Forschungsthemen
[GB] Building a Runtime Debugger for Behavioral Programs using Models@run.time
Behavioral Programming is a software development approach, which allows to incrementally develop applications in terms of scenarios described by sets of behavior threads, which interact in a special way with one another. This reduces the gap between problem and solution space, as scenarios as main building block in behavioral programming actually crosscut problem and solution space.
Models@run.time on the other hand is a software development approach promoting the use of models and modelling techniques at system runtime. Each system based on models@runt.ime has a key architectural constituent: a causal connection between a runtime model and the running system. Changes in the runtime model are propagated to the running system and vice versa. The main purpose of using a runtime model is to have an adequate abstraction of a running system for a given task, which could, e.g., be reasoning for self-adaptation.
The goal of this thesis is to examine how both approaches can be combined in a meaningful way. Behavioral programming provides means to incrementally develop applications in terms of scenarios, but the individual threads swiftly become very complex although running isolated from each other. Models@run.time could be used to get an abstract runtime view of such a system and, by this, help in debugging behavioral programs.
- Literature review on behavior programming and models@run.time
- Development of a runtime metamodel for behavioral programming
- Development of a runtime dashboard for behavioral programs using the runtime metamodel
- Realization of a case study
- Drafting a concept for further applications of the runtime metamodel
Betreuer: Sebastian Götz