Two papers have been accepted to the CSCW 2013 conference: “The work of developing cyberinfrastructure middleware projects” and “Globally distributed system developers: Their trust expectations and processes.”
Bietz, M. J., Paine, D., & Lee, C. P. “The work of developing cyberinfrastructure middleware projects.”
Abstract: Middleware, which provides an abstraction layer between low-level computational services and domain-specific applications, is a key component of cyberinfrastructure. This paper presents a qualitative study of how cyberinfrastructure middleware development is accomplished in two supercomputing centers. Our investigation highlights key development phases in the lives of middleware projects. Middleware development is typically undertaken as part of collaborations between technologists and domain scientists, and middleware developers must balance the pressure to meet specific scientific needs and the desire to explore their own R&D agendas. We explore how developers work to sustain an ongoing development trajectory by aligning their own work with particular domain science projects and funding streams. However, we find that the key transition from being a component in a domain-specific project to a stand-alone system that is useful across domains is particularly challenging for middleware development. We provide organizational and national policy implications for how to better support this transition.
Al-Ani, B., Bietz, M. J., Wang, Y., Trainer, E., Koehne, B., Marczak, S., et al. “Globally distributed system developers: Their trust expectations and processes.”
Abstract: Trust remains a challenge in globally distributed development teams. In order to investigate how trust plays out in this context, we conducted a qualitative study of 5 multi-national IT organizations. We interviewed 58 individuals across 10 countries and made two principal findings. First, study participants described trust in terms of their expectations of their colleagues. These expectations fell into one of three dimensions: that socially correct behavior will persist, that team members possess technical competency, and that individuals will demonstrate concern for others. Second, our study participants described trust as a dynamic process, with phases including formation, dissolution, adjustment and restoration. We provide new insights into these dimensions and phases of trust within distributed teams which extend existing literature. Our study also provides guidelines on effective practices within distributed teams in addition to providing implications for the extension of software engineering and collaboration tools.
See you in San Antonio!