What's a Typical Commit? A Characterization of Open Source Software Repositories

The research examines the version histories of nine open source software systems to uncover trends and characteristics of how developers commit source code to version control systems (e.g., subversion). The goal is to characterize what a typical or normal commit looks like with respect to the number of files, number of lines, and number of hunks committed together. The results of these three characteristics are presented and the commits are categorized from extra small to extra large. The findings show that approximately 75% of commits are quite small for the systems examined along all three characteristics. Additionally, the commit messages are examined along with the characteristics. The most common words are extracted from the commit messages and correlated with the size categories of the commits. It is observed that sized categories can be indicative of the types of maintenance activities being performed.

• Type: Conference
• Year: 2008
• Where: IWPC
• Authors: Abdulkareem Alali, Huzefa H. Kagdi, Jonathan I. Maletic
• Keywords: Commit Messages, Developers Commit Source, IWPC 2008, Normal Commit, Software Engineering