No abstract

No abstract

The success of software development projects depends on carefully and effectively coordinating the effort of many individuals across the multiple stages of the development process. In software engineering, modularization is the traditional technique intended to reduce the interdependencies among modules that constitute a system. Reducing technical dependencies, the theory argues, results in a reduction of work dependencies between teams developing interdependent modules. Organizational researchers have proposed similar theoretical arguments. Although such research streams have been quite influential, they have taken a coarse-grain and static view of the problem of coordination in engineering activities. This paper proposes a new perspective on coordination where fine-grain and evolving dependencies are front and central. Our empirical analyses demonstrate that considering dependencies at a fine-grain level of analysis provides us deeper insight to the relationship between technical and work dependencies. Moreover, our examination of two large scale projects from two distinct companies showed that (a) logical dependencies among software entities are significantly more important in terms of coordinating development work compared to syntactic dependencies and (b) satisfying coordination needs arising from those logical software dependencies with appropriate coordinating actions results in significant improvements on development productivity as well as a significant reduction in the failure proneness of the software systems.

This study investigates the topological form of a network and its impact on the uncertainty entrenched in descriptive measures computed from observed social network data, given ubiquitous data-error. We investigate what influence a network's topology, in conjunction with the type and amount of error, has on the ability of a measure, derived from observed data, to correctly approximate the same of the ground-truth network. By way of a controlled experiment, we reveal the differing effect that observation error has on measures of centrality and local clustering across several network topologies: uniform random, small-world, core-periphery, scale-free, and cellular. Beyond what is already known about the impact of data uncertainty, we found that the topology of a social network is, indeed, germane to the accuracy of these measures. In particular, our experiments show that the accuracy of identifying the prestigious, or key, actors in a network-according observed data-is considerably predisposed by the topology of the ground-truth network.

In this paper, we seek to shed light on how communication networks in geographically distributed projects evolve in order to address the limits of the modular design strategy. We collected data from a geographically distributed software development project covering 39 months of activity. Our analysis showed that over time a group of developers emerge as the liaisons between formal teams and geographical locations. In addition to handling the communication and coordination load across teams and locations, those engineers contributed the most to the development effort.