1,721,203 research outputs found
Why Software Reliability Predictions Fail
No full textSoftware reliability reflects a customer's view of the products software engineers build and test, as it is usually measured in terms of failures experienced during regular system use. However, this testing strategy is often based on early product measures, since failures cannot be measured until the software is placed in the field. Thus, this measurement is not effective at predicting the likely reliability of the delivered software
Empirical evaluation of software maintenance technologies
No full textTechnology evaluation is part of the decision-making process of any software organization. Unlike conventional wisdom, empirical evaluation strives to avoid biased conclusions by relying on observation and looking for pitfalls in the evaluation process. In this paper, we provide a summary of the maintenance studies presented in the session "Study and assessment of (new) technologies" of the International Workshop on Empirical Studies of Software Maintenance (WESS '96), and also report on the working group discussion which focused on common problems and open issues in the field of technology evaluation. These empirical studies are then classified according to a multi-dimensional framework to synthesize the state of the research in technology evaluation and ultimately discover interesting patterns
Building Trust through Social Awareness: The SocialCDE Project
No full textTrust is paramount in distributed software development to prevent geographically distributed sites to feel and act like distinct, distant teams. Nevertheless, how to build trust among developers with few or no chances to meet is an open issue. To overcome such a challenge we hypothesize that increased social awareness may foster trust building in global software teams. Here, we first present SocialCDE, a tool that aims at augmenting Application Lifecycle Management (ALM) platforms with social awareness to facilitate the establishment of interpersonal connections by disclosing developers’ personal interests and contextual information. Then, we present two different empirical studies, specifically designed to test our hypothesis
- …
