Skip to content

Latest commit

 

History

History
75 lines (44 loc) · 2.5 KB

ISSUE-EOL-POLICY.md

File metadata and controls

75 lines (44 loc) · 2.5 KB

End-of-Life Policy on Issues

jetsam (n): the part of a ship, its equipment, or its cargo that is cast overboard to lighten the load in time of distress

Finite resources demand some level of pruning. This document describes the basic principles used by the Containers team to identify and close stale issues.


Definitions

stale: open, but no activity in the last thirty days.

crickets: closed due to lack of response from reporting party.

jetsam: closed without being implemented. A deliberate decision made in recognition of human limitations.


Process

In order of judgment, from least to most.

→ stale

Issues are marked with the label stale-issue by a github action that runs daily at 00:00 UT. This also triggers an email alert to subscribers on that issue.

Judgment: typically a team member will skim the issue, then decide whether to:

  • remove the label; or
  • close the issue (see below); or
  • do nothing.

This is informal: there is no guarantee that anyone will actually do this.

→ crickets

Typically done by a team member after receiving a stale-issue email.

Judgment:

  • there is not enough information to act on the issue; and
  • someone on the team has asked the reporter for more details (like NEEDINFO); and
  • the reporter has not responded.

There is no actual crickets label. There is no automated way to find issues that have been closed for this reason.

→ jetsam

Last-resort closing of an issue that will not be worked on.

Factors:

  • issue has remained open for over sixty days; and
  • reporter is responsive, and still wishes to have the issue addressed (as does the team).

Judgment:

  • the issue is too difficult or complicated or hard to track down.
  • decision should be made by two or more team members, with discussion in the issue/PR itself.

When such an issue is closed, team member should apply the jetsam label.


Grayer Areas

stalebot isn't perfect. It often misses issues, and we end up with some that have been open a long time and do not have the stale-issue label.

These are hard to find manually. There is no defined process for identifying or acting on them. If and when someone finds these, they should be handled through the process defined above.