Bug 200848

Summary: Automatically set 'timeouts' based on maintainer-{approval,feedback} flags + time delay
Product: Services Reporter: Yuri Victorovich <yuri>
Component: Bug TrackerAssignee: Oleksandr Tymoshenko <gonzo>
Status: In Progress ---    
Severity: Affects Some People CC: brad, bugmeister, gonzo, koobs, ygy
Priority: --- Keywords: feature
Version: unspecified   
Hardware: Any   
OS: Any   
Description Flags
check_maintainer_feedback_timeout.pl none

Description Yuri Victorovich freebsd_committer 2015-06-14 07:13:45 UTC
Bug reports which are unanswered by maintainers stay this way for a very long time. Could you add "Maintainer Timeout" flag on bugs? Ideally Bugzilla should automatically set this flag after 30 days.

Here are few examples:

Ideally, committers should work off the combined report including all port updates/new ports/maintainer timeout bugs, and they should work in chronological order.
Comment 1 Kubilay Kocak freebsd_committer freebsd_triage 2015-06-14 09:50:59 UTC
This feature has been (and is) planned, leveraging any existing flags, as flags are exactly about request/response cycles, that all warrant 'timeout' actions being taken.

We'll continue to track this here, thank you for your report
Comment 2 Kubilay Kocak freebsd_committer freebsd_triage 2015-12-30 10:22:49 UTC
Lower priority until we get Stage 1 (low hanging fruit) issues sorted out
Comment 3 Guangyuan Yang freebsd_committer 2017-10-17 05:25:12 UTC
Got here through randomly browsing. Wondering if this can be done now.
Comment 4 Oleksandr Tymoshenko freebsd_committer freebsd_triage 2018-02-16 20:10:54 UTC
Created attachment 190698 [details]
Comment 5 Oleksandr Tymoshenko freebsd_committer freebsd_triage 2018-02-16 20:17:30 UTC
This functionality can be implemented but I need more details first because I am not quite familiar with ports workflow. Also there were some changes since original request.

- Is it still required? I can see named search for Maintainer timeout port reports on main page.

- maintainer-approval "-" value means "rejected", there is no "timeout" state. Should this flag be left out of scope of this change?

- There may be several maintainer-feedback requests. Should script wait for the most recent one to lapse before setting single maintainer-feedback timeout flag or should it replace them with "timeout" flags one by one?

I attached tentative implementation of automation script just in case I drop the ball