Practices and Advantages of Submitting Images in OSS projects : A Systematic Mapping Study and a Survey

Abstract: Background: With the increasing number of software users using social media forums, providing feedback about the OSS projects, the developer’s need to address this feedback to understand the requirements of an OSS project. As different tools support different structures for the feedback, the need to classify, prioritize and filter them into a fundamental set of categories persists. Some of the feedback includes images from users, along with the text. These images may vary from a screenshot of the bug, encountered by the user to a code snippet modification as required by the user. The significance of how these images help the developers in fixing the bug is not clear. Objectives: This thesis aims to identify the underlying advantages of using images in the feedback or bug report submitted by the user for an OSS project to the developers. The goal is to find the extent to which different image attributes help the developer’s in understanding the issue suggested in the feedback or bug report. The research also aims to classify the view of practitioners regarding which image attributes affect the most and to propose a simple DSS model that can possibly be used by users and developers while attaching images in the feedback or bug reports. Methods: In this research, we have conducted an empirical study using systematic mapping and a survey study. We identified 28 research articles form systematic mapping using a search string and snowballing process to extract different image attributes. To triangulate and verify the results of the systematic map, we have conducted an online questionnaire replied by 32 respondents experienced in contributing to the OSS community. The usability of the image attributes has been evaluated from the responses received. Both quantitative and descriptive statistical analysis techniques were used to analyze the results. Results: From the 28 research articles identified for the systematic mapping study, we have extracted 11 image attributes that influence the developers in interpreting the software requirements from the images attached to feedback or bug reports. Of the identified image attributes, image quality and image resolution are considered to be the most useful attributes by the survey respondents. Moreover, two new image attributes (timestamp and steps to reproduce) are reported from the survey study. Conclusions: The identification and validation of the image attributes suggest the potential use of images in feedback and bug reports. Furthermore, these image attributes provide additional information to the developers in understanding the software requirements from the users perspective clearly. We propose a simple DSS model that can be used by the users and the developers before attaching an image along with the feedback or the bug reports to the developing OSS communities to promote further usage of images in feedback and bug reports for OSS.