Evaluating workflow and process automation in wide-area software development
The effectiveness of any tool, whether workflow or not depends on the how well it co-exists with the processes it supports. If it is congruent with those processes, then there's a chance that it enhance them. If not then it may be a severe hindrance.
HyperCode automates several aspects of the current manual process and so is prima facae consistent with the process it supports. In fact, the initial response from users has so been overwhelmingly positive that it's difficult to constrain the tool's use to fit our experimental design. We attribute this response to two things: first, the elimination of the paper as the medium, and second, the elimination of synchronous interactions.
We also notice, however, scheduling difficulties, physical distance between reviewers, and physical distance between developers who must negotiate the repair of defects appears to affect the benefits of the tool. The exact extent to which this tool works, therefore, must be assessed via our rigorous experiments.
Unable to display preview. Download preview PDF.
- 1.“IEEE Standard for software reviews and audits”. IEEE Std 1028-1988, Soft. Eng. Tech. Comm. of the IEEE Computer Society, 1989.Google Scholar
- 2.Adam A. Porter and Lawrence G. Votta and Harvey P. Siy and Carol A. Toman. “An Experiment to Assess the Cost-Benefits of Code Inspections in Large Scale Software Development”. The Third International Symposium on the Foundations of Software Engineering Washington, D.C., October 1995.Google Scholar