Capiluppi, Andrea and Boldyreff, Cornelia and Beecher, Karl and Adams, Paul (2008) Quality factors and coding standards - a comparison between open source forges. In: Second International Workshop on Software Quality and Maintainability, 1-4 April 2008, Athens, Greece.

Preview

PDF sqm2008.pdf Download (178Kb)

Abstract

Enforcing adherence to standards in software development in order to produce high quality software artefacts has long been recognised as best practice in traditional software engineering. In a distributed heterogeneous development environment such those found within the Open Source paradigm, coding standards are informally shared and adhered to by communities of loosely coupled developers. Following these standards could potentially lead to higher quality software. This paper reports on the empirical analysis of two major forges where OSS projects are hosted. The first one, the KDE forge, provides a set of guidelines and coding standards in the form of a coding style that developers may conform to when producing the code source artefacts. The second studied forge, SourceForge, imposes no formal coding standards on developers. A sample of projects from these two forges has been analysed to detect whether the SourceForge sample, where no coding standards are reinforced, has a lower quality than the sample from KDE. Results from this analysis form a complex picture; visually, all the selected metrics show a clear divide between the two forges, but from the statistical standpoint, clear distinctions cannot be drawn amongst these quality related measures in the two forge samples.

Item Type:	Conference or Workshop Item (Paper)
Additional Information:	Enforcing adherence to standards in software development in order to produce high quality software artefacts has long been recognised as best practice in traditional software engineering. In a distributed heterogeneous development environment such those found within the Open Source paradigm, coding standards are informally shared and adhered to by communities of loosely coupled developers. Following these standards could potentially lead to higher quality software. This paper reports on the empirical analysis of two major forges where OSS projects are hosted. The first one, the KDE forge, provides a set of guidelines and coding standards in the form of a coding style that developers may conform to when producing the code source artefacts. The second studied forge, SourceForge, imposes no formal coding standards on developers. A sample of projects from these two forges has been analysed to detect whether the SourceForge sample, where no coding standards are reinforced, has a lower quality than the sample from KDE. Results from this analysis form a complex picture; visually, all the selected metrics show a clear divide between the two forges, but from the statistical standpoint, clear distinctions cannot be drawn amongst these quality related measures in the two forge samples.
Keywords:	Software Complexity, Coding Standards, Open Source Software
Subjects:	G Mathematical and Computer Sciences > G600 Software Engineering
Divisions:	College of Sciences > Faculty of Science > Lincoln School of Computer Science
Depositing User:	Bev Jones
Date Deposited:	15 Jul 2008 18:05
Last Modified:	13 Mar 2013 08:29
URI:	http://eprints.lincoln.ac.uk/id/eprint/1614

Actions (login required)

View Item