This talk provides an overview of how reproducible research is implemented in computational sciences. It will clarify key concepts such as reproducibility, replicability, and repeatability within this field, and discuss the main challenges involved in producing and evaluating reproducible algorithms, as well as the best practices to address them. Finally, it will highlight several initiatives in France that aim to promote reproducibility in scientific research. These include Guix, which enables precise control of execution environments; Notebooks, which facilitate the execution of algorithms and the presentation of results; fully peer-reviewed, reproducible journals such as IPOL; Software Heritage, which ensures the perpetual archival of source code and provides reliable, intrinsic, and permanent identifiers; and MOOCs, which offer training opportunities, among others.