Lucy

Lucy is a python application for managing student assignments delivered through GitLab. Lucy includes many time-saving and useful features for course convenors and tutors.

Lucy has been developed actively by ANU academics and tutors since 2017 (> 500 commits) and applied in both small and large courses.

This site includes:

• Documentation for installing, setting up, and using Lucy for course convenors and tutors.
• Explanation of the “Lucy way of doing things”.
• Explanation of Lucy’s features, and development roadmap.

Go have a look: https://gitlab.comp.anu.edu.au/lucy/lucy

• Integrates with Canvas
  • Sync student details and lab enrolments from Canvas with one command (or import a MyTimetable export)
  • Upload marks and comments straight to the Canvas gradebook
• Never touch student repositories
  • Marking happens in a private marking fork owned by the course’s marker user
  • Easily check completion/validity of tutor marking
  • Feedback sent back to students as a branch of their repo
• Flexible marking workflows
  • Allocate submissions to tutors (own lab, other labs, random, weighted)
  • Tutors can mark locally with lucy, or entirely in the GitLab web UI with a published progress report
  • Pre-marking moderation round so tutors calibrate before they mark
  • Collated marks with summary statistics
• Deadline accuracy and flexibility
  • Mark content of latest push before deadline
  • Track student progress (forks, pushes, CI jobs)
  • Easy to add extensions
• Forum Integration
  • List students in correct format to add them to Ed
• Academic Integrity Tools
  • Consistent reporting
  • MOSS Integration
• Course plugins
  • Add course-specific commands in your course’s plugins/ directory

Lucy is a Python CLI app distributed through GitLab. Here’s a checklist for setting it up for your course. There’s more details on the installation page, but here is the short version:

1. Fork the Lucy repository for your course (e.g., https://gitlab.comp.anu.edu.au/comp1720/2021/lucy )
2. Clone your copy of Lucy
3. Configure Lucy by copying data/templates/course_info.yml to data/course_info.yml and filling it out
   1. Course details (Canvas course ID, GitLab marker username, tutor group)
   2. Assessment task details (deadlines, number of marks, marking files)
   3. Tutor details (names and UIDs, which labs)
   4. Lab group details (activity and times)
   5. MOSS settings (id, language)
4. Put the marker user’s GitLab token and your Canvas token in data/secrets.yml, starting from data/templates/secrets.yml
5. Install Lucy’s python environment with poetry (poetry install)
6. Run Lucy: poetry run ./lucy.py --help
7. Pull the class list from Canvas: poetry run ./lucy.py canvas-sync

There are many approaches to automatically managing computing student assignments. Why choose Lucy?

1. It’s useful, reliable, and adaptable.
2. It has many eyes checking code and many brains available to fix problems.
3. When you make Lucy better, others benefit.

Vision: Make Lucy so good that it is used in all core computing courses.

• Group Projects
  • So far doesn’t support group projects Needed for core subjects (e.g., COMP1110)
• Improve Docs
  • Create an easily navigated docs website/wiki Need quick reference and deeper documentation
• Core Features: Lab Reports
  • Lab Reports applied in COMP2300/6300 and COMP1720/6720 Need to upgrade to be a core feature
• Student Communication and Forums
  • Can this interact with other typical communication channels? (Canvas, Email, Ed)
• Fit in with ANU World
  • Harmonising Lucy with Canvas
  • Developing helpful tools to help with Academic Integrity

Lead Contact: Charles Martin

BDFL: Ben Swift

Devs:

• Julian Crosby (current contributor)
• Ashleigh Johannes (current contributor)
• Brent Schuetze (lifetime achievement award)

Hall of Fame:

• Harrison Shoebridge
• Peter Davis
• John Zhang
• Kieran Browne
• Zixian Cai