Software Research and Development Organizations (or SRDs) have unique goals that differ from the goals of Production Software Organizations. SRDs focus on exploring the unknown, while Production Software Organizations focus on implementing solutions to known problems. These unique goals call for reevaluating the role of Software Engineering Process for SRDs. This paper presents six common Software Engineering Processes then analyzes their strengths and weaknesses for SRDs. The processes presented include: Waterfall, Rational Unified Process (RUP), Evolutionary Delivery Cycle (EDLC), Team Software Process (TSP), Agile Development and Extreme Programming (XP). The results indicate that an ideal software process for SRDs is iterative, emphasizes visual models, uses a simple organization structure, produces working software (with limited functionality) early in the lifecycle, exploits individual capabilities, minimizes artifacts, adapts to new discoveries and requirements, and utilizes collective code ownership among developers. The results also indicate that an ideal software process for SRDs does NOT define rigid personnel roles or rigid artifacts, is NOT metric-driven and does NOT implement pair programming. This paper justifies why SRDs require a unique software process, outlines the ideal SRD software process, and shows how to tailor existing software processes to meet the unique needs of SRDs. / text

Thesis (M.S.)--West Virginia University, 2008. / Title from document title page. Document formatted into pages; contains ix, 160 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 154-160).

A área de melhoria de processos de software (MPS) tem sido investigada sistematicamente, dadas as evidências de que a qualidade do processo pode influenciar significativamente na qualidade do produto final. Modelos e guias com boas práticas para a MPS têm sido sintetizados. Ainda assim, a literatura carece de trabalhos que estabeleçam estratégias de como implementar as boas práticas sugeridas por tais modelos e guias na indústria. Em paralelo, o desenvolvimento distribuído de software (DDS) tornou-se uma realidade, aumentando a complexidade e a importância do processo de desenvolvimento de software e demandando estratégias que permitam uma MPS também distribuída. O objetivo deste trabalho é investigar estratégias e mecanismos que possam promover uma MPS distribuída e colaborativa. A ColabSPI, uma estratégia colaborativa e distribuída para MPS, é proposta para apoiar a evolução estruturada do processo; o tratamento de propostas de melhorias de processo; e a comunicação e participação dos desenvolvedores na MPS. Durante a investigação, fatores que podem influenciar a MPS foram identificados e a relação entre eles foi mapeada, tanto a partir da literatura quanto a partir de um estudo em campo. Dois contextos foram explorados durante a investigação: (i) a MPS em uma organização de grande porte, com unidades distribuídas; e (ii) a MPS no desenvolvimento de software livre (SL). Contribuições para a MPS no desenvolvimento de SL foram geradas no contexto do projeto internacional Qualipso, com a co-autoria do Qualipso Open Source Maturity Model (OMM) e a adequação de ColabSPI para evoluir o OMM. Em ambiente industrial, estudos experimentais foram conduzidos para validar a aplicação de ColabSPI e revelaram que algumas práticas do desenvolvimento de software, inclusive do DDS e do desenvolvimento de SL, podem ser aplicadas com sucesso à MPS, trazendo ganhos de eficácia e eficiência para a melhoria de um processo de desenvolvimento de software. A estratégia proposta está sendo base também para a definição do ambiente de MPS do processo Demoiselle, um processo livre para desenvolvimento de software / Software Process Improvement (SPI) has become an active research field, given the evidences that the quality of software processes can significantly influence the final product quality. Therefore, SPI models and guides have been disclosed. Still, there is a lack of studies establishing strategies on how to implement SPI models\' best practices. Meanwhile, distributed software development (DSD) is becoming widespread, increasing the complexity and importance of software processes and demanding distributed SPI strategies. This study aims to investigate strategies and mechanisms to promote a distributed and collaborative SPI. ColabSPI, a collaborative and distributed approach to SPI is proposed to supporting process evolution, handling process improvement proposals, and promoting the developers participation in SPI and communication. Accordingly, during the investigation possible influences to SPI programs were explored and the relationship between them has been mapped, both from the literature and from a field study. SPI was investigated in two contexts: (i) a large organization, counting on distributed development units; and (ii) the FLOSS development environment. Contributions to SPI in the FLOSS development environment took place in the international project Qualipso resulting on a ColabSPI customization to evolve the Qualipso Open Source Maturity Model (OMM) and in the co-authoring of the model. In industrial environment, experimental studies were conducted to validate ColabSPI, revealing that some of the software development practices, including DSD and FLOSS practices, can be successfully applied to SPI and yield a more efficient and effective improvement of the software development process. ColabSPI features are also being considered to the definition of a SPI environment to the Demoiselle process, an open process for software development

Software process improvement is an important activity which starts when an organization plans to enhance/purify the capabilities of its ongoing processes. When improvement or change is planned or started, there are a number of factors which originate and affect the effectiveness of software process improvement. If they are not identified and dealt with proper expertise, they can leave adverse effects in the form of low effectiveness, high cost, low quality etc. Such factors change the intensity or the expected results of software process improvement. It is very important to understand and highlight these factors in order to make the process improvement effective and efficient in an organization during change. This master thesis is written to answer the research questions which elaborate the understanding of software process improvement and factors affecting the software process improvement during change in an organization. Moreover, a systematic review is also performed in identifying and prioritizing the factors which affect the software process improvement activity in change. The outcomes of this research can be helpful in understanding the factors affecting SPI during change.

Current software process modeling tools lack the capabilities of generating word processing documents that can represent model semantics in a computer process-able and human understandable way. This results into inefficient use of word processors for editing and reviewing a model’s textual data. In an attempt to resolve this problem, this thesis presents an approach for representing software process models in word processing documents. The development of the approach is based on a set of issues that can hinder the generation of human understandable and computer process-able documents from software process models. The approach is validated through its implementation for a software process modeling tool. The implementation allows for the generation of word processing documents from software process models and their re-import into the process modeling tool. / Email: msaqibsaeed@hotmail.com

In the competitive global economy, increasing customer value by improving software product quality is the aim for organizations. Academia and industry are collaboratively striving to find new ways to achieve this goal. The last two decades have seen a mind shift where emphasis is now on improving processes to improve the quality of products. This approach has its roots in the realization that a better development process would lead to a better quality product. Based on this process improvement frameworks were created. These frameworks enable organizations to systematically improve their development processes and strive towards the ultimate goal of improving product quality. However, these initiatives were not always successful. The deriving factors that make the adoption and implementation of such frameworks a success were still unknown. In this regard, several primary studies have been conducted to identify the factors that make software process improvement initiative successful. Researchers have been investigating factors that affect success of Software Process Improvement (SPI). There was still a lack of a secondary study where systematic approach has been used to find factors that affect the success of SPI. In our thesis project we present results of the systematic review to accumulate success factors for SPI. These findings were validated by conducting an industrial survey. With our secondary study we have identified 31 important success factor categories of SPI. We have contributed to the body of knowledge by devising formal definitions for important factors. We have also unveiled the relationship of SPI success factors not only to different SPI framework but also to software process areas.

Context. Software Process Improvement (SPI) initiatives like Capability Maturity Model (CMM), Capability Maturity Model Integration (CMMI), Bootstrap etc., have been developed on the primary agenda of continuous software process improvement. Similarly, about two decades ago, the United Kingdom Accreditation Services (UKAS) have laid down a set of guidelines based on the ISO quality standards for providing certification to organizations named TickIT. TickIT is now obsolete with its successor scheme TickITplus taking up its place with a lot of significant additions. All the companies which were certified based on TickIT guidelines (more than 1000 companies) were asked to move to TickITplus in order to keep their TickIT certification. However, until now it has been three years since the inception of TickITplus and only 70 companies have adopted TickITplus. This is way below relative to the number of TickIT certified organizations. The present thesis is done in order to find the factors why most of the companies have not adopted TickITplus and also why the 70 organizations have moved to TickITplus.   Objectives In this study, an attempt has been made to accomplish the following objectives: Identifying the changes that have been brought about in the new scheme. The factors that a software organization looks into while adopting or migrating to a new software quality certification scheme are identified. Validate these factors with the help of survey and interviews. Analyze the results of survey and interviews to provide the reasons why most of the organizations haven’t adopted TickITplus certification scheme. Methods. This research is done by using a mixed method approach by incorporating both quantitative and qualitative research methods. An online survey is conducted with the help of an online questionnaire as part of the quantitative leg. Two survey questionnaires have been framed to gather responses. With respect to the qualitative research method interviews are conducted to get a wider understanding about the factors that led an organization to migrate or not to migrate to TickITplus. The gathered data is analyzed using statistical methods like bivariate and univariate analysis for the quantitative method and thematic coding has been applied for the qualitative method. Triangulation method is used to validate the data obtained by correlating the results from the survey and interviews with those extracted from the literature review. Results. Results pertaining to the reasons why companies have moved to and also why other companies haven’t taken up TickITplus have been gathered from the survey and interviews. It was identified that high costs and low customer demand were the main reasons for the organizations not to choose TickITplus while among the organizations which have moved to TickITplus have also chosen the scheme based on customer requirement. However, few other reasons apart from these have also been identified which are presented in this document Conclusions. Conclusions have been drawn citing the importance of costs incurred for implementing TickITplus as a reason for not selecting TickITplus as it was considered very expensive. Among other reasons customer requirement was also low which was identified as a factor for the relatively low number of TickITplus certified organizations. On the other hand, among the TickITplus certified firms, customer demand forms the prominent reason for moving to TickITplus and lack of appropriate people to take up the work was considered as an important hindrance while implementing TickITplus. Several other reasons and challenges have also been identified which are clearly detailed in the document.

Um crescente número de empresas produtoras de software tem adotado um modelo organizacional de Fábrica de Software (FS), que facilita a terceirização por intermédio da segmentação das atividades e adoção de um sistema de produção mais flexível, dinâmico e controlado. Uma FS pode atender a múltiplas demandas de natureza e escopos distintos com o intuito de prover as necessidades específicas de cada cliente. Em face da diversidade e complexidade deste contexto, uma das questões principais a resolver é como mapear todos os processos envolvidos, identificando claramente o que, quem e, sobretudo como cada trabalho deve ser executado e controlado, visando o alinhamento dos processos à estrutura organizacional e conceitual de FS com foco nos processos de integração, alinhamento e reuso. Este estudo tem como objetivo mapear, definir, reestruturar e estabelecer processos de Fábrica de Software, conduzidos pelo método de pesquisaação em uma organização produtora de software do setor público, considerando seu contexto operacional, técnico e cultura organizacional. O desenvolvimento, execução, acompanhamento e resultados são descritos pelos ciclos da pesquisa-ação envolvendo o estudo de estruturas organizacionais, metodologias de desenvolvimento de sistemas, e mudanças organizacionais. Para isso foram definidas uma Estrutura Organizacional de Referência de FS, uma Arquitetura de Definição de Processos para FS e uma Metodologia de Desenvolvimento de Sistemas e Integração com foco organizacional, concebidos e desenvolvidos com base na literatura, experiência profissional e pesquisa empírica, que serviram como guia de condução da pesquisa-ação, culminando na associação do estudo empírico com o estudo teórico. As contribuições empíricas e teóricas geradas referem-se à melhoria dos processos organizacionais e operacionais de uma empresa de TI do setor público com base nos conceitos de FS. / A growing number of software producing companies have adopted an organizational model of Software Factory (SF), which facilitates outsourcing by segmenting activities and by adopting a more flexible, dynamic and controlled production system. A SF can serve the multiple demands of different nature and scopes with the purpose of providing each customers specific needs. Because of the diversity and complexity of this context, one of the main issues to be solved is how to map all the processes involved, clearly identifying what, who and mainly how each work must be executed and controlled, aiming at aligning the processes to the organizational and conceptual structure of SF focusing on the integration, aligning and reuse processes. The purpose of this study is mapping, defining, restructuring and establishing the Software Factory process, conducted by the action research method in a software producing organization of the public sector, considering its operational, technical and organizational culture context. The development, execution, follow-up and results are described by the action research cycles involving the study of organizational structures, system development methodologies and organizational changes. For this, a Reference Organizational Structure of SF, and Processes Definition Architecture for SF and a Integration System Development Methodology with organizational focus were defined, conceived and developed based on the literature, professional experience and empirical research, which served as a guiding line of the action research, peaking at the association of the empirical study with the theoretical study. The generated empirical and theoretical contributions refer to the improvement of the organizational and operational processes based on the SF concepts.

Uma das grandes dificuldades do desenvolvimento de software é adequar, de forma sistemática, os esforços na resolução de problemas e nas melhorias dos processos internos aos objetivos do negócio. O Seis Sigma é uma metodologia estruturada que incrementa a qualidade dos processos e o alinhamento com as estratégias empresariais, proporcionando altos ganhos financeiros para quem o utiliza. Surgido inicialmente na manufatura, o Seis Sigma vem sendo aplicado também noutros setores. No desenvolvimento de software, o seu uso é ainda restrito, mas há indícios de que possa ser utilizado com o mesmo sucesso. O objetivo desta dissertação de mestrado é propor um roteiro específico para a implementação do Seis Sigma nas organizações desenvolvedoras de software. Além do levantamento teórico, foi realizado um estudo de casos múltiplos para analisar as razões da sua adoção, as modalidades de seu uso, os seus benefícios e as suas limitações para os processos de software. Os resultados observados mostram que o Seis Sigma pode ser aplicado total ou parcialmente na melhoria dos processos de software. Além disso, é necessário alguma customização nos processos do Seis Sigma convencional, uma vez que os projetos de melhoria podem estar orientados para novos produtos, novos clientes ou, ainda, que os trabalhos de desenvolvimento fiquem a cargo de novas equipes, que muitas vezes são terceirizadas. O trabalho conclui que o Seis Sigma pode contribuir efetivamente para a melhoria dos processos de software, independentemente da organização objetivar níveis superiores de maturidade. O estudo sugere também um roteiro específico, chamado SW-DMAIC, para a aplicação do Seis Sigma em empresas desenvolvedoras de software. Este roteiro é composto por três etapas: implementação da filosofia do Seis Sigma, reciclagem desta filosofia, quando necessário e execução de projetos específicos. / Aligning sistematically efforts to solve problems and improve the software process with business goals is one of the challenges of software development. Six Sigma is a structured methodology that increases the process quality and aligning with enterprise strategies, providing high financial returns for those who uses it. Firstly applied in manufacturing companies, Six Sigma has been applied in other sectors too. It is not widely used in the software companies yet, but there are already some successful cases. The main objective of this master degree dissertation is to propose a specific roadmap for the Six Sigma application at software companies. Besides the theoretical research, it is accomplished a multiple case study to analyze the adoption reasons, the modalities of the use, the benefits and the limitations of Six Sigma for the software processes. The observed results show that Six Sigma could be applied totally or partially for software process improvement. Therefore it is necessary some tayloring in the Six Sigma processes, because improvement projects could refer for new products or new customers or some of the development work is in charge of new teams, specially when they are a outsoursing. The work has concluded that Six Sigma could contribute indeed to software process improvement, independently the enterprise is going to reach higher level maturity. This study suggests also a specific three-stage-roadmap, called SW-DMAIC, for the Six Sigma application at software companies containing processes for implementation of Six Sigma?s philosophy, recycling this philosophy when it?s necessary and, execution of specific Six Sigma projects.

A maioria das micro e pequenas empresas (MPEs) de desenvolvimento de software brasileiras são voltadas para o desenvolvimento de software on-demand. Normalmente, essas MPEs têm dificuldades em formalizar um processo de software padrão. O presente trabalho de pesquisa tem por objetivo apresentar uma proposta de integração de modelos/normas de capacidade de processo com modelos de ciclo de vida de software, em um contexto de gestão de processos de negócio. Como resultado, obteve-se o modelo de gestão do processo de venda e desenvolvimento de software on-demand para MPEs (ProcSoftVD - Gestão), composto pelo método de melhoria de processo de software (ProcSoftVD - Melhoria) e pelo modelo de processo de venda e desenvolvimento de software on-demand para MPEs (ProcSoftVD). O método de melhoria de processo de software foi criado a partir de abordagens existentes e complementares, voltadas às MPEs. O ProcSoftVD foi originado com base no framework unified process, possibilitando a visualização do processo em duas perspectivas - fases e áreas de conhecimento, e nos modelos/normas de capacidade de processo CMMI-DEV e ISO/IEC 15504-5, elaborado em um processo iterativo e evolutivo de pesquisa-ação com a participação de duas MPEs. Um diferencial dessa proposta é a consideração de atividades de comercialização do software junto ao processo de desenvolvimento de software, o que auxilia na delimitação do escopo do projeto de desenvolvimento de software para um acordo contratual. Outro diferencial é o detalhamento das atividades do processo por meio de tarefas, sugestão de papéis desempenhados pelos responsáveis das atividades e disponibilização de templates com exemplos para cada um dos artefatos elaborados na execução da atividade. / The majority of the software development micro and small companies are turned to the development of on-demand software. Usually, for these small companies the formalization of a standard process for software development is very difficult. For that reason the goal of this research is to propose an integration of capability processes models/standards within software life cycle models, in a perspective of business process management. The main result of this research is a management model of selling and on-demand software development process, which embraces the software process improvement method and the selling and on-demand software development process model. The software process improvement method was build from complementary and existents tailored approaches for micro and small companies. The selling and on-demand software development process model was created based on the models/standards of process capability CMMI-DEV and ISO/IEC 15504-5 and on the unified process framework, allowing the process view from two perspectives - phases and knowledge areas. The model was elaborated in an iterative and evolutionary action-research process carried out within two micro and small companies. The originality of this proposal is the consideration of software sale activities jointly with software development process, assisting the scope delimitation of a software development project for contractual agreement. Other aspect of this research which makes it distinctive is the detail of the process activities by mean of tasks, suggestions of people roles for each activities and provision of templates with examples for each artifact created during the activities.