You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There's a substantial backlog of obsolete or stale issues and we could use a heuristic to close a significant number ("issue bankruptcy"). The policy would initially cause a large number to be closed, then stay in place periodically closing issues that meet the criteria. Several major repos including non Microsoft repos have such a policy.
Experiment
We will be trialing an issue bankruptcy experiment in dotnet/runtime focused on area-System.Collections and area-System.Text.Json. There are a number of parameters to be considered for the experiment:
Selection Criteria
Focus on Future milestone issues that have not been updated in over 6 months.
Exclude popular issues, e.g. issues that contain a large number of responses or upvotes.
Possibly exclude issues where the ball is clearly in our court (although that is probably vague and might be impossible to express in a query).
Execution
FabricBot applies a no recent activity label to selected issues and writes a post explaining that the issue has been marked for closure, linking to documentation related to this experiment, potentially encouraging users to submit their feedback on the process.
If no response has been provided within 14 days, the issue will be closed automatically by FabricBot. Otherwise, FabricBot will remove the no recent activity label from the issue.
Timing
Ideally we would want this experiment to coincide with the planning phase of a next release.
Success criteria
Reduction in the number of open issues in the Future milestone.
The text was updated successfully, but these errors were encountered:
Suggest to post a sticky issue first in the repo to explain the experiment, what we're going to do and why and encourage folks to help us do it right and learn from it -- and give the community some time (eg., a week) to share their thoughts.
Note that in dotnet/runtime, closed issues currently can be reopened by the original author within 30 days of the last activity -- after that they're locked. We can consider adjusting that automation of course.
Hypothesis
There's a substantial backlog of obsolete or stale issues and we could use a heuristic to close a significant number ("issue bankruptcy"). The policy would initially cause a large number to be closed, then stay in place periodically closing issues that meet the criteria. Several major repos including non Microsoft repos have such a policy.
Experiment
We will be trialing an issue bankruptcy experiment in dotnet/runtime focused on
area-System.Collections
andarea-System.Text.Json
. There are a number of parameters to be considered for the experiment:Selection Criteria
Future
milestone issues that have not been updated in over 6 months.Execution
no recent activity
label to selected issues and writes a post explaining that the issue has been marked for closure, linking to documentation related to this experiment, potentially encouraging users to submit their feedback on the process.no recent activity
label from the issue.Timing
Ideally we would want this experiment to coincide with the planning phase of a next release.
Success criteria
Reduction in the number of open issues in the
Future
milestone.The text was updated successfully, but these errors were encountered: