Kevin Feichtinger, Chico Sundermann, Thomas Thüm, Rick Rabiser,
"It?s Your Loss: Classifying Information Loss During Variability Model Roundtrip Transformations. Proc. of the 26th ACM International Systems and Software Product Line Conference (SPLC 2022), Graz, Austria, ACM, 2022."
: Proceedings of the 26th ACM International Systems and Software Product Line Conference (SPLC 2022), ACM, New York, USA, Seite(n) 67-68, 9-2022, ISBN: 978-1-4503-9443-7
Original Titel:
It?s Your Loss: Classifying Information Loss During Variability Model Roundtrip Transformations. Proc. of the 26th ACM International Systems and Software Product Line Conference (SPLC 2022), Graz, Austria, ACM, 2022.
Sprache des Titels:
Englisch
Original Buchtitel:
Proceedings of the 26th ACM International Systems and Software Product Line Conference (SPLC 2022)
Original Kurzfassung:
In software product lines, variability models are used to explicitly capture commonalities and variability of a set of software systems. Many variability modeling approaches have been developed over a period of more than 30 years. Most of them are only described in academic papers, which makes it difficult to assess their properties and find the right approach for a specific use case. New approaches are developed regularly, adding to the ever-growing plethora of variability modeling approaches. Transforming variability models, i.e., of one type to another, would help to better understand and compare existing approaches and would also enable users to switch between approaches. Since variability modeling approaches differ especially in terms of scope and expressiveness, it is difficult to implement transformations without information loss. Thus, in this paper, we analyze concrete variability modeling approaches, present a mapping of key concepts between them, and identify and classify the information lost in one-way and roundtrip transformations. We evaluate their applicability by transforming different models of varying size and complexity using an existing implementation of transformations. We argue that our classification of information loss contributes to a better understanding of different variability modeling approaches, simplifies the comparability, and allows users to grasp the impact of transformations.
Forschungs-