Skip to content
Ani edited this page Jan 16, 2024 · 11 revisions

Given that there is no hardline triage procedure, it would be beneficial to at least have an organized form of categorizing and prioritizing issues/PRs for data.table, or in general managing both incoming (primarily) and existing ones for this GitHub repository.

Related roles

A separate role (namely, 'Triage Manager') would be assigned to one who actively handles the triage management aspects of data.table. Responsibilities that fall under this specialized role can include, but are not restricted to:

  • Tagging new issues appropriately for ease of filtering.
  • Prioritizing PRs and issues to be worked on with a focus on user needs, breaking changes, and shared concerns amongst the developers and active contributors.
  • Cherry-picking features and fixes to include for the release next in line.
  • Assisting in issue contributions from new users (such as reducing the length of a supplied reproducible example or making it minimal where possible; could even be a simpler case like suggesting to show a reprex if no such examples are provided).
  • Re-running the code supplied in regression or bug-oriented issues and commenting on the results achieved (could lead to potentially useful information, but even if not it helps to validate that the issue indeed exists and is not specific to the author).
  • Being actively observant of PRs and issues, with or without engagement. This helps to reference former related issues (these could be based on some other R-package or software used in the toolchain and not specific to data.table, but to keep track of this repository's issue tracker is what I'm referring to) for more context or helpful information, and is seen done frequently by ones that are already in the habit.