Parallel software is now required to exploit the abundance of threads and processors in modern multicore computers. Unfortunately, manual parallelization is too time-consuming and error-prone for all but the most advanced programmers. While automatic parallelization promises threaded software with little programmer effort, current auto-parallelizers are easily thwarted by pointers and other forms of ambiguity in the code. In this dissertation we profile the loops in SPEC CPU2006, categorize the loops in terms of available parallelism, and focus on promising loops that are not parallelized by IBM's XL C/C++ V10 auto-parallelizer. For those loops we propose methods of improved interaction between the programmer and compiler that can facilitate their parallelization. In particular, we (i) suggest methods for the compiler to better identify to the programmer the parallelization-blockers; (ii) suggest methods for the programmer to provide guarantees to the compiler that overcome these parallelization-blockers; and (iii) evaluate the resulting impact on performance.

Parallel software is now required to exploit the abundance of threads and processors in modern multicore computers. Unfortunately, manual parallelization is too time-consuming and error-prone for all but the most advanced programmers. While automatic parallelization promises threaded software with little programmer effort, current auto-parallelizers are easily thwarted by pointers and other forms of ambiguity in the code. In this dissertation we profile the loops in SPEC CPU2006, categorize the loops in terms of available parallelism, and focus on promising loops that are not parallelized by IBM's XL C/C++ V10 auto-parallelizer. For those loops we propose methods of improved interaction between the programmer and compiler that can facilitate their parallelization. In particular, we (i) suggest methods for the compiler to better identify to the programmer the parallelization-blockers; (ii) suggest methods for the programmer to provide guarantees to the compiler that overcome these parallelization-blockers; and (iii) evaluate the resulting impact on performance.

No description available.

Two studies of factors affecting the programming performance of first- and second year Computer Science students were conducted. In one study students used GIL, a simple application framework, for their programming assignments in a second-semester programming course. Improvements in student performance were realized. In the other study, students submitted detailed logs of how time was spent on projects, along with their programs. Software metrics were computed on the students' source code. Correlations between student performance and the log data and software metric data were sought. No significant indicators of performance were found, even with factors that are commonly expected to indicate performance. However, results from previous research concerning variations in individual programmer performance and relationships between software metrics were obtained. / Master of Science

This study dealt with the problem of investigating the validity of two recently developed paper-and-penci1 tests of programmer aptitude , the Aptitude Test for Programmer Personnel (ATPP) and the Computer Programmer Aptitude Battery (CPAB), in predicting success in computer programming.

he main purpose of this semestral thesis is to get to know programming of Atmel AVR microcontroller, working with SD/MMC cards, working with ethernet (TCP/IP) to create a standalone programmer of AVR microcontroller with ethernet connectivity. The first part deals with the questions of programming theory of memories inside microcontrollers and description of programming algorithm. The most used programming method is called „In System Programming“ - ISP. This method provides easy and fast manipulaton. Concept of this programmer is designed for this ISP method. In this thesis is too explain how to work SD/MMC card and file systém FAT16. This knowledges are necesary to store binary program inside standalone programmer. In final part and in attachments there is hardware design, electrical scheme, bill of materials, printed circuit board.

No description available.

This thesis argues the case for the use of a pattern language based on the basic features of the programming language used in instruction for the teaching of programming. We believe that the difficulties that novices are known to have encountered with the task of learning to program ever since the inception of computers derive from a basic misfit between the language used to communicate with a computer, the programming language, and the way that humans think. The thrust of the pattern language idea is that patterns are the essential element in understanding how the mind words in that they are the source of that relationship that we call `meaning'. What an entity or event `means' to us derives from the effect that it has on us as living biological beings, a relationship that exists in the `real world', not from any linguistic relationship at the symbolic level. Meaning, as a real world relationship, derives from the patterns of interactions that constitute being. The meaning that an entity has for an individual is more than can be expressed in a formal definition, definitions are matters of agreement, convention, not the pattern of experience that the individual has acquired through living. What is missing for a novice in any skill acquisition process is meaning, the pattern of experience. All that we can give them using a formal linguistic system like a programming language is definitions, not meaning. Pattern language is the way that we think because it exists at that fundamental level of experience as living beings. The patterns of experience become the patterns of thought through recurrence, not through definition. But this takes time, so in presenting new material to a person trying to learn, we have to present it in the form of a pattern language, the 'cognitive map' that drives the problem solving process. Creativity is always a function of combining ideas, what is really being created is new meaning, not a program, or a house, or a poem, or a sculpture - these things are mere implementations of meaning. Ultimately meaning can derive only from experience, the pattern of life around us, so creativity is the language of experience, pattern language. The mind is the product of experience, creativity its modus operandi.

Soft skills are characteristics associated with an individual’s personality. They are relevant to professional qualification because they improve the performance at work. Since they are important, Information Technology companies need to understand the soft skills to each role in software development process. Additionally, during the hiring process these companies need to identify soft skills in candidates to find out which one of them have the required characteristics to fit the available jobs. However, soft skills identification is a hard task because it takes time to know an individual’s behavior and normally uses interviews or recommendations. Therefore, we notice a lack of automatic approaches in this context. This dissertation proposes a strategy to minimize the problem of soft skills identification. The strategy focus on the role of software programmers and it aims to find ways to automatically identify soft skills of individuals in this role. To do so, we propose a set of metrics that evaluate soft skills. We collect the metrics from an online judge system, according to its users’ performance and activities. To evaluate the metrics, we conduct an empirical study regarding 56 students of programming courses. Our results indicate that the metrics to identify Analytical and solving problems skills, Attention to details, Fast learning and Persistence are satisfactory. On the other hand, Communication and Work independently skills did not reach significant results. / Soft skills são características associadas a personalidade de um indivíduo. Consideradas relevantes para compor o perfil de um profissional qualificado, elas melhoram o desempenho no trabalho. Diante de sua importância, empresas de Tecnologia da Informação precisam entender quais soft skills são necessárias para cada papel no processo de desenvolvimento de software. Além disso, durante o processo de contratação, essas empresas precisam identificar soft skills em candidatos a fim de descobrir quais deles possuem as características exigidas para os cargos disponíveis. No entanto, a identificação de soft skills é uma tarefa difícil, pois exige conhecer um indivíduo e seu comportamento por um tempo. Normalmente também requer esforços como entrevistas e recomendações, tendo sido observada a falta de abordagens automáticas nesse contexto. Esta dissertação propõe uma estratégia para minimizar o problema da identificação de soft skills. Tal estratégia foca no papel do programador de software e tem como objetivo encontrar formas para identificar automaticamente soft skills de indivíduos nesse papel. Para isso, propomos um conjunto de métricas que pontuam soft skills. Coletamos essas métricas a partir de um juiz online, de acordo com o desempenho e atividades de usuários no sistema. Para avaliar as métricas propostas, conduzimos um estudo empírico envolvendo 56 estudantes de cursos de programação. Nossos resultados indicam que as métricas para identificar as soft skills Análise e resolução de problemas, Atenção a detalhes, Aprendizagem rápida e Persistência são satisfatórias. Por outro lado, as métricas relativas às soft skills de Comunicação e Trabalho independente não alcançaram resultados significativos.

Scania, one of the heavy engines manufacturers, produces Scania Diagnos Programmer 3 (SDP3) to facilitate repair process in their workshops. SDP3 is localizable software which challenges to separate User Interface strings (UI strings) during development process and later combine with the localized strings for local user access. The objective of this report is to provide knowledgeable solution for Graphical User Interface (GUI) development, especially with respect to synchronization of UI strings in SDP3.The migration of SDP3 from .NET 3.0v framework to .NET 3.5v framework satisfies modern standards and needs. With regards to migration of SDP3’s localization process, I have attempted to summarize major .NET 3.5v framework methods that can be used for localization of GUI text in SDP3. Experiments show that tools used to facilitate the localization process also lack important features. Although pre-build process and post-build process provide promising solutions for localization, using them along with some proprietary localization tool should result in more features, better and faster production cycle. However, proprietary localization tool have to be used with anyone of the localization methods.