Skip to content

Latest commit

 

History

History
74 lines (54 loc) · 4.66 KB

triage.md

File metadata and controls

74 lines (54 loc) · 4.66 KB

Introduction

This document will outline the process of triaging GitHub issues and provide an explanation of GitHub labels.

GitHub Labels

Labels that start with a: are area labels. Area labels help categorize issues based on their scope. Each issue should ideally have at least one area label. The description of the area label ends with a code that indicates the maintainer or backup maintainer of that area. For example, if a label description ends with (mX,bY), it means person X is the maintainer, and person Y is the backup maintainer for that area.

Apart from area labels, there can be other labels used to provide additional context to issues.

GitHub Assignee

Here are some guidelines for GitHub assignees:

  • Assign issues to individuals who are actively working on them or going to actively work on them in the near future.
  • Avoid assigning too many issues to a single person to prevent overload.
  • Assign an issue to yourself if you plan to work on it to inform others that the issue is being addressed.
  • Assign an issue to someone else if you expect a quick action or if someone else is better suited to handle it.
  • Unassign issues that you are not actively working on to allow others to pick them up.

Issue Description

A clear and concise issue description is crucial for effective communication.

  • Edit the title and description of the issue to ensure clarity and conciseness.
  • If modifications have been made to the original issue, indicate it by adding "Edited:" to the description.
  • Consider adding "acceptance criterial" so that it's easy to understand what needs to be accomplished in order to close the issue as done.

Triage Process

  1. The first person who identifies an issue can assign labels to it. Make sure to include at least one area label. Add the Need Triaging label to the issue if triaging is not complete, so that the individual responsible for the area can continue with the process.
  2. Investigate the issue by trying to reproduce and analyze it. (Is it for a specific platform? what part of the code is involved? Narrowing down to the root cause)
  3. If the issue can be done in less than one hour, do it now. (Unless it's a low importance bug or some feature that would be a good candidate for a good-first-issue)
  4. Otherwise, evaluate if the area is correct, if this is a bug or an enhancement, set the right tag, maybe a priority. If needed edit the description and title to make it more clear.
  5. Answer the issue by thanking the reporter for reporting the issue. Asking relevant question if any. The comment can add context and information from the analyze step so that the next person looking at the issue can understand it better.

Issues should be responded to reasonably fast. Ideally, within one business day. Remember, effective communication and prompt action are essential for successful issue triaging.

If an issue seems like a good candidate for a good-first-issue, add the good first issue label to it and add instructions in a comment on how to do it, ideally with link to the code.

Filters

Use the following scripts to generate the filters.

Filter for all the issues not assigned to an area:

curl -H 'Accept: application/vnd.github.v3+json' "https://api.github.com/repos/slint-ui/slint/labels?per_page=100&page=1" | jq -r '.[].name'  | grep "^a:" | sed 's/^\(.*\)$/-label:\\\"\1\\\"/' | xargs echo

Filter of all issues for which X is a maintainer (replace the X in "mX" with the right letter name, or bX for the backup)

curl -H 'Accept: application/vnd.github.v3+json' "https://api.github.com/repos/slint-ui/slint/labels?per_page=100&page=1" | jq -r '.[] | select(.description | contains("mX")) | .name' | awk '{printf "\"%s\",", $0}' | sed 's/^\(.*\),$/label:\1\n/'