Search

Context: Context-awareness challenges the engineering of contemporary software systems and jeopardizes their testing. The variation of context represents a relevant behavior that deepens the limitations of available software testing practices and technologies. However, such software systems are mainstream. Therefore, researchers in non-academic settings also face challenges when developing and testing contemporary software systems. Objective: To understand how researchers deal with the variation of context when testing context-aware software systems developed in non-academic settings. Method: To undertake a secondary study (Rapid Review) to uncover the necessary evidence from primary sources describing the testing of context-aware software systems outside academia. Results: The current testing initiatives in non-academic settings aim to generate or improve test suites that can deal with the context variation and the sheer volume of test input possibilities. They mostly rely on modeling the systems’ dynamic behavior and increasing computing resources to generate test inputs to achieve this. We found no evidence of test results aiming at managing context variation through the testing lifecycle process. Conclusions: So far, the identified testing initiatives and strategies are not ready for mainstream adoption. They are all domain-specific, and while the ideas and approaches can be reproduced in distinct settings, the technologies are to be re-engineered and tailored to the context-awareness of contemporary software systems in different problem domains. Further and joint investigations in academia and experiences in non-academic set- tings can evolve the body of knowledge regarding the testing of contemporary software systems in the field.

Context Kanban is increasingly being used in diverse software organizations. There is extensive research regarding its benefits and challenges in Software Engineering, reported in both primary and secondary studies. However, these results have not been synthesized yet. Goal to investigate and identify the actual benefits and challenges of using Kanban in Software Engineering to support practitioners in understanding and analyzing the benefits and challenges of adopting Kanban in their software projects. Method to use the Structured Synthesis Method to aggregate existing empirically-grounded evidence in the published primary studies regarding using Kanban in Software Engineering. Results from the 20 selected primary studies in which over 16 benefits were identified, four had the most robust results in the aggregation, i.e., with the most confidence associated, namely: ‘work visibility,’ ‘control of project activities and tasks,’ ‘flow of work,’ and ‘time-to-market.’ Furthermore, the ‘organizational culture’ was identified as the most dominant challenge in Kanban implementations. Conclusions Syntheses studies represent a fundamental step in organizing the body of evidence as an empirically-grounded reference for decision-making in practice. The benefits with most confidence indeed appear to be the ones intrinsically linked to the Lean thinking and the Kanban approach. As Kanban originated in the manufacturing, it is interesting to observe this kind of confirmation in the software domain. Still, there are several benefits and challenges which still lacks the appropriate level of evidence. We also noticed the absence of negative results reported in the technical literature. These aspects need the additional attention of the research community.

Genetic Improvement is an area of Search Based Software Engineering that aims to apply evolutionary computing operators to the software source code to improve it according to one or more quality metrics. This article describes challenges related to experimental studies using Genetic Improvement in JavaScript (an interpreted and non-typed language). It describes our experience on performing a study with fifteen projects submitted to genetic improvement with the use of a supercomputer. The construction of specific software infrastructure to support such an experimentation environment reveals peculiarities (parallelization problems, management of threads, etc.) that must be carefully considered to avoid future research threats to validity such as dead-ends, which make it impossible to observe relevant phenomena (code transformation) to the understanding of software improvements and evolution.

The relevance of Requirements Engineering (RE) research to practitioners is vital for a long-term dissemination of research results to everyday practice. Some authors have speculated about a mismatch between research and practice in the RE discipline. However, there is not much evidence to support or refute this perception. This paper presents the results of a study aimed at gathering evidence from practitioners about their perception of the relevance of RE research and at understanding the factors that influence that perception. We conducted a questionnaire-based survey of industry practitioners with expertise in RE. The participants rated the perceived relevance of 435 scientific papers presented at five top RE-related conferences. The 153 participants provided a total of 2,164 ratings. The practitioners rated RE research as essential or worthwhile in a majority of cases. However, the percentage of non-positive ratings is still higher than we would like. Among the factors that affect the perception of relevance are the research's links to industry, the research method used, and respondents' roles. The reasons for positive perceptions were primarily related to the relevance of the problem and the soundness of the solution, while the causes for negative perceptions were more varied. The respondents also provided suggestions for future research, including topics researchers have studied for decades, like elicitation or requirement quality criteria., Accepted at IEEE Transactions on Software Engineering

National audience; The user interfaces (UIs) promote the interaction with the software system to achieve the users' goals. In this way different types of interaction are provided, such as direct manipulation, web UI or tangible interaction. These interfaces have evolved, including artificial intelligence and adaptation mechanisms to answer the evolution of the technological areas. From this evolution emerged the intelligent user interfaces (IUIs) that aim to be more effective, efficient, and natural. Considering the importance of IUI nowadays, we have performed a systematic literature review (SLR) to investigate the design trends of IUIs in the context of contemporary software systems (CSS), such as software systems based on internet of things (IoT) or dedicated to smart cities. Preliminary results show which models and technologies are most used to develop IUIs, and which application domain is mostly represented.; Les interfaces utilisateur (IU) visent à favoriser l'interaction avec le système pour permettre aux utilisateurs d'atteindre leurs buts. Ainsi, différents types d'interaction sont possibles, tels que la manipulation directe, les interfaces web ou l'interaction tangible. Ces interfaces ont évolué au fil du temps, tout en intégrant des mécanismes issus de l'intelligence artificielle, tels des mécanismes d'adaptation, pour répondre à l'évolution des domaines technologiques. De cette évolution ont emergé les interfaces utilisateur intelligentes (IUI) qui visent à être efficaces et naturelles. Ainsi nous sommes en train de mener une revue systématique de la littérature pour étudier les tendances de conception des IUI dans le contexte des systèmes logiciels contemporains (SLC), tels que les systèmes basés sur l'internet des objets ou dédiés aux villes intelligentes. Les résultats préliminaires montrent quels sont les modèles et technologies les plus utilisés pour développer des IUIs, ainsi que le domaine d'application le plus représenté.

International audience; Contemporary software systems, such as the Internet of Things (IoT), Industry 4.0, and Smart Cities represent a technology changing that offer challenges for their construction since they are calling into question our traditional form of developing software. They are a promising paradigm for the integration of devices and communications technologies. It is leading to a shift in the classical monolithic view of development where stakeholders used to receive a software product at the end (that we have been doing for decades), to software systems incrementally materialized through physical objects interconnected by networks and with embedded software to support daily activities. Therefore, we need to revisit the classical way of developing software and start to consider the particularities required by these new sorts of applications. Since such software systems involve different concerns, this paper presents the results of an investigation towards defining a framework to support the software systems engineering of IoT applications. To support its representation, we evolved the Zachman’s Framework as an alternative to the organization of the framework architecture. The filling of such a framework is supported by a) 14 significant concerns of IoT applications, recovered from the technical literature, practitioner’s workshops and a Government Report; b) seven structured facets emerged from IoT data analysis, that together represent the engineering challenges to be faced both by researchers and practitioners towards the advancement of IoT in practice.

Security and performance are critical nonfunctional requirements for software systems. Thus, it is crucial to include verification activities during software development to identify defects related to such requirements, avoiding their occurrence after release. Software verification, including testing and reviews, encompasses a set of activities that have a purpose of analyzing the software searching for defects. Security and performance verification are activities that look at defects related to these specific quality attributes. Few empirical studies have been focused on how is the state of the practice in security and performance verification. This paper presents the results of a case study performed in the context of Brazilian organizations aiming to characterize security and performance verification practices. Additionally, it provides a set of conjectures indicating recommendations to improve security and performance verification activities.

ContextCurrent software systems have increasingly implemented context-aware adaptations to handle the diversity of conditions of their surrounding environment. Therefore, people are becoming used to a variety of context-aware software systems (CASS). This context-awareness brings challenges to the software construction and testing because the context is unpredictable and may change at any time. Therefore, software engineers need to consider the dynamic context changes while testing CASS. Different test case design techniques (TCDT) have been proposed to support the testing of CASS. However, to the best of our knowledge, there is no analysis of these proposals on the advantages, limitations and their effective support to context variation during testing. ObjectiveTo gather empirical evidence on TCDT concerned with CASS by identifying, evaluating and synthesizing knowledge available in the literature. MethodTo undertake a secondary study (quasi-Systematic Literature Review) on TCDT for CASS regarding their assessed quality characteristics, used coverage criteria, test type, and test technique. ResultsFrom 833 primary studies published between 2004 and 2014, just 17 studies regard the design of test cases for CASS. Most of them focus on functional suitability. Furthermore, some of them take into account the changes in the context by providing specific test cases for each context configuration (static perspective) during the test execution. These 17 studies revealed five challenges affecting the design of test cases and 20 challenges regarding the testing of CASS. Besides, seven TCDT are not empirically evaluated. ConclusionA few TCDT partially support the testing of CASS. However, it has not been observed evidence on any TCDT supporting the truly context-aware testing, which that can adapt the expected output based on the context variation (dynamic perspective) during the test execution. It is an open issue deserving greater attention from researchers to increase the testing coverage and ensure users confidence in CASS.

International audience; Context-awareness feature has been widely employed in the last years due to their usefulness to enhance systems functional reliability and user experience. Different software and hardware technologies can be integrated and used to capture context information, without user intervention. However, the heterogeneity of technology makes interoperability a great development challenge. Moreover, its evaluation for the context-aware system is also not trivial. Considering the relevance of the topic, we performed a characterization research (secondary study) aiming at revealing evidence on how interoperability has been addressed in context-aware software systems (CASS). Despite the observed conceptual divergence on interoperability, we present a set of evidence-based characteristics that can support decision-making and evaluation of interoperability on such systems. This work also introduces some research challenges aiming at to improve what we know about interoperability in CASS.

The development of software not only needs to consider the construction process, but also other aspects such as cost, human resources and communication among stakeholders. The lack of simplicity into this context becomes explicit when some restrictions, such as service oriented architecture, must be considered as the basic style to build sustainable applications into environments were practitioners are not aware of this software technology. In addition to this, most of the available software processes are not directly applicable nor are they reusable, so learning times becomes risk for the development of the project. Therefore, Tape Mbo’e (TME) has been proposed to support the building of such applications, into development environments like developing countries where we can have economic constraints and scarcity of proficient practitioners. The first application of TME has been to develop a service-based application whose goal is to provide the interoperability among legacy systems of different public agencies in Paraguay. Initial results of this experience indicated the feasibility and simplicity of TME when applied in this field. The evaluation process, its results and conclusions are described in this paper. Resumen El desarrollo de software no solo necesita considerar el proceso de construccin, sino tambin otros aspectos tales como los costos, los recursos humanos y la comunicacin entre los interesados. La falta de simplicidad en este contexto se vuelve explicito cuando algunas restricciones tales como la arquitectura orientada a servicios, debe ser considerada como el estilo bsico para construir aplicaciones sustentables en entornos donde los profesionales no estn al tanto de esta tecnologa. Adems, la mayora de los procesos de software no son directamente aplicables o no son reutilizables, por lo que el tiempo de aprendizaje se vuelve un riesgo para el desarrollo del proyecto. Por lo tanto, Tape Mbo’e (TME) ha sido propuesto para apoyar la construccin de tales aplicaciones, en contextos como el de pases en vas de desarrollo donde se puede tener limitaciones econmicas y escasez de profesionales competentes. La primera aplicacin de TME has sido para desarrollar una aplicacin basada en servicios cuyo objetivo es la interoperabilidad entre sistemas legados de diferentes instituciones pblicas en Paraguay. Los resultados iniciales de esta experiencia indicaron la viabilidad y simplicidad de TME cuando se aplica en este campo. El proceso de evaluacin, sus resultados y conclusiones son descriptos en este artculo.

Incorporar agilidade aos processos de software é um tema que vai além do debate sobre a aplicação de métodos ágeis pré-definidos. Visando reforçar as evidências sobre a viabilidade de introduzir características de agilidade e práticas ágeis previamente mapeadas para obtenção de agilidade em processos de software, este artigo apresenta a reexecução em larga escala de um survey sobre a pertinência e a relevância destas características e práticas, coletando a opinião de 292 participantes. Os resultados sugerem um consenso sobre a pertinência da maioria das características e das práticas investigadas, embora a relevância delas não seja percebida na mesma intensidade. Em adição, novas características e práticas foram propostas.

A especificação de aplicações de software contemporâneas costuma envolver uma quantidade de informação que necessita ser devidamente capturada e representada. Utilizar casos de uso e diagramas de atividades representa uma alternativa viável desde que se possa garantir sua qualidade. Entretanto, esta garantia está limitada pela carência atual de técnicas que apoiem a identificação de defeitos, principalmente, em diagramas de atividades. Desta forma, neste artigo apresentamos ActCheck, uma técnica configurável de inspeção baseada em checklist desenvolvida para apoiar a identificação de defeitos em diagramas de atividades usados na especificação de requisitos. Um estudo de viabilidade indicou que inspeções com ActCheck permitem identificar defeitos diferentes dos detectados por inspeções ad-hoc, apesar da necessidade de reorganização dos itens dos checklists visando otimizar o tempo de inspeção.

Model-based testing (MBT) approaches help automatically generate test cases using models extracted from software artifacts, and hold the promise to greatly affect how we build software. A review of the literature shows that certain specialized domains are applying MBT, but it does not yet seem to be a mainstream approach. The authors therefore conducted a systematic review of the literature to investigate how much evidence is available on MBT's costs and benefits, especially regarding how these techniques compare to other common testing approaches. They use these results to derive suggestions regarding what types of studies might further increase the deployment of these techniques.

SUMMARY We suggest that empirical studies of maintenance are difficult to understand unless the context of the study is fully defined. We developed a preliminary ontology to identify a number of factors that influence maintenance. The purpose of the ontology is to identify factors that would affect the results of empirical studies. We present the ontology in the form of a UML model. Using the maintenance factors included in the ontology, we define two common maintenance scenarios and consider the industrial issues associated with them. Copyright © 1999 John Wiley & Sons, Ltd.

An adequate representation and a feasible aggregation procedure of evidence represents a challenging problem in many disciplines. The right representation can help scientists discuss and present the results of their findings and, if it is simple enough, it can be useful for practitioners to base their decisions on improvement implementations. The aggregation strengthens confidence in comparison to single evidence and is an important contribution to the body of knowledge. In this paper, we present a preliminary proposal to use empirically-based theories and belief functions as a means to represent and aggregate evidence. By having evidence explained by the same theory, we used belief functions to combine them in a way that the theory propositions (cause-effect values) result from combined evidence. We suggest this can be an useful way to obtain a good estimate of multiple evidence combination. In addition, we indicate its possible usefulness for practitioners to formalize and reuse their experiences. A real-case application of the approach is presented by formulating a theory for Usage-Based Reading inspection technique and aggregating the evidence acquired in three related empirical studies. This application indicated that the approach can give compatible results with the aggregated evidence.

Establishing representative samples for Software Engineering surveys is still considered a challenge. Specialized literature often presents limitations on interpreting surveys’ results, mainly due to the use of sampling frames established by convenience and non-probabilistic criteria for sampling from them. In this sense, we argue that a strategy to support the systematic establishment of sampling frames from an adequate source of sampling can contribute to improve this scenario. A conceptual framework for supporting large scale sampling in Software Engineering surveys has been organized after performing a set of experiences on designing such strategies and gathering evidence regarding their benefits. The use of this conceptual framework based on a sampling strategy developed for supporting the replication of a survey on characteristics of agility and agile practices in software processes is depicted in this paper. A professional social network (Linkedln) was established as the source of sampling and its groups of interest as the units for searching members to be recruited. It allowed to deal with a sampling frame composed by more than 110,000 members (prospective subjects) distributed over 19 groups of interest. Then, through the similarity levels observed among these groups, eight strata were organized and 7745 members were invited, from which 291 have confirmed participation and answered the questionnaire. The heterogeneity and number of participants in this replication contributed to improve the strength of original survey’s results. Therefore, we believe the sharing of this experience, the instruments and plan can be helpful for those researchers and practitioners interested on executing large scale surveys in Software Engineering.

DevOps and continuous delivery have greatly impacted the organizational structures of development and infrastructure groups in software-producing organizations. Our research aims at revealing the different options adopted by the software industry to organize such groups, understanding why different organizations adopt distinct structures, and discovering how organizations handle the drawbacks of each structure. By interviewing 68 carefully-selected, skilled IT professionals and analyzing these conversations through a Grounded Theory process, we identified conditions, causes, reasons to avoid, consequences, and contingencies related to each discovered structure (segregated departments, collaborating departments, API-mediated departments, and single department). We, then, offer a theory to explain organizational structures for development and infrastructure professionals. This theory can support practitioners and researchers in comprehending and discussing the DevOps phenomenon and its related issues; it also provides valuable input to practitioners decision-making. DevOps e entrega contínua têm impactado as estruturas organizacionais dos grupos de desenvolvimento e infraestrutura em organizações produtoras de software. Nossa pesquisa visa revelar as diferentes opções adotadas pela indústria de software para organizar tais grupos, entender por que diferentes organizações adotam estruturas distintas e descobrir como as organizações lidam com as desvantagens de cada estrutura. Ao entrevistar 68 qualificados profissionais de TI, cuidadosamente selecionados, e analisar essas conversas por meio de um processo de Teoria Fundamentada (Grounded Theory), identificamos condições, causas, razões para evitar, consequências e contingências relacionadas a cada estrutura descoberta (departamentos segregados, departamentos colaborativos, departamentos mediados por API e departamento único). Esta tese, então, propõe uma teoria para explicar estruturas organizacionais para profissionais de desenvolvimento e infraestrutura. Essa teoria pode apoiar profissionais e pesquisadores na compreensão e discussão do fenômeno DevOps e seus problemas relacionados, e também fornece informações valiosas para a tomada de decisões dos profissionais.

The Internet of Things (IoT) paradigm brings many expectations, challenges, and a multidisciplinarity of technical knowledge. It involves several knowledge areas or facets that should be combined precisely and coherently to develop more autonomous and smarter software systems. In this work, 117 recommendations distributed in 29 categories were organized in an IoT Roadmap instrument to support IoT development. The IoT Roadmap is an evidence-based artifact to comprise different expertise to deal with IoT in a multi-faceted way. The IoT Roadmap guides what to consider while specifying, designing, and implementing IoT systems. The Systems Engineering Body of Knowledge life cycle (concept definition, system definition, and realization) supports the roadmap elements' organization and temporal dimension to guide IoT systems engineering effectively. The analyses and experimental studies resulted in a large set of information on multidisciplinary IoT, organized into a body of knowledge that can benefit researchers and professionals involved in IoT.; Dans cette thèse, nous avons présenté la conceptualisation, le développement et l'évaluation d'un artefact fondé sur des preuves nommé IoT Roadmap pour prendre en charge la spécification, la conception et la mise en oeuvre de systèmes logiciels IoT. La feuille de route IoT a été organisée sur la base de preuves acquises grâce à des études expérimentales et a évolué avec les études primaires menées dans son évaluation. La feuille de route IoT englobe l'IoT multidisciplinaire impliquant les facettes des objets, de l'interactivité, de la connectivité, du comportement, de l'intelligence, de l'environnement et des données avec des recommandations conçues individuellement pour chaque facette. De plus, les recommandations ont une composition temporelle, couvrant les phases génériques du cycle de vie d'un projet d'ingénierie système (définition du concept, définition du système et réalisation du système). La motivation d'un tel artefact a émergé de l'intérêt croissant pour l'IoT et de la demande de technologies logicielles qui prennent en compte les particularités et les caractéristiques de ce paradigme. De plus, nous avons observé que les défis révélés par les études primaires et rapportés dans la littérature technique renforcent le besoin de technologies logicielles pour prendre en charge l'ingénierie des systèmes logiciels IoT. Par conséquent, la feuille de route IoT peut aider les chercheurs et les praticiens à faciliter la compréhension, la planification et le développement de systèmes logiciels IoT. Les recommandations suggérées par la feuille de route IoT peuvent contribuer à avoir une orientation plus claire pour le projet, en fournissant des directives du domaine problématique à la solution IoT matérialisée. Les chercheurs et les praticiens peuvent définir les facettes et les éléments les plus pertinents pour un projet spécifique et une phase spécifique, en sélectionnant ce qui peut s'appliquer à leurs objectifs. La feuille de route IoT a été organisée pour donner de la visibilité à ce qui a été fait avec un espace pour ajouter des commentaires et des preuves pour chaque élément. Cela peut être une alternative pour percevoir et gérer les besoins, les demandes et les risques associés à l'ingénierie d'une solution pour un système logiciel IoT. La feuille de route IoT définie dans cette thèse fournit sept facettes, dirigées par le domaine du problème, influençant les activités de conceptualisation et de réalisation. Les connaissances derrière la feuille de route IoT montrent que de tels projets doivent (1) définir le domaine du problème en soulignant pourquoi l'IoT est utilisé pour atteindre un objectif, (2) considérer quels composants seront utilisés pour atteindre un tel objectif, (3) définir l'identification, la détection, l'action et d'autres ensembles de comportements à exécuter par ces composants, (4) identifier tous les acteurs impliqués dans la solution et leurs méthodes respectives, (5) établir un support adéquat pour que tout soit connecté, (6) définir l'intelligence, l'intelligence et l'automatisation nécessaires à un tel objectif, (7) mettre en oeuvre les stratégies pour faire face à la capture, l'analyse et le traitement des données ; et (8) considérer l'influence sur et à partir de l'environnement dans lequel la solution est installée. Les facettes sont organisées en catégories qui contribuent à la compréhension et à la compréhension du paradigme IoT. De plus, chaque catégorie est composée d'éléments fournissant des recommandations et des actions que les organisations de logiciels peuvent utiliser pour prendre en charge l'ingénierie des systèmes logiciels IoT. Les recommandations actuelles sont simples et peuvent être utilisées dans l'ordre ou uniquement selon les facettes souhaitées, en fonction de l'objectif du projet et des compétences de l'équipe. Cette organisation permet un flux de raisonnement des objectifs du projet aux résultats et aux résultats à travers les discussions et la prise de décision. ... [etc]; O paradigma da Internet das Coisas (IoT) traz muitas expectativas, desafios e uma multidisciplinaridade do conhecimento técnico. Envolve várias áreas ou facetas do conhecimento que devem ser combinadas de forma precisa e coerente paradesenvolver sistemas de software mais autônomos e inteligentes. Neste trabalho, um conjunto de 117 recomendações distribuídas em 29 categorias foram organizadas em instrumento chamado de IoT Roadmap, para apoiar o desenvolvimento da IoT. O IoT Roadmap é um artefato baseado em evidências que compreende diferentes experiências para lidar com a IoT de uma maneira multifacetada. O Roadmap é um guia do que considerar ao especificar, projetar e implementar sistemas IoT. O ciclo de vida do Corpo de Conhecimento em Engenharia de Sistemas (definição conceitual, definição e realização do sistema) oferece suporte à organização dos elementos do roteiro e à dimensão temporal para orientar a engenharia de sistemas IoT de maneira eficaz. As análises e estudos experimentais resultaram em um grande conjunto de informações sobre IoT multidisciplinar, organizado em um corpo de conhecimento que pode beneficiar pesquisadores e profissionais envolvidos na IoT.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO Contexto: Exceções não tratadas afetam a confiabilidade, usabilidade e segurança em aplicações web. Diversos estudos mediram a confiabilidade de aplicações web em uso em relação a exceções não tratadas, mostrando a recorrência deste problema durante a fase de manutenção. Detectar exceções não tratadas latentes de forma automatizada é uma tarefa difícil e específica de cada aplicação. Assim, abordagens gerais para tratar defeitos em aplicações web não tratam exceções não tratadas latentes apropriadamente. Objetivos: Projetar e avaliar um método que possa suportar encontrar, corrigir e prevenir exceções não tratadas em aplicações web. Método: Nós aplicamos o ciclo de engenharia do design science para projetar o método chamada Manutenção Orientada a Padrões (PDM). PDM consiste em identificar padrões de defeitos se baseando nos logs do servidor de aplicação, produzindo regras de análise estática que podem ser utilizadas para a prevenção de defeitos. Nós aplicamos PDM em duas aplicações web na indústria envolvendo empresas e tecnologias diferentes, medindo a melhoria confiabilidade das aplicações e a precisão das regras de análise estática produzidas. Nós também avaliamos o reuso das regras de análise estática produzidas durante a aplicação do PDM em software da mesma empresa e de outras empresas. Finalmente, nós estudamos a eficácia e aceitação de mantenedores novatos aplicando o método PDM. Resultados: Nos dois casos industriais, nossa abordagem permitiu a identificação de padrões de defeitos e exceções não tratadas latentes para correção no código fonte, permitindo eliminar completamente as falhas relacionadas a exceções não tratadas latentes e melhorando assim a confiabilidade da aplicação. Algumas regras de análise estática produzidas pela aplicação do método PDM foram reutilizadas em software na mesma empresa e em outra empresa. Nós identificamos os conhecimentos e experiências que influenciam em aplicar os passos do método PDM de forma eficaz. A maior parte dos mantenedores novatos acharam o método PDM útil, mas não fácil de aplicar, dificultando a aceitação do método entre novatos. Conclusões: Os resultados fortalecem nossa confiança que o PDM pode ajudar os mantenedores a melhorar a confiabilidade em relação a exceções não tradadas em aplicações web existentes. Nós disponibilizamos orientações sobre como utilizar o método, reutilizar as regras de análise estática produzidas, e quais conhecimentos e experiências são necessárias para aplicar o PDM com eficácia. Background: Unhandled exceptions affect the reliability, usability, and security of web applications. Several studies have measured the reliability of web applications in use against unhandled exceptions, showing a recurrence of the problem during the maintenance phase. Detecting automatically unhandled latent exceptions is difficult and application-specific. Hence, general approaches to deal with defects in web applications do not treat unhandled exceptions appropriately. Aims: To design and evaluate a method that can support finding, correcting, and preventing unhandled exceptions in web applications. Method: We applied the design science engineering cycle to design a method called Pattern-Driven Maintenance (PDM). PDM relies on identifying defect patterns based on application server logs and producing static analysis rules that can be used for prevention. We applied PDM to two industrial web applications involving different companies and technologies, measuring the reliability improvement and the precision of the produced static analysis rules. We also evaluated reuse of static analysis rules produced during PDM application on within- and cross-company software. Finally, we studied the effectiveness and acceptance of novice maintainers on applying the PDM method Results: In both industry cases, our approach allowed identifying defect patterns and finding unhandled latent exceptions to be fixed in the source code, enabling to eliminate the pattern-related failures and improving the application reliability completely. Some of the static analysis rules produced by PDM application were reused on within- and cross-company software. We identified knowledge and experiences that influence on effectively applying steps of the PDM method. Most of the novice maintainers find PDM useful, but not easy to apply, thus hindering PDM acceptance among novices. Conclusions: The results strengthen our confidence that PDM can help maintainers to improve the reliability for unhandled exceptions in existing web applications. We provide guidance on how to apply PDM, reuse the produced static analysis rules, and the knowledge and experiences needed to apply the PDM method effectively.

Em um sistema de software, as anomalias de código indicam problemas estruturais que podem ser resolvidos através da refatoração. No entanto, desenvolvedores podem negligenciar ou acabar criando novas anomalias ao refatorar. Pouco foi relatado sobre os efeitos benéficos e prejudiciais da refatoração de anomalias de código. Evidências sugerem que os desenvolvedores frequentemente precisam aplicar uma sequência de refatorações (refatoração em lote) para remover completamente as estruturas anômalas. Assim, nesta tese, realizamos uma série de estudos para entender o impacto de refatorações simples e em lote em anomalias de código. Em nossos primeiros estudos, analisamos com que frequência os tipos de refatoração comumente usados afetam a densidade de anomalias ao longo das histórias de dezenas de projetos. Mesmo que 79,4 por cento das refatorações tenham tocado em elementos anômalos, 57 por cento não reduziram suas ocorrências. Surpreendentemente, apenas 9,7 por cento das refatorações removeram anomalias de código, enquanto 33 por cento induziram a introdução de novas. Por um lado, observamos padrões nocivos de introdução de anomalias. Por outro lado, observamos que muitas anomalias podem ser removidas apenas por refatorações em lote. Assim, nossos últimos estudos investigam o impacto de refatorações em lote nas anomalias. Mesmo quando aplicadas em lotes, as refatorações tendem a não afetar ou mesmo aumentar a densidade de anomalias. Também identificamos padrões entre tipos de lotes e tipos de anomalias, levando-nos à criação de heurísticas que podem orientar os desenvolvedores durante tarefas de remoção de anomalias de código. O último estudo avaliou essas heurísticas e concluímos que os resultados são promissores. Code smells in a program represent indications of structural quality problems, which can be addressed by software refactoring. However, developers may neglect or end up creating new code smells through single refactoring. Little has been reported about recurring beneficial and harmful effects of refactoring on the program structural quality. As a consequence, developers still miss guidance along non-trivial smell-removing tasks. In fact, evidence suggests developers often need to apply a sequence of refactorings, so-called batch refactoring, to entirely remove a smelly code structure. Thus, in this thesis, we have conducted a series of studies to understand the impact of single and batch refactorings on code smells. In our first studies, we analyze how often commonly-used types of single refactoring affect the density of code smells along the version histories of dozens of projects. Even though 79.4 percent of the refactorings touched smelly elements, 57 percent had no impact on the smell removal. Surprisingly, only 9.7 percent of refactorings removed smells, while 33 percent induced the introduction of new ones. On one hand, we observed that harmful refactoring-smell patterns could be used to guide developers to avoid smell-inducing refactoring. On the other hand, we observed that many smells can be removed only through batch refactoring. Thus, our last studies investigate the impact of batch refactorings on smells. Even when applied in batches, refactorings tend to maintain or even increase the density of code smells. We also identified common batch-smell patterns, which enable us to create heuristics that can guide developers through smell-removing tasks. The last study evaluated those heuristics, and we conclude the outcomes are promising.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO Mesmo sistemas de software desenvolvidos com um controle de qualidade rigoroso podem apresentar falhas durante seu ciclo de vida. Quando uma falha é observada no ambiente de produção, mantenedores são responsáveis por produzir o diagnóstico e remover o seu defeito correspondente. No entanto, em um serviço crítico este tempo pode ser muito longo, logo, se for possível, a assinatura da falha deve ser utilizada para gerar um mecanismo de recuperação automático capaz de detectar e tratar futuras ocorrências similares, até que o defeito possa ser removido. Nesta tese, a atividade de recuperação consiste em restaurar o sistema para um estado correto, que permita continuar a execução com segurança, ainda que com limitações em suas funcionalidades. Para serem eficazes, as tarefas de diagnóstico e recuperação requerem informações detalhadas sobre a execução que falhou. Falhas que ocorrem durante a fase de testes em um ambiente controlado podem ser depuradas através da inserção de nova instrumentação e re-execução da rotina que contem o defeito, tornando mais fácil o estudo de comportamentos inesperados. No entanto, falhas que ocorrem no ambiente de produção apresentam informações limitadas à situação específica em que ocorrem, além de serem imprevisíveis. Para mitigar essa adversidade, informações devem ser coletadas sistematicamente com o intuito de detectar, diagnosticar para recuperar e, eventualmente, diagnosticar para remover a circunstância geradora da falha. Além disso, há um balanceamento entre a informação inserida como instrumentação e a performance do sistema: técnicas de logging geralmente apresentam baixo impacto no desempenho, porém não provêm informação suficiente sobre a execução; por outro lado, as técnicas de tracing podem registrar informações precisas e detalhadas, todavia são impraticáveis para um ambiente de produção. Esta tese propõe uma abordagem hibrida para gravação e extração de informações durante a execução do sistema. A solução proposta se baseia no registro de eventos, onde estes são enriquecidos com propriedades contextuais sobre o estado atual da execução no momento em que o evento é gravado. Através deste registro de eventos com informações de contexto, uma técnica de diagnóstico e uma ferramenta foram desenvolvidas para permitir que eventos pudessem ser filtrados com base na perspectiva de interesse do mantenedor. Além disso, também foi desenvolvida uma abordagem que utiliza estes eventos enriquecidos para detectar falhas automaticamente visando recuperação. As soluções propostas foram avaliadas através de medições e estudos conduzidos em sistemas implantados, baseando-se nas falhas que de fato ocorreram enquanto se utilizava o software em um contexto de produção. Even software systems developed with strict quality control may expect failures during their lifetime. When a failure is observed in a production environment the maintainer is responsible for diagnosing the cause and eventually removing it. However, considering a critical service this might demand too long a time to complete, hence, if possible, the failure signature should be identified in order to generate a recovery mechanism to automatically detect and handle future occurrences until a proper correction can be made. In this thesis, recovery consists of restoring a correct context allowing dependable execution, even if the causing fault is still unknown. To be effective, the tasks of diagnosing and recovery implementation require detailed information about the failed execution. Failures that occur during the test phase run in a controlled environment, allow adding specific code instrumentation and usually can be replicated, making it easier to study the unexpected behavior. However, failures that occur in the production environment are limited to the information present in the first occurrence of the failure. But run time failures are obviously unexpected, hence run time data must be gathered systematically to allow detecting, diagnosing with the purpose of recovering, and eventually diagnosing with the purpose of removing the causing fault. Thus there is a balance between the detail of information inserted as instrumentation and the system performance: standard logging techniques usually present low impact on performance, but carry insufficient information about the execution; while tracing techniques can record precise and detailed information, however are impracticable for a production environment. This thesis proposes a novel hybrid approach for recording and extracting system s runtime information. The solution is based on event logs, where events are enriched with contextual properties about the current state of the execution at the moment the event is recorded. Using these enriched log events a diagnosis technique and a tool have been developed to allow event filtering based on the maintainer s perspective of interest. Furthermore, an approach using these enriched events has been developed that allows detecting and diagnosing failures aiming at recovery. The proposed solutions were evaluated through measurements and studies conducted using deployed systems, based on failures that actually occurred while using the software in a production context.