Clone-and-Own software product derivation based on developer preferences and cost estimation

Abstract

Clone-and-own is a common reuse practice that is widely adopted for evolving a family of software systems. However, this practice loses its effectiveness if not supported with valuable indicators that guide the derivation of new products. In this paper, we propose an approach to support the derivation of new product variants based on clone-and-own, by providing the possible scenarios in terms of operations to perform to accomplish the derivation. We generate a constraints system prior to a product derivation, to facilitate the software engineer selection of the suitable scenario and operations based on his preferences. In addition, we propose a cost estimation for each operation and respectively for each scenario, thus, a software engineer can rely on it as an additional parameter to achieve the derivation. The proposed scenarios and cost estimation are based on indicators retrieved after an automated identification of the mappings between the features implemented by the family of software products and the assets in which they are implemented. We preliminarily validate our approach on a case study where results show that the provided support can considerably reduce the amount of time and efforts that can be required to achieve a product derivation. Index Terms-Clone-and-own, product derivation, software reuse.

Taming the Diversity of Computational Notebooks

“…In the longer term, we aim to evolve the SPL by involving data scientists in the enrichment process. This includes using automated reasoning techniques based on past configurations [5,14,24,35] and exploring model evolution [2,6,13,16,36]. Additionally, we plan to support multiple implementations within ML components using the Multi-Level Feature Trees approach [10,31].…”

Evolving software system families in space and time with feature revisions

“…An approach to support the composition of new variants based on opportunistic reuse, namely clone-and-own methodology, is presented by Ghabach et al (2018). It supports mappings between features and artifacts in an automated and incremental way.…”

“…A new product variant is needed when no existing variant implements the requested features, or when some of the features were modified to update existing variants. Thus, the composed product variant will contain the same features, but not the same implementation as before (Ghabach et al 2018). In this context, feature location techniques can ease evolution and maintenance tasks (Bennett and Rajlich 2000).…”

Featured based CAVE software factory

“…Given the above a multitude of similar but separate VR software products could be considered -with the variety of optional features to be implemented to satisfy the end user needs. With the traditional approach to handling that developers would reuse pieces of code (small-grained reuse) [Mor02a] or larger parts of another product to build a new one (clone and own approach) [Gha18a]. However, it may be considered ineffective for CAVE products, as two applications built from the same base would be deployed and maintained in the same CAVE separately, with high duplication of effort to fix possible errors in multiple products multiple times and often implementing independently the same enhancement in different ways in similar products.…”