Crescencio Lima

Teaching Globally Distributed Software Development: An Experience Report

Companies around the world routinely distribute their software development across different sites. Students, however, rarely get a chance to learn the potential problems that arise, and the potential solutions to those problems, when conducting distributed development. It is especially difficult to simulate the situation for students when development is distributed across time zones and cultures. We have developed a course that requires teams of students at widely separated universities to collaborate with each other to complete a software development project. Instances of the course have been presented four times using combinations of five different universities, and we are seeking to create a larger pool of universities interested in and capable of presenting it. This paper discusses our goals, the characteristics of the course and the results of teaching it, with a primary result that all the universities want to and will offer the course again.

Investigating the Recovery of Product Line Architectures: An Approach Proposal

Due to the complexity of managing architectural variability, maintaining the Product Line Architecture (PLA) up-to-date and synchronized with the project source code is a key challenge. Moreover, allow the variability traceability in architectural level of large-scale projects is a costly task. The systematic use of Software Architecture Recovery (SAR) techniques enables PLA recovery and keeps the PLA aligned with development. In this context, we present an initial proposal that consists of an approach to recover PLAs based on SAR techniques. We performed literature reviews and exploratory studies to investigate the relationship between SAR and PLA to identify gaps and define state-of-the-art. Learn how to combine SAR and PLA is an important strategy to address some issues of PLA design. We identified that few studies address architectural variability and provide empirical evaluation For this reason, more empirical research is still necessary.

Product Line Architecture Recovery: An Approach Proposal (Extended Abstract)

The Product Line Architecture (PLA) is an important asset for the success of Software Product Line (SPL) projects. Due to the complexity of managing the architectural variability, maintain the PLA up-to-date and synchronized with the project source code is a hard problem. The systematic use of Software Architecture Recovery (SAR) techniques enables the PLA recovery and keeps the PLA aligned with the development. In this context, we present our initial proposal that consists of an approach to recover PLAs based on the use of (bottom-up) SAR techniques. We performed some studies (such as surveys, literature reviews, and exploratory studies) to investigate the relationship between SAR and PLA to identify gaps and define the research area state-of-the-art. The combination of SAR and PLA is an important strategy to address some issues of PLA design. We identified that few studies address architectural variability, PLA variability traceability, and empirical evaluation such as experiments, surveys, mixed-methods, and so on.

A Systematic Review on Metamodels to Support Product Line Architecture Design

Product Line Architecture (PLA) design is a key activity for developing successful Software Product Line (SPL) projects. PLA design is a difficult task, mostly due to the complexity of the software systems that SPLs deal with, and their variabilities. Metamodels have been used to support the representation of assets that compose a PLA, SPL variability and the relationships among them. The goal of this study is to characterize the use of metamodeling on PLA design, aiming to identify the main characteristics of metamodels, the elements used for PLA and variability representation and trace the evolution of metamodels. We conducted a systematic literature review to identify the primary studies on the use of metamodels in PLA Design. Thirty-five studies that proposed metamodels to support PLA design were selected. The review main findings are: (i) it is difficult to identify the existence of research trends because the number of publication varies and metamodels lack standardization; (ii) several metamodels support feature representation; (iii) the majority of studies addressed variability representation with variation points in UML diagrams; and, (iv) five evolution lines that describe how metamodels evolved over the years were identified.

MERCI: A Method to Evaluate Combinatorial Interaction Testing Tools for Software Product Lines

Testing a system is a routine activity, and it plays an important role in the software quality assurance process. However, testing highly-configurable systems, such as Software Product Lines (SPL), is a rather complex activity, due to the presence of variability in its engineering process, which increases the number of product configurations to test. The underlying idea to make testing feasible in SPL engineering is to select a small but representative subset of products to test, by employing techniques such as combinatorial interaction testing (CIT). This paper presents Method to Evaluate Combinatorial Interaction (MERCI), a novel method to evaluate the adequacy of existing CIT tools for SPL engineering, with respect to three measures: defect detection, test coverage, and test execution length. We carried out an empirical evaluation to compare four CIT tools: ACTS, CATS, PICTMaster and VPTag. The results show that the method may serve as an affordable strategy to evaluate how the CIT tools could behave in an SPL testing scenario.