Project

software visualization; collaborative software engineering; software evolution; mining software archives; software development

Lead
Lay summary
Modern software development has in the past two decades gone through radical changes due to higher-productivity tools and techniques to develop and analyze software. In practice this prolongs the lifetime of systems, and therefore generates novel problems affecting them. It is nowadays indeed difficult to make a distinction between developing and maintaining software, because modern techniques have made systems undergo a transition towards ``permanent evolution''. In recent times software industry is facing a new phenomenon: distributed, collaborative software engineering. Indeed, software is nowadays often built by geographically separated development teams, which leads to a number of challenges, namely awareness (the understanding of other's activities to provide a context for one's activities), communication (the exchange of information between developers working remotely), and synchronization (the versioning of the system and the coordination of development activities). Despite recent efforts by a number of researchers, these and other issues concerning collaborative software engineering remain largely unaddressed.The goal of this project is to build on our previous work on "Change-based Software Evolution" in the context of the predecessor project REBASE, and extend it to "Global Synchronous Software Development", a collaborative context in which we model in a fine-grained way the real-time changes that are performed in parallel by developers working on a system.

Direct link to Lay Summary

Last update: 21.02.2013

Active participation

Self-organised

Active participation

Title	Year

Number	Title	Start	Funding scheme

Modern software development has in the past two decades gone through radical changes due to higher-productivity tools and techniques to develop and analyze software. In practice this prolongs the lifetime of systems, and therefore generates novel problems affecting them. It is nowadays indeed difficult to make a distinction between developing and maintaining software, because modern techniques have made systems undergo a transition towards ``permanent evolution''. In recent times software industry is facing a new phenomenon: distributed, collaborative software engineering. Indeed, software is nowadays often built by geographically separated development teams, which leads to a number of challenges, namely awareness (the understanding of other's activities to provide a context for one's activities), communication (the exchange of information between developers working remotely), and synchronization (the versioning of the system and the coordination of development activities). Despite recent efforts by a number of researchers, these and other issues concerning collaborative software engineering remain largely unaddressed.The goal of this continuation project is to build on our previous work on "Change-based Software Evolution" in the context of the predecessor project REBASE, and extend it to "Global Synchronous Software Development", a collaborative context in which we model in a fine-grained way the real-time changes that are performed in parallel by developers working on a system. In particular, the GSync project aims to answer the following research questions: 1. How can we model a collaborative change-centric approach? How should concurrent changes be recorded, stored, and shared among workspaces? How can we detect arising commit conflits before they happen, and how can we resolve them? 2. How can we visually augment workspace awareness? How can we assist teams to coordinate development activities? How can we model the relationships of trust that are in place when developers work collaboratively? Can we augment the productivity of developers working in parallel? 3. How can we use 3D visualization to render the complex information generated by collaborative development? How can we step beyond the current state-of-the-art IDEs and create an immersive collaborative development environment?For the GSync project we propose to investigate the following research tracks:1. Enabling Synchronous Development. We will continue our work on change-based software evolution by devising the infrastructure to take into account collaborative aspects, by addressing the issues that arise when multiple developers work at the same time on the same system. In detail, this implies the creation of a client-server application that records every change happening to a system in real time and then broadcasts this information to all developers. The client side is represented by Syde, the Eclipse IDE plugin we have been developing, which we want to extend to model in a more accurate way, namely at the abstract syntax tree (AST) level, what happens to a system as it is being changed. The server side will serve as central repository for all the changes, and will also broadcast changes to all clients. We will also investigate how in a collaborative context commit conflicts can be detected and resolved before they happen, and will use this as a basis to develop the means to provide assistance to developers working in a collaborative context.2. Immersive Synchronous Development. Our goal is to use visualization to convey to developers the events that pertain to the first research track, which deals mostly with infrastructural issues that must remain invisible to the developers to minimize information overload. We will devise visual means to augment workspace awareness, to warn of potential conflicts, and in a latter moment to proactively assist developers. In the last part of the project, our goal is to build, on top of the Eclipse IDE, an immersive 3D environment that supports synchronous development activities.3. Demonstrator & Industrial Validation. This track integrates the first two, as they are intrinsically intertwined, i.e., the first track will provide the infrastructure for the second one, which in turn will provide valuable feedback for the first one. Our goal is to periodically publicly release an integrated version of our software prototypes, which in turn is also needed to potentially attract industrial partners for a real-world validation of our research endeavours.