by Deepak Dhungana, Paul Grünbacher, Rick Rabiser
Abstract:
The variability of a product line is typically defined in models. However, many existing variability modeling approaches are rigid and dont allow sufficient domain-specific adaptations. We have thus been developing a flexible and extensible approach for defining product line variability models. Its main purposes are to guide stakeholders through product derivation and to automatically generate product configurations. Our approach is supported by the DOPLER (Decision-Oriented Product Line Engineering for effective Reuse) meta-tool that allows modelers to specify the types of reusable assets, their attributes, and dependencies for their specific system and context. The aim of this paper is to investigate the suitability of our approach for different domains. More specifically, we explored two research questions regarding the implementation of variability and the utility of DOPLER for variability modeling in different domains. We conducted a multiple case study consisting of four cases in the domains of industrial automation systems and business software. In each of these case studies we analyzed variability implementation techniques. Experts from our industry partners then developed domain-specific meta-models, tool extensions, and variability models for their product lines using DOPLER. The four cases demonstrate the flexibility of the DOPLER approach and the extensibility and adaptability of the supporting meta tool.
Reference:
The DOPLER meta-tool for decision-oriented variability modeling: a multiple case study (Deepak Dhungana, Paul Grünbacher, Rick Rabiser), In Automated Software Engineering, volume 18, 2011.
Bibtex Entry:
@ARTICLE{Dhungana2011,
author = {Deepak Dhungana and Paul Grünbacher and Rick Rabiser},
title = {The DOPLER meta-tool for decision-oriented variability modeling:
a multiple case study},
journal = {Automated Software Engineering},
year = {2011},
volume = {18},
pages = {77-114},
number = {1},
abstract = {The variability of a product line is typically defined in models.
However, many existing variability modeling approaches are rigid
and dont allow sufficient domain-specific adaptations. We have
thus been developing a flexible and extensible approach for defining
product line variability models. Its main purposes are to guide stakeholders
through product derivation and to automatically generate product
configurations. Our approach is supported by the DOPLER (Decision-Oriented
Product Line Engineering for effective Reuse) meta-tool that allows
modelers to specify the types of reusable assets, their attributes,
and dependencies for their specific system and context. The aim of
this paper is to investigate the suitability of our approach for
different domains. More specifically, we explored two research questions
regarding the implementation of variability and the utility of DOPLER
for variability modeling in different domains. We conducted a multiple
case study consisting of four cases in the domains of industrial
automation systems and business software. In each of these case studies
we analyzed variability implementation techniques. Experts from our
industry partners then developed domain-specific meta-models, tool
extensions, and variability models for their product lines using
DOPLER. The four cases demonstrate the flexibility of the DOPLER
approach and the extensibility and adaptability of the supporting
meta tool.},
doi = {10.1007/s10515-010-0076-6},
keywords = {CD Lab ASE}
}