No description available.

This work demonstrates specific ways that the design of computer user interfaces can influence how individuals structure the problem solving process. In particular, an observational study of expert users of an image manipulation application indicates that current user interfaces make it difficult to explore sets of alternatives in parallel, despite this being a common problem solving practice. As a consequence, individuals tend to engage in highly linear problem solving processes. To address this problem, this work introduces the concept of a set-based interface, or an interface that facilitates the generation, manipulation, evaluation, and management of sets of alternative solutions. The concepts of a set-based interface are demonstrated in two tools, Side Views and Parallel Pies, both designed for use in the domain of image manipulation. Side Views automatically generates sets of previews for one or more commands and their parameters, enabling side-by-side comparison of alternatives. Parallel Pies streamlines the process of forking, or the act of creating new, standalone alternatives, and provides a visualization to evaluate results. Two controlled laboratory studies and a third think-aloud study reveal that these tools lead to users more broadly exploring the solution space and developing more optimal solutions for some types of tasks. These studies also show that the ability to broadly explore can initially be overused, adversely affecting solution quality if not enough time is spent maturing a single solution instance. This enthusiastic use of exploration tools is especially notable because such features are entirely optional to developing a solution. As such, these results suggest the need to further research ways user interfaces can support individuals in rapidly generating sets of alternative solutions. To support future research in this direction, this work contributes a set of metrics for quantifying breadth and depth of exploration; backtracking; and dead-ends in the problem solving process. A visualization called a process diagram aids in communicating these concepts.

Los patrones de usabilidad son un aspecto central en el desarrollo de software, pues estos son los encargados de sentar las bases de un conjunto de principios validados y establecidos para la creación de una apropiada interfaz de usuario. En este sentido el esfuerzo por desarrollar un modelo de patrones de usabilidad esta justificado y permite sumar una instancia mas al proceso de automatización en el desarrollo de software. En esta tesis doctoral se propone un mecanismo que ha evolucionado desde los patrones hasta una arquitectura de usabilidad, detallando el proceso evolutivo que ha llevado el mismo, define en detalle el patrón de usabilidad UNDO/REDO y construye los cimientos para extender este modelo a otros patrones de usabilidad. / Usability patterns are a central aspect of software development, as these are responsible for laying the foundations of a set of validated and established principles for creating an appropriate user interface. In this sense the effort to develop a model of usability patterns is justified and can then add another instance to process automation in software. This PhD thesis proposes a mechanism that has evolved from an architecture patterns to usability, detailing the evolutionary process that has led it defines in detail the usability pattern UNDO/REDO and builds the foundation for extending this model to other usability patterns. / La tesis contó con la doble dirección de la Dr. Patricia Pesado (UNLP) y del Prof. Oscar Dieste (UPM).

During past thirty years, several models for non-linear undo models have been presented, but almost none solves undoing and redoing actions in environments, where multiple history buffers are involved and when there are causal dependencies among separate actions. This thesis focuses on developing a new model, which allows a user to select any action from any history buffer. The key part of the model is a smart command design and undo manager, which searches for dependencies and offers possible solutions to the user. The results are presented in the context of the DaemonX framework.

Multi-user computer-aided design (CAD) is an emerging technology that promises to facilitate collaboration, enhance product quality, and reduce product development lead times by allowing multiple engineers to work on the same design at the same time. The BYU site of the NSF Center for e-Design has developed advanced multi-user CAD prototypes that have demonstrated the feasibility and value of this technology. Despite the possibilities that this software opens up for enhanced collaboration, there are now a new variety of challenges and opportunities to understand and address. For multi-user CAD to be used effectively in a modern engineering environment, it is necessary to understand and address both human and technical collaboration challenges. The purpose of this dissertation is to understand and address these challenges. Two studies were performed to better understand the human side of engineering collaboration: (1) engineers from multiple companies were interviewed to assess the collaboration challenges they experience, and (2) players of the multi-player game Minecraft were surveyed and studied to understand how a multi-user environment affects design collaboration. Methods were also developed to address two important technical challenges in multi-user CAD: (1) a method for detecting undo conflicts, and (2) additional methods for administering data access. This research addresses some of the important human and technical collaboration challenges in multi-user CAD. It enhances our understanding of collaboration challenges in engineering industry and how multi-user CAD will help address some of those challenges. It also enhances our understanding of how a multi-user design environment will affect design collaboration. The method developed for detecting conflicts that occur during local undo in multi-user CAD can be used to block conflicts from occurring and provide the user with some information about the cause of the conflict so they can collaborate to resolve it. The methods developed for administering data access in multi-user CAD will help protect against unauthorized access to sensitive data.

Undo is a command that users rely on in most software applications. Its importance in multi-user CAD is no different. However, due to its unique nature, users may have different expectations regarding how undo should behave. This research seeks to better understand users' expectations regarding undo in multi-user CAD by having users participate in collaborative design exercises and then asking them how they would expect undo to behave. In addition, users participated in a survey in which they watched 8 videos showing users interacting within multi-user CAD and were then asked about how they would want undo to behave. Based on these findings, various recommendations are made for how undo ought to be implemented in multi-user CAD. These include recommendations regarding the user experience/user interface as well as proposing an equation that seeks to quantify whether a user expects to share an undo stack with another user or if they should only be able to undo their own actions.

L'importance des systèmes collaboratifs a considérablement augmenté au cours des dernières années. La majorité de nouvelles applications sont conçues de manière distribuée pour répondre aux besoins du travail collaboratif. Parmi ces applications, nous nous intéressons aux éditeurs collaboratifs temps-réel (RCE) qui permettent la manipulation de divers objets partagés, tels que les pages wiki ou les articles scientifiques par plusieurs personnes réparties dans le temps et dans l'espace. Bien que ces applications sont de plus en plus utilisées dans de nombreux domaines, l'absence d'un modèle de contrôle d'accès adéquat limite l'exploitation de leur plein potentiel. En effet, contrôler les accès aux documents partagés de façon décentralisée et sans alourdir les performances du système collaboratif représente un vrai challenge, surtout que les droits d'accès peuvent changer fréquemment et de façon dynamique au cours du temps. Dans cette thèse, nous proposons un modèle de contrôle d'accès générique basé sur l'approche de réplication optimiste du document partagé ainsi que sa politique de contrôle d'accès. Pour cela, nous proposons une approche optimiste de contrôle d'accès dans la mesure où un utilisateur peut violer temporairement la politique de sécurité. Pour assurer la convergence, nous faisons recours à l'annulation sélective pour éliminer l'effet des mises à jour illégales. Vu l'absence d'une solution d'annulation générique et correcte, nous proposons une étude théorique du problème d'annulation et nous concevons une solution générique basée sur une nouvelle sémantique de l'opération identité. Afin de valider notre approche tous nos algorithmes ont été implémentés en Java et testés sur la plateforme distribuée Grid'5000 / The importance of collaborative systems in real-world applications has grown significantly over the recent years. The majority of new applications are designed in a distributed fashion to meet collaborative work requirements. Among these applications, we focus on Real-Time Collaborative Editors (RCE) that provide computer support for modifying simultaneously shared documents, such as articles, wiki pages and programming source code by dispersed users. Although such applications are more and more used into many fields, the lack of an adequate access control concept is still limiting their full potential. In fact, controlling access in a decentralized fashion for such systems is a challenging problem, as they need dynamic access changes and low latency access to shared documents. In this thesis, we propose a generic access control model based on replicating the shared document and its authorization policy at the local memory of each user. We consider the propagation of authorizations and their interactions. We propose a optimistic approach to enforce access control in existing collaborative editing solutions in the sense that a user can temporarily violate the access control policy. To enforce the policy, we resort to the selective undo approach in order to eliminate the effect of illegal document updates. Since, the safe undo is an open issue in collaborative applications. We investigate a theoretical study of the undo problem and propose a generic solution for selectively undoing operations. Finally, we apply our framework on a collaboration prototype and measure its performance in the distributed grid GRID?5000 to highlight the scalability of our solution

Main goal of this master´s thesis is to propose and implement improvements to an original purely raster version of Flame Painter editor. An enhancement of the undo/redo system with emphasise to its functionality and memory requirements has been implemented. Moreover, the editor was extended by adding an ability to edit brush strokes using vector primitives which effectively assist the user in future stroke editing. This project was created in cooperation with employees of the Escape Motions Company.