Promoting Science Software Best Practices

In addition to creating digital data, contemporary scientists often develop software/code as part of their research activities. While the code is generally well understood by the creator, it is often not easily understood by others who might use it. In addition, most scientists are not trained software developers and their code may break easily or not be easily linked to or modified to extend its functionality. These challenges are analogous to those faced on the data side of the science enterprise. It is, therefore, a natural fit for ESIP to look at science software best practices as a complement to the work ESIP does related to data.

The goal of this session is to examine some of the dimensions of science software best practices and to gauge interest for creating a cluster devoted to promoting science software best practices.

Some dimensions of science software of potential interest include the following:

• User perspectives

• Software engineering

• Model Coupling

• Community model standards

• Agile methodologies

• Open Source

This cluster will naturally complement other ESIP activities such as data preservation, technology committees, and complement efforts of sponsor organizations in these areas.

The format of the session will consist of short presentations elucidating some of these topics followed by a town-hall or open forum for discussing the formation of a cluster devoted to this topic.