslav: we should explain why run-time adapation is important, but we don't need to provide a detailed model beyond what Jean suggests.
Fabio: a brief explanation is sufficient.
Jaroslav: we should be more concrete in respect to our examples/use cases.
Fabio: yes, that is important.
We need to know who is responsible to do what, otherwise, we won't progress.
Gerrit: in my opinion, the people who provided the use cases should refine the description.
Jean: overlap between some use cases
Gaelle: each use case needs characterising in the table, so that we can decide what we need.
Jean: I see that the table is really relevant
Fabio: the table focuses on adaptation rather than the model-based approach, do we need another table?
Gaelle: either we focus on the state of the art or we envision interactive systems for which model-based approaches is key
Jean: the table should have 2 sets of criteria separating the goals and the means
Fabio: that's a good idea. This document is aimed at people who are unfamiliar with model-based approaches, and we need to cover the features and the results.
Fabio: how should we proceed. Extending the table would be a good start
Jaroslav: we extracted a list of benefits from the Kaiserslautern F2f, but I need to compare with what we have now.
Jean: the goal is to start with high level goals and a more extensive structured list.
Gerrit: you can send me the document you wrote with Vivian, perhaps you could send it to me, and I will attempt to harmonize and extend it.
Task Model document
Jean sent a reference to a taxonomy of tasks.
Fabio reviews. The main question is whether a task is automatic, interactive or something internal to the user.
We need to understand why we want a classification of different task types. We also need to allow for extensibility to allow people to add further task types as needed.
Jean: the goal is to provide guidance and as much precision as possible.
Fabio: we can of course evaluate your reference to how effective the types are
It is easy to find new task types, but it is a challenge to have a good set that isn't too long
Jean: there may be some useful new types for us to consider. We have 2 criteria: precision and scalability.
Fabio: they are in conflict
Jean: for the moment, we don't force people to use task types, right?
Fabio: they can choose whether to use the predefined types or to come up with their own.
Jean: but the task type itself is optional
Fabio: yes
Paolo: for interoperability we want to encourage the use of a common set of task types
Fabio: if we find a good reason to add/refine types we can do so
Current efforts of the OMG concerning IFML
deferred to next week's call
end of meeting