Summary of the first meeting

Presentations

  • Eddy’s presentation on AOP to the design of self-adaptive architectures
  • Jorge’s brief summary about feedback loop in programming language approaches.

Brainstorming

  • It would be interesting to hear the feedback loop story from people working with it at the UCL (e.g. Boris)
  • We need a taxonomy for self-adaptive systems
  • In this group we could also cover how to map the domain analyis modeling to the domain architecture modeling level(e.g. the modeling of the time-dependent adaptations (as in CODA this is partially already the case): how can this be supported at the implementation level?)
  • An interesting idea for self-adaptive systems would be to combine a crosscutting pointcut mechanism (such as the one of aspect technology, more specifically trace matches) with a crosscutting advice mechanism (such as layers in ContextL or context in Ambience).
  • We could see how Plonemeeting could help us to define an architecture for self-adaptive systems.
  • The goal of this group is to define a software product line for self-adaptive systems

For the next meeting

  • Alfredo gives a presentation about his master thesis
  • Each one of us (or by university) describe a first attempt of a taxonomy of self-adaptive systems. For some members of the group this taxonomy will be first based on context-aware adaptations. This taxonomy should enable the classification of existing techniques for self-adaptiveness.
  • Each university should bring an case study in order to provide meat/inspiration for the taxonomy. Suggestions: Plonemeeting (VUB), Cell phone (VUB-UCL), dynamic organization (KUL), Alfredo’s example (of his thesis).
  • This is not just for fun! An initial target for our results would be the Journal on Software Evolution.
wp3/view.txt · Last modified: 2009/02/18 17:44 by jorge.vallejos