Repository navigation
Retire the old reporter's News output #274
Description
Part of #271 (Phase 1).
Why. Once the split is merged, two things are left over:
-
The old workflows. The reporter in QuantEcon/reports-activity has two workflows that would keep opening News drafts against
_posts/:news-watch.ymlruns daily at 07:25 UTC and opens a draft the morning after any library release;news-weekly.ymlruns Sundays at 08:55 UTC.
Both were active on 2026-09-29.
-
The drafts. Its 14 drafts are superseded: 13 by the migrated Activity data, and News drafts — lectures, website (2026-07-12 → 2026-07-19) #240 because none of its items meets the Activity definition.
Work.
-
Disable both workflows:
gh workflow disable news-watch.yml -R QuantEcon/reports-activitygh workflow disable news-weekly.yml -R QuantEcon/reports-activity
QuantEcon/reports-activity#91 replaces them with one daily Activity workflow, and replays the paused days.
-
Close the 14 drafts unmerged:
- News drafts — software releases 2026-07-20 #241, News drafts — lectures (2026-07-19 → 2026-07-26) #242, News drafts — software releases 2026-07-26 #243, News drafts — lectures (2026-07-26 → 2026-08-02) #246, News drafts — software releases 2026-08-02 #247, News drafts — lectures (2026-08-02 → 2026-08-09) #251, News drafts — software releases 2026-08-12 #252, News drafts — lectures (2026-08-16 → 2026-08-23) #255, News drafts — software releases 2026-09-14 #258, News drafts — software releases 2026-09-16 #259, News drafts — software releases 2026-09-23 #261, News drafts — software releases 2026-09-24 #262 and News drafts — lectures (2026-09-20 → 2026-09-27) #263, with a note that their content is on
/activity/; - News drafts — lectures, website (2026-07-12 → 2026-07-19) #240, with a note that none of its items qualifies for Activity.
- News drafts — software releases 2026-07-20 #241, News drafts — lectures (2026-07-19 → 2026-07-26) #242, News drafts — software releases 2026-07-26 #243, News drafts — lectures (2026-07-26 → 2026-08-02) #246, News drafts — software releases 2026-08-02 #247, News drafts — lectures (2026-08-02 → 2026-08-09) #251, News drafts — software releases 2026-08-12 #252, News drafts — lectures (2026-08-16 → 2026-08-23) #255, News drafts — software releases 2026-09-14 #258, News drafts — software releases 2026-09-16 #259, News drafts — software releases 2026-09-23 #261, News drafts — software releases 2026-09-24 #262 and News drafts — lectures (2026-09-20 → 2026-09-27) #263, with a note that their content is on
-
Mark NEWS: move to a single weekly news post to reduce noise + only report changes to lectures and major software (QuantEcon.py, mystmd etc.) #225, which raised the problem, as completed, with a pointer to Split updates into News and an Activity feed #265 and this project.
Acceptance criteria
- Both workflows show as disabled.
- The 14 drafts are closed, and no new News draft has been opened.
- NEWS: move to a single weekly news post to reduce noise + only report changes to lectures and major software (QuantEcon.py, mystmd etc.) #225 is closed with a pointer.
Activity
mmcky commentedon Sep 29, 2026
Done on 2026-09-29, after #265 merged.
Reporter paused.
news-watch.ymlandnews-weekly.ymlin QuantEcon/reports-activity are disabled, and both read back asdisabled_manually. One leftover: anews-weeklyrun from 2026-09-13 has shown as queued, with no jobs, since that day, and a normal cancel reports it as already completed.Drafts closed without merging. Before closing, each draft was checked against the Activity data on
main:Items the definitions exclude (translation-sync tooling, analytics, editor-of-record metadata) were left out, and each closing note says so.
NEWS: move to a single weekly news post to reduce noise + only report changes to lectures and major software (QuantEcon.py, mystmd etc.) #225 closed as completed, with a pointer to Split updates into News and an Activity feed #265 and News and Activity: announcements on News and a self-updating Activity feed #271.
No new News draft. No reporter PR has been opened since the merge.
Where the old reporter stopped. It keeps no state, so this is the starting point for the replay in QuantEcon/reports-activity#91:
_data/activity/2026-09-28-software.ymlalready exists: it is the hand-added backfill of five releases the reporter missed. A replay that names files by the day covered would write the same name for 2026-09-28, so it must not overwrite that file.