This is a writing sample by “nycghostwriter,” AKA Barbara Finkelstein. It is a news story published on the IBM Researcher website. You can get professional ghostwriting services from an established business writer. Email me or fill out the short form on my contact page.
An Open Collaborative Research project at IBM is helping developers address the challenges of writing code in a world where effective teamwork influences the success of complex projects
Every scientist knows that the lone inventor is a myth. This is especially the case for computer scientists and software engineers for whom global collaboration has become the norm in development work. Now an Open Collaborative Research project at IBM is helping developers address the challenges of writing code in a world where effective teamwork influences the success of complex projects.
IBM is working with researchers from Carnegie Mellon University, Harvard Business School and the University of Virginia to study the collaborative relationships that define successful development teams in a flat and highly interconnected world.
The OCR project, which began exploring the interplay between software architecture and team structure in 2009, has evolved to include studying the effect of the development “ecosystem” on projects and the teams responsible for them. Software researchers are learning that the various players in this rich ecosystem are helping to shape the software development process from start to finish.
“We cannot underestimate the importance of understanding the enterprise’s socio-technical concerns and decisions in a software project,” says Clay Williams, manager of governance science research at IBM. “The cost of failed software projects in the U.S. alone is upwards of $75 billion a year in re-work costs and abandoned systems.”
Three factors for every development project
The OCR project is aligned with a strategic initiative at IBM Research that focuses on governance of software development. The OCR team is using Rational Team Concert, an IBM development tool, as a basis for studying three important factors that affect globally distributed development:
- Modularity. Distributed software development would be impossible without distinct functional units, or modules, that can be developed independently. Modules allow systems to evolve more easily over time — a valuable feature in all development processes. The OCR project includes Carliss Y. Baldwin, an economist at the Harvard Business School, who is working with her OCR colleagues to study how modular structures permit experimentation and evolution of systems as well as ensure economic value by supporting adaptation.
- Adaptation and Decisions. Replacing a module in a software system can have serious repercussions if the architecture cannot accommodate change. The team is working with Kevin Sullivan, associate professor, computer science, University of Virginia, to explore how development decisions can be understood and captured. Having such information is key to characterizing a system and understanding the kind of adaptation it has undergone — and can undergo in the future.
- Coordination and Collaboration. The structure of a system and structure of a team share significant interdependencies, especially when development is distributed. Alignment between system structure and organizational structure is a necessary condition for coordinating work effectively. Carnegie Mellon University computer scientist Jim Herbsleb is helping the OCR team consider effective coordination as a core architectural concern — along with modularity and adaptation.
Linking development activities to the business
“Ultimately, software governance is really about managing value,” says OCR team lead Clay Williams. “Understanding the interplay between technical architecture, decisions, organizational dynamics and project outcomes is key, especially when investment dollars are scarce. Our group is linking development concerns to business issues. Therein lies our value.”
IBM established the Open Collaborative Research program to support open-source software research between IBM and universities. The first open source component of the software governance project is a framework for mining Jazz® code, the foundation for Rational Team Concert®.
Read this story on an IBM Research website.
