Reproducibility is the hallmark of good science. Maintaining a high degree of transparency in scientific publications is essential not just for gaining trust and credibility within the scientific community but also to facilitate development of novel science (not sure if it will be obvious to naive readers why transparency facilitates novel science). Sharing data and computer code associated with publications is becoming increasingly common, motivated partly in response to data deposition requirements from journals and mandates from funders. Despite this increase in transparency, it is still difficult to reproduce or build upon the central finding (or any findings?) of most scientific publications without access to a complete workflow. 
                                        
Version control systems (VCS), which have long been used to maintain code repositories in the software industry, are now finding new applications in science. One such open-source VCS, git (do you have to argue why git instead of svn?), stands out in particular because it provides a robust framework that allows scientists to track every component (data, code, figures, and text - though aren't more complicated file types harder to deal with using git?) of a research endeavor from start to finish, with the ability to revert any file or an entire project back to any stage in its development. Since the system is decentralized, every copy of a repository not only contains all the data and code but also the entire history of changes along with detailed notes documenting each of those decisions. The power of a git repository can be further extended by linking it to a git hosting service (e.g. GitHub) which makes it easy for multiple collaborators to work asynchronously and merge their changes as needed. Among its numerous features, a git based workflow is particularly powerful for science because it is designed to protect against data loss, quickly retrace errors, and allow multiple ideas and methods to co-exist in parallel. (the word branches may not be familiar to non VCS users?). 
                                        
In this paper I review how git benefits science and why more scientists should make git-based workflows an integral part of their research. I also provide several real world use-cases for how sharing git repositories can foster collaboration, increase accountability, track contributions, in addition to lowering barriers to data reuse and fostering novel synthesis.