An effort has begun to collect the source code to as much software as possible, with the aim of building the canonical reference database of software.
Dubbed the Software Heritage Project, and created by the French national science and technology institute INRIA, efforts towards collating source code for reference and posterity began last year. 'We decided to start working on Software Heritage more than a year ago, and we have now shown its feasibility,
' explained INRIA chief executive Antoine Petit of the project's history. 'In order to scale up worldwide, the time has now come to open it up to the widest, national and international contributions.
Software Heritage already has two major partners in the bag: DANS, a subsection of the Netherlands Organisation for Scientific Research, and Microsoft. 'Microsoft has been involved in open source initiatives by enabling, integrating, releasing and contributing to many open source projects and communities for well over a decade,
' explained Microsoft's Jean Paoli of his company's involvement. 'We applaud Software Heritage as an open project that will help curate and conserve human knowledge in the form of code for future generations as well as help today’s generations of developers find and re-use code worldwide. We are proud to be one of the first industry partners for this initiative and to provide the Azure infrastructure to ensure the data is highly available.
'Software Sustainability is one of the core elements to accelerate discovery. Reproducibility of scientific results, but also re-usability and findability of software are core principles in science DANS seeks to support,” added DANS director Peter Doorn. '[i]We are committed to long-time preservation of scientific information and are delighted to be the first international academic partner to support the mission of Software Heritage.
Software Heritage is far from the first such effort to save software for posterity - The Internet Archive
has a software sub-section and the Software Preservation Society
was born out of the Classic Amiga Preservation Society, to name but two - but its concentration on source code, rather than precompiled binaries, potentially gives its data a longevity others would struggle to match.
More information on the project is available on the official website