1 October 2020 Standard for Public Code community call

Update from the Foundation

• We gave several conference talks recently:
  • State of the Source 2020: How Open Source in Public Administrations and Government is different by Boris
  • Comparing ecosystems: MariaDB and Public Code - MariaDB Server Fest 2020 by Eric
  • Open Sourcing our Way to the SDGs, a panel Ben was in.
• OpenZaak had their first Product steering meeting and Technical steering meeting.

Attendees

• Jan Ainali (Chair)
• Elena Findley-de Regt
• Eric Herman
• Laura Scheske

Notes

Distinguish the requirements of the “Create reusable and portable code” criterion

This issue arose during a Standard gap analysis where it was unclear what the difference between the terms portable and reusable was. They are intended to have a different focus so new wording was proposed to make it more clear. We all liked this proposed change, but we’ll leave it open for a few more weeks in case anyone else has a comment.

Add a requirement to consider the ethics of your software

We’ve decided to apply to have the Standard recognized as a Digital Public Good. Being a Digital Public Good comes with ‘do no harm’ requirements, so we made some proposals for changes to the Standard to ensure that ethics are not forgotten. We believe that no new requirements are necessary, since public organizations usually have stringent ethical principles and policies, which can be referred to. We adjusted the wording of the proposal slightly.

We also acknowledged that new suggestions for ‘further reading’ are welcome but intentionally left them out of this pull request to keep it small and easy to move forward.

General discussion about public code

Following this, we talked a little bit about our definition of public code and how policies should be referenced in codebases for us to consider them as such. This came up in light of the policy we just discussed and the criterion Bundle policy and source code.

The discussion was mostly philosophical, and it became clear that codebases are diverse: for some it is very obvious that a specific policy is being implemented, and for others it is equally obvious that none exist. The interesting parts are all those cases in between where either:

• it is plausible that a codebase implements policy but it is not being stated explicitly
• a generic codebase is used as a tool to execute policy, but the policy is not implemented in the codebase itself

This discussion was enlightening but not conclusive, so we need to continue it in the future.