Software sustainability != project sustainability

At OSS Watch, we take the quality of our documents very seriously. When we publish documents, we don’t just leave them alone on the website, but we nurture them, reviewing them every six months. This ensures that every one of our published briefing notes remains up-to-date and relevant.

Sometimes, reviewing a document means that we make some changes to reflect current thinking. This happened recently, when we were reviewing our document on advice for project bids. In that document, we discuss the sustainability aspects of funded software projects. The focus was on ‘project sustainability’ but reading that back it felt like an ambiguous term.

Our experience in working with projects like the DataFlow has confirmed that sustainability of the project, where this refers to the project as funded by eg. JISC or a research council, is related to but not the same as the sustainalibity of the software that is being developed as part of the project.

DataFlow, for example, creates a two-stage data management infrastructure that “makes it easy for you and your research group to manage your research data.” Two separate pieces of software, DataStage and DataBank, are being developed as part of the project. From our perspective, the value is in the software that is being developed and the communities that can be created around these communities to collaboratively develop the software further using the open development methodology. It is likely that DataFlow as a project will end. But the value of the software will remain and the ongoing sustainability will be through the software that has been developed. The software, and its associated community, documentation, etc., will be what can attract people from outside the funded project to become interested and involved. Of course, the software and development around it is a project in its own right, but to prevent ambiguity we now distinguish explicitly between software sustainability (ie. sustainability of the software being developed) and project sustainability (where this regards the project that funds the initial software development).

Sometimes it can even be just a small part of the software that gets picked up and becomes sustainability. If you read our Wookie case study, you’ll see that Apache Wookie (Incubating) originally was only part of the software that had been developed in the TenCompetence project. TenCompetence itself does not show a lot of activity, whereas the Wookie project is on its way to graduate from the ASF’s Incubator to become a top-level Apache project.

Our work with the DataFlow project is continuing, and we help organise a workshop in Oxford to which we invite everyone who is interested in the solutions that are being developed in both the DataFlow and VIDaaS project. Register now if you are interested!

One thought on “Software sustainability != project sustainability

  1. Pingback: What I’ve starred this month: February 28, 2012 – MASHe

Comments are closed.