Very little research in software engineering has focused on the model of a programmer working alone. These so-called cowboys are disdained for not working in teams to build software. In reality many cowboys work by themselves due to the circumstances of their work environment, not because they are unable or unwilling to work with well with others. These solo programmers could benefit from a methodology to assist them in consistently developing reliable software on time and within budget while satisfying the customer's needs. Cowboy was designed to help fill this void. This agile-based system incorporates the benefits of agile methodologies into a lightweight, customer-centered approach to software development for the lone developer. This thesis describes Cowboy and its successful application in developing a prototype for a web application.

Dans de nombreux domaines tels que l’avionique, la médecine ou la domotique, les applications logicielles jouent un rôle de plus en plus important, allant jusqu’à être critique pour leur environnement. Afin de pouvoir faire confiance à ces applications, leur développement est contraint par des exigences de sûreté de fonctionnement. En effet il est nécessaire de démontrer que ces exigences de haut-niveau sont prises en compte tout au long du cycle de développement et que des solutions concrètessont mises en œuvre pour parvenir à les respecter. De telles contraintes rendent le développement d’applications sûres de fonctionnement particulièrement complexe et difficile. Faciliter ce processus appelle à la recherche de nouvelles approches dedéveloppement qui intègrent des concepts de sûreté de fonctionnement et guident les développeurs lors de chacune des étapesnécessaires à la production d’une nouvelle application digne de confiance.Cette thèse propose ainsi de s’appuyer sur une approche dirigée par la conception pour guider le développement des applications sûres de fonctionnement. Cette approche est concrétisée à travers une suite d’outils nommée DiaSuite et offre du support dédié à chaque étape du développement. En particulier, un langage de conception permet de décrire à la fois les aspects fonctionnels et non-fonctionnels des applications en se basant sur un paradigme dédié et en intégrant des concepts de sûreté de fonctionnement tels que le traitement des erreurs. A partir de la description d’une application, du support est généré pour guider les phases d’implémentation et de vérification. En effet, la génération d’un framework de programmation dédié permet de guider l’implémentation tandis que la génération d’un modèle formel permet de guider la vérification statique de l’application et qu’un support de simulation permet de faciliter les tests. Cette approche est évaluée grâce à des cas d’études réalisés dans les domaines de l’avionique et de l’informatique ubiquitaire. / In many domains such as avionics, medecine or home automation, software applications play an increasingly important rolethat can even be critical for their environment. In order to trust these applications, their development is contrained by dependability requirements. Indeed, it is necessary to demonstrate that these high-level requirements are taken into account throughout the development cycle and concrete solutions are implemented to achieve compliance. Such constraints make the development of dependable applications particularly complex and difficult. Easing this process calls for the research of new development approaches that integrate dependability concepts and guide the developers during each step of the development of trustworthy applications.This thesis proposes to leverage a design-driven approach to guide the development of dependable applications. This approachis materialized through a tool-suite called DiaSuite and offers dedicated support for each stage of the development. Inparticular, a design language is used to describe both functional and non-functional applications. This language is based on adedicated paradigm and integrates dependability concepts such as error handling. From the description of an application, development support is generated to guide the implementation and verification stages. Indeed, the generation of a dedicated programming framework allows to guide the implementation while the generation of a formal model allows to guide the static verification and simulation support eases the testing. This approach is evaluated through case studies conducted in the domains of avionics and pervasive computing.

Unlike many proprietary component systems, the academic ones do not have sufficient support in integrated development environments. This the- sis analyzes development of component-based applications in terms of the SOFA 2 component system and it finds out that the main issue is an in- sufficient connection between processes of common application design and creation of particular components. Based on this analysis, it defines a subset of the UML, a universal language for application design, and its semantics regarding entities of the SOFA 2 component system. Furthermore, it creates a tool integrated to the Eclipse IDE, which enables a developer to automat- ically generate these entities from a UML component model as well as to connect this model with already existing entities enabling their automatic correction in case of model changing. This tool is designed modularly so that it is possible to easily change semantics of the model or using it for other models. Finally, this thesis analyzes possibilities of extensions of this tool for other component systems, code generation and component behaviour verification.

The need for a robust system architecture to support software development is well known. In enterprise software development, this must be realized in a multi-tier environment for deployment to a software framework. Many popular integrated development environment (IDE) tools for component-based frameworks push multi-tier partitioning by assisting developers with convenient code generation tools and software deployment tools which package the code. However, if components are not packaged wisely, modifying and adding components becomes difficult and expensive. To help manage change, vertical partitioning can be applied to compartmentalize components according to function and role, resulting in a grid partitioning. This thesis is to advocate a design methodology that enforces vertical partitioning on top of the horizontal multitier partitioning, and to provide guidelines that document the grid partitioning realization in enterprise software development processes as applied in the J2EE framework.

FOOPS is a concurrent object oriented specification language with an executable subset. In this thesis we propose an extension of FOOPS with features for specifying systems of distributed and autonomous objects. This extension supports most features of concurrent object oriented programming, including classes of objects with associated methods and attributes, object identity, dynamic object creation and deletion, overloading, polymorphism, inheritance with overriding, dynamic binding, concurrency, nondeterminism, atomic execution, evaluation of method expressions as background processes, and object protection. The main contribution of this thesis is to develop a framework for supporting formal development of software in the extension of FOOPS mentioned above. In particular, we introduce a structural operational semantics for FOOPS, a notion of refinement for concurrent object oriented programs, congruence properties of refinement of FOOPS programs, and tools for mechanising refinement proofs. The operational semantics is the core of the formal definition of FOOPS. It is used to define notions of refinement for FOOPS states, programs, and specifications. Those notions and associated proof techniques for proving refinement are used to illustrate stepwise formal development of programs in FOOPS. The congruence properties of refinement (with respect to some of FOOPS operators) justify compositional development of software in FOOPS. The tools help to validate the framework introduced in this thesis and motivate its use in practice.

When delivering Software as a Service (SaaS), increased challenges regarding responsiveness and flexibility are introduced to software providers. In order to address these challenges and attain customer satisfaction, the Lean Software Development principle Eliminate waste can be applied. One organization that has recognized the potential of applying the Eliminate waste principle when developing software is the Integration & BI unit of the Enterprise Resource Planning system provider Infor M3. However, the implementation of the Eliminate waste principle at the Integration & BI unit’s five teams is still in an early stage. Consequently, the intended purpose of the thesis was to identify waste and suggest approaches to reduce waste at the case organization, Infor M3 Integration & BI. In order to collect the in-depth knowledge required, the thesis utilized a qualitative case study methodology, whereby a literature review, interviews and observations were conducted. The literature review created a foundation of knowledge regarding waste in software development, that subsequent culminated as a basis for the analysis and recommendations. It could be concluded that the subject of waste identification and reduction in software development is in an early stage, largely driven by practitioners, with few verifying studies that support the subject’s applicability. However, by utilizing a waste categorization model various wastes could be identified at all of Integration & BI’s teams during the interviews, whereupon Partially done work, Delays, Task switching and Relearning was considered as the most prominent wastes. Moreover, it could be established that one team had developed successful approaches that eliminates much of the team’s waste whilst the other teams’ approaches were generally deficient. In order to more successfully reduce waste, the Integration & BI unit is suggested to create awareness of the concept of waste within the unit. The teams need a common definition and an increased understanding of waste in order to reach this awareness. Additionally, the unit is suggested to use more comprehensive indicators, like cumulative flow diagram, in order to facilitate identification and root-cause analysis of waste. Lastly, the unit is recommended to reduce waste by continuous improvements with activities structured as a PDSA-cycle. / Programvaruföretag som levererar Software as a Service (SaaS) står inför ökade utmaningar med avseende på receptivitet och flexibilitet. För att bemöta dessa utmaningar och uppnå hög kundnöjdhet kan Lean Software Development principen Eliminera slöseri appliceras. En organisation som har insett potentialen av att tillämpa principen Eliminera slöseri vid utveckling av programvara är enheten Integration & BI hos affärssystemleverantören Infor M3. Implementeringen av principen vid enhetens fem utvecklingslag är emellertid fortfarande i ett tidigt skede. Avsikten med examensarbetet var således att identifiera slöseri och föreslå tillvägagångssätt för att minska slöseri hos fallorganisationen, Infor M3 Integration & BI. För att samla in en djup förståelse om ämnet och fallföretaget utnyttjade examensarbetet en kvalitativ fallstudiemetodik, där en litteraturstudie, intervjuer och observationer genomfördes. Litteraturstudien skapade kunskap angående slöseri inom mjukvaruutveckling, vilken senare skapade en grund för både analys och rekommendationer. Slutsatsen att ämnet rörande identifiering och minimering av slöseri inom mjukvaruutveckling är i ett tidigt skede kunde göras i och med att ämnet i stor utsträckning drivs av yrkesutövare och inte av akademin. Dessutom finns det få vetenskapliga studier som verifierar ämnets applicerbarhet. Genom att använda en modell för att kategorisera olika slöseri inom mjukvaruutveckling kunde dock olika typer av slöseri identifieras hos samtliga av Integration & BIs utvecklingslag, varpå delvist gjort arbete, förseningar, uppgiftsbyte och återinlärning betraktades som de mest framträdande. Dessutom kunde det fastställas att ett av utvecklingslagen etablerat framgångsrika tillvägagångssätt för att eliminera slöseri under tiden de andra utvecklingslagens metoder generellt var bristfälliga. För att mer framgångsrikt minska slöseri, föreslås enheten Integration & BI att använda en gemensam definition och öka förståelsen om konceptet slöseri för att skapa en medvetenhet inom enheten. Dessutom rekommenderas enheten att använda mer omfattande indikatorer så som kumulativt flödesschema för att underlätta identifiering och rotorsaksanalys av slöseri. Slutligen föreslås enheten att eliminera slöseri genom ständiga förbättringar med aktiviteter strukturerade efter PDSA-cykeln.

This work deals with methodologies of software development, particularly web applications. Its main objective is to propose a methodology that helps ensure quality web development and prevents a number of frequently recurring problems. The first chapter deals with characteristics of current state and defines the need for process and discipline, the need for team development and the importance of appropriate tools. The second chapter describes the specifics of the web development. It defines the quality criteria in relation to web applications, particularly the aspects of accessibility, usability and security are concerned. It deals with web frameworks and the benefits and risks arising from their usage. It also outlines requirements for the chosen methodology for web development and describes types of projects to be applied to the methodology. The third chapter analyzes the suitability of the selected life cycle models and methodologies in software development for web applications. The life cycle models consider a model of "program and repair", the waterfall model and incremental model. It also deals with the assumptions of agile methodologies and their level of satisfaction for web development. For the analysis is used the methodology evaluation system METES with which is examined Rational Unified Process, Open Unified Process, Extreme Programming, Scrum and Feature Driven Development. The fourth chapter deals with the characteristic of Open Unified Process methodology. It describes the principles, features and development process. The main chapter deals with the proposal of methodology for web development. As the basis is chosen Open Unified Process methodology, whose elements are supplemented and adapted for web development. As the elements of the methodology are elaborated principles, practices, roles, work products, tasks, delivery process, manuals and guidelines and tools. Each element is adapted to fit the specific needs of web applications. In order to achieve greater control over the quality of web applications there are also included detailed checklists as an integral part of products which are aimed at preventing the most frequently occurring errors and problems. The chapter concludes with an overview of tools that have been validated in web development.

[pt] Na literatura disponível hoje, encontram-se exemplos da utilização do Scrum em projetos e times pequenos, deixando um questionamento sobre a possibilidade de utilização desta metodologia em projetos e/ou times grandes. Esta dissertação tem por objetivo examinar casos práticos de empresas conhecidas onde foi utilizado Scrum em projetos grandes, enfatizando as dificuldades encontradas ao longo de todo processo e as soluções adotadas, destacando as práticas que levaram os projetos a obter sucesso. / [en] In today s writings, there are many examples of using Scrum in small teams and projects, leaving a knowledge gap about the possibility of applying Scrum in big teams and large projects. This work aims at presenting case studies in known companies where Scrum has been applied in large projects, emphasizing the roadblocks found throughout the process, and the solutions adopted, highlighting the practices that lead the projects to success.

A Research Report Submitted to the Faculty of Engineering and the Built Environment in partial fulfilment of the Requirements for the degree of Master of Science in Engineering / The research centred on improving the current software systems development life cycle (SDLC) of the credit scoring process at a financial institution based on systems engineering principles. The research sought ways to improve the current software SDLC in terms of cost, schedule and performance. This paper proposes an improved software SDLC that conforms to the principles of systems engineering. As decisioning has been automated in financial institutions, various processes are developed according to a software SDLC in order to ensure accuracy and validity thereof. This research can be applied to various processes within financial institutions where software development is conducted, verified and tested. A comparative analysis between the current software SDLC and a recommended SDLC was performed. Areas within the current SDLC that did not comply with systems engineering principles were identified. These inefficiencies were found during unit testing, functional testing and regression testing. An SDLC is proposed that conforms to systems engineering principles and is expected to reduce the current SDLC schedule by 20 per cent. Proposed changes include the sequence of processes within the SDLC, increasing test coverage by extracting data from the production environment, filtering and sampling data from the production environment, automating functional testing using mathematical algorithms, and creating a test pack for regression testing which adequately covers the software change. / MT2016

Context. Software Development has been witnessing great innovations over past few years with good number of technologies, tools and practices invading the industry. Client demands and collaboration in the development process are also seen increasing. So many new practices and methodologies are coming up and Agile is one of prominent practices adopted by many. Even in Agile, Scrum methodology is picking up more demand presently. As Software Development process and practices are changing so are the leadership styles in the same. Leadership is critical for success of any team. This study intended to explore this model and requirement of leadership in Scrum based Software Development in a practical scenario. Objectives. Leadership, which is essential in any Software project, differs from Traditional Methodology to Scrum Methodology of Agile practices in Software Development. Through a case study, the author attempted to investigate and explore the perspective of Leadership requirement in Scrum based Software Development in a practical scenario. The study aimed to gather and analyze the Leadership model implemented in two domestic projects in an Indian company, involved in two distinctive domains, and sum up the impressions gained in the same. The study aimed to assess whether the gathered knowledge adds up to the existing body of knowledge on the phenomenon or on the contrary whether any suggestions for improvement can be given to the case units. Methods. Case Study method was chosen for undertaking this explorative study. A literature review was conducted prior to the case study to gain knowledge on the phenomenon, which also answered one of the Research Questions and helped partially the other. A multiple case study was conducted through semi structured personal interviews, tools analysis and direct observation in the case units. Qualitative data analysis was made using Grounded Theory on this three orders of collected data. The results were compared with the Literature and conformity or variance analyzed. This comparative analysis is used for making recommendations to the case units for improvement or for additions to the existing body of knowledge. Results. Through the results of Literature Review, Leadership models in Software Development including Agile Scrum were summed up. And through the results of the case study, the leadership models and features implemented in the case units have been identified. These results are further validated and contrasted with the results of the literature review. How the literature models and the case unit models of leadership differed is studied. The justification for the implemented leadership model in the practical situation is also analyzed. Following, a review of the models employed at the case environments, the perspective of leadership in the two Scrum based Software Development projects is summed up. At the end, it is assessed what effect the case study would have on the existing body of knowledge on the phenomenon and modifications that can be proposed to the case units based on the results and analysis. Conclusion. It is concluded that the Case Units are implementing Situational Leadership and Transformational Leadership in a mixed way. Telling and Selling models in Situational Leadership are prominent while Participating and Delegating are ranking less. Some of the important features of Transformational Leadership like Self Management, Organizational Consciousness, Adaptability and Proactive are in implementation but not all features of the model are assumed. Even Scrum is implemented in a modified way, extending only controlled autonomy with higher monitoring and it had a direct effect on the leadership. On the whole it is directive leadership that is in play with co-existence of collaborative one situationally.