Michael Nieke, new postdoc in SQUARE, will give a talk about his work on consistent feature-Model driven software product line evolution. Details below.
SPEAKER
Michael Nieke, Postdoctoral Researcher, ITU.
TITLE
Consistent Feature-Model Driven Software Product Line Evolution
ABSTRACT
Software Product Lines (SPLs) are a concept to manage complex families of variable software systems. SPL evolution is particularly challenging as it impacts all of its variants. Thus, thorough evolution planning in advance is well-advised. Feature models lean themselves as central planning artifact, driving the SPL evolution. However, typically do not go as intended and replanning is required. Additionally, feature-model inconsistencies and anomalies may arise. Finally, SPL evolution may impact behavior of existing products that have been derived from the SPL.
In this talk, I will present my Ph.D. thesis in which I devised a concept to track, enact, and plan feature-model evolution as driver for consistent SPL evolution. In particular, I will present:
A generic concept to model entire evolution timelines (past, present, future) of feature models.
– An efficient to guarantee feature-model consistency in presence of evolution timeline replanning.
– A method to reduce complexity of feature-model anomaly explanations by concentrating on causing evolution operations.
– A methodology to automate the process of updating product configurations semantically after SPL evolution.