Paper short abstract:

Software is much more than just code, and it is undervalued to understand the digitalization of research practices. The dynamics of circulation of software in Computational Chemistry unveil how software packages were and are distributed, maintained and licensed under sometimes conflicting norms.

Paper long abstract:

Because of its advent in times of entrepreneurship science incentives, because of its proximity to the pharmaceutical industry, computational chemistry has always been dealing with academic norms as well as business norms. The community of computational chemists emerged in times of a plural hardware ecosystem. They blossomed in between two epochs. They were no longer in the age of supercomputers, when they were remote and dependent on the attribution of calculation time provided by computing centers. They were not yet in the age of ubiquitous desktop computing. In this specific computing context, the circulation of software has proved pivotal for practitioners.

It is epistemically important to describe how software packages were and are distributed, maintained and licensed under sometimes conflicting academic and business norms. Computational chemistry is thus a field with peculiarities regarding software which highlight computing dimensions that have been understudied until now. In media such as newsletters, mailing lists or journals’ op-eds, computational chemists passionately debate about issues of transparency, openness, dissemination... in short software issues, and the tensions they entail. For example, some scientists advocate for open-source software as a necessary condition for sound science, others defend proprietary packages as a warrant for reliable scientific software as a scientific tool. A lot lie in between, trying to reconcile academic norms and a sustainable business model. These heated debates help unveiling tensions about software in science that are otherwise invisibilized.