Software Supply Chain Map: How Reuse Networks Expand

Abstract: Clone-and-own is a typical code reuse approach because of its simplicity and efficiency. Cloned software components are maintained independently by a new owner. These clone-and-own operations can be occurred sequentially, that is, cloned components can be cloned again and owned by other new owners on the supply chain. In general, code reuse is not documented well, consequently, appropriate changes like security patches cannot be propagated to descendant software projects. However, the OpenChain Project defined identifying and tracking source code reuses as responsibilities of FLOSS software staffs. Hence supporting source code reuse awareness is in a real need. This paper studies software reuse relations in FLOSS ecosystem. Technically, clone-and-own reuses of source code can be identified by file-level clone set detection. Since change histories are associated with files, we can determine origins and destinations in reusing across multiple software by considering times. By building software supply chain maps, we find that clone-and-own is prevalent in FLOSS development, and set of files are reused widely and repeatedly. These observations open up future challenges of maintaining and tracking global software genealogies.