Scientific software has become an indispensable commodity for the production, processing and analysis of
empirical data but also for modelling and simulation of complex processes. Software has a significant influence
on the quality of research results. For strengthening the recognition of the academic performance of scientific
software development, for increasing its visibility and for promoting the reproducibility of research results,
concepts for the publication of scientific software have to be developed, tested, evaluated, and then transferred into
operations. For this, the publication and citability of scientific software have to fulfil scientific criteria by means
of defined processes and the use of persistent identifiers, similar to data publications.
The SciForge project is addressing these challenges. Based on interviews a blueprint for a scientific software
publishing platform and a systematic implementation plan has been designed. In addition, the potential
of journals, software repositories and persistent identifiers have been evaluated to improve the publication and
dissemination of reusable software solutions. It is important that procedures for publishing software as well as
methods and tools for software engineering are reflected in the architecture of the platform, in order to improve the
quality of the software and the results of research. In addition, it is necessary to work continuously on improving
specific conditions that promote the adoption and sustainable utilization of scientific software publications. Among
others, this would include policies for the development and publication of scientific software in the institutions but
also policies for establishing the necessary competencies and skills of scientists and IT personnel.
To implement the concepts developed in SciForge a combined bottom-up / top-down approach is considered
that will be implemented in parallel in different scientific domains, e.g. in earth sciences, climate research and
the life sciences. Based on the developed blueprints a scientific software publishing platform will be iteratively
implemented, tested, and evaluated. Thus the platform should be developed continuously on the basis of gained
experiences and results. The platform services will be extended one by one corresponding to the requirements of
the communities. Thus the implemented platform for the publication of scientific software can be improved and
stabilized incrementally as a tool with software, science, publishing, and user oriented features.
EPIC Alfred Wegener Institut