Page Comparison

...

Last sprints (S56) retro is here: 2023-02-08 (S56) Retrospective

Retrospective

...

S57 Discussion topics:

• New thoughts on Draft PRs?

  • Thoughts from last retro: Joyce Sato-Reinhold attempted to do a draft PR this sprint as a way to get feedback from the team on an approach before getting too deep into a particular solution. In the end, the approach she thought would work didn’t so she ended up coding two solutions that both look like they’ll work. There are other solutions that she thinks will work, too. Concerned it may not be super readable since it’s very commented. Structuring it was pretty difficult too. Some things to consider:

    • Does this help your personal process?

    • Is the ticket impactful enough to code out several approaches?

  • Joyce did all of the different approaches in the same branch and thought this was easier. Could also potentially do them as different branches.

  • Sayaka always creates Draft PRs first, reviews her own code, and then switches it to ready when she’s happy with the solution.

  Feedback on Mwaya’s unit testing guidance: Jest unit test guidance

  • took a different approach this sprint, she:

    • Only did 1 approach instead of multiple and very thoroughly documented her approach in comments on the PR (https://github.com/LiteFarmOrg/LiteFarm/pull/2461 )

    • Only Duncan reviewed the draft PR; one theory is when there are many PRs vying for your attention a draft one becomes de-prioritized

    • A few other approaches we could employ draft PRs for:

      • Sayaka Ono ‘s approach for work that’s “in progress” but not ready for approval

      • Jimmy’s approach for more experimental features that aren’t necessarily on the roadmap

    • Final guidance on using Draft PRs: use them if it helps your process!

• Feedback on Mwaya’s unit testing guidance: Jest unit test guidance

  • Add a section on “what not to test”, what is the threshold to not create a test

  • Would be nice to know how are unit tests part of a more holistic testing strategy?

  • Would be nice to have a seed database with pre-made farms, users, crops, etc. that create a few baseline environments for testing. A few examples include “Farm with wild crops”, “Farm with paid and unpaid users”, etc.

    • Joyce Sato-Reinhold: I'm very fond of this idea because a large volume of the code in each test file involves the setup of the farms, users, etc. Moving this setup into a database seeding script would make the tests much shorter and more readable. But it would also introduce the complexities of keeping the testing files in sync with the seed script and knowing when to use the current approach with the mocks vs. relying on the seed database.

  • Combine into this document / section? https://lite-farm.atlassian.net/wiki/spaces/LITEFARM/pages/1190428675/The+api+package#Backend-tests

• Thoughts on a LiteFarm calendar for tracking OOO, meetings, and events?

  • Q: How would this incorporate For folks with @litefarm.org you can automatically view their calendars on google calendar

  • For folks without @litefarm.org email addresses?you can request to view (“subscribe”) their calendar on google calendar

Action items:

•  Lite Farm to auto-assign Duncan and Sayaka to PRs
•  Lite Farm to give Sayaka Ono merge access