Archiving and referencing source code with Software Heritage

Software, and software source code in particular, is widely used in modern research. It must be properly archived, referenced, described and cited in order to build a stable and long lasting corpus of scientic knowledge. In this article we show how the Software Heritage universal source code archive provides a means to fully address the first two concerns, by archiving seamlessly all publicly available software source code, and by providing intrinsic persistent identifiers that allow to reference it at various granularities in a way that is at the same time convenient and effective. We call upon the research community to adopt widely this approach.