Skip to content

Instantly share code, notes, and snippets.

@MichaelDimmitt
Last active April 3, 2024 17:53
Show Gist options
  • Save MichaelDimmitt/2597ac2f50f7f43ef414dd83b7cf1ff9 to your computer and use it in GitHub Desktop.
Save MichaelDimmitt/2597ac2f50f7f43ef414dd83b7cf1ff9 to your computer and use it in GitHub Desktop.
my thoughts: guidelines for code reviews

My Thoughts: Guidelines for Code Reviews

Does anyone have a document they follow for having guidelines reviewing code within a team?

I read through this document today and it was helpful for this task:
https://github.com/mawrkus/pull-request-review-guide?tab=readme-ov-file

Different ways developers can review a pr

  • sanity check (confirm it looks good and logs are on the pr, tests are written or a checklist followed)
    • devs should check this themselves but it sometimes slips.
  • reviewing a/c's
    • devs should check this themselves but it sometimes slips.
  • an extensive manual test
    • usually a developer asks for another pair of eyes on a manual regression test.
    • comment with approvals a code suggestion
    • hold the line, if someone is learning a new pattern and deviating significantly from that pattern.
    • other ways to review ...
  • the point is different pr's should be evaluated differently
    • the goal is to help inform the person making the pr
      • the person who made the pr is responsible for the change.
      • usually reviewing a pr you are there to assist and help them be informed.

If you do not know how to review a pr. Ask the person who submitted it, how they want it to be reviewed.

When should a pr be declined or reassigned?

  • What is your current measure of when a pr should not be merged?
    (My measure: If there is a known breaking change)
  • What is your current measure of when a pr should be reviewed by someone else?
    (My measure: If a significant change is presented on something someone else worked on like water system group)
  • What is your current measure of when a pr should be reviewed by only the lead developer?
    (My measure: rarely, maybe only assign the lead to evaluate a huge refactor)

More considerations regarding pr's:

  1. how many users are impacted by your codebase in general.
  2. how many users are impacted by your change?
  3. how far away are you from a release?
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment