Course Repository Playbook

Course Repository Playbook

For conveners and tutors running a course in the organisation. Teaching material is one of the organisation's main uses, and it carries one risk above all others: student personal data must never enter a repository. This page covers how to run a course repository safely.

Setting up

• Create a team named with the course code, for example EEE2046F, and add the teaching staff and tutors. See Setting up a Team.
• Create the course repository and give the team access. Grant tutors Write, or Maintain if they manage issues and releases but should not change settings.
• Keep memoranda, unreleased assessments and solutions in a private repository, with access limited to the teaching team.

Adding a cohort

• If students need direct access, add the cohort to the course team for the duration of the course only.
• Remove them once results are finalised, or archive the repository, so that access does not linger past the course.

What may and may not go in

Marking and analysis tools are welcome. Course notes, tutorial material, LaTeX sources and problem sets all belong here.

The following must never be committed, in any repository, at any point in its history:

• Student names, student numbers and EMPL IDs.
• Marks, grades and feedback tied to a student.
• Student submissions.

Keep all of that on UCT-controlled storage, such as OneDrive or Amathuba. This is both a privacy obligation and a consequence of GitHub being hosted outside South Africa.

At the end of the course

• Finalise results and remove student access.
• Archive the repository if the course is being retired or substantially restructured, so that the active version is never in doubt.
• Carry improvements to notes and tutorials forward into the next offering, rather than leaving several live copies.

See also the Contributing guide.