Retire the old reporter's News output

[mmcky]

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.yml runs daily at 07:25 UTC and opens a draft the morning after any library release;
  • news-weekly.yml runs 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 Its 14 drafts are superseded: 13 by the migrated Activity data, and News drafts — lectures, website (2026-07-12 → 2026-07-19) #240 because none of its items meets the Activity definition.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-activity
  • gh 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.

• 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.

[mmcky]

Done on 2026-09-29, after #265 merged.

• Reporter paused. news-watch.yml and news-weekly.yml in QuantEcon/reports-activity are disabled, and both read back as disabled_manually. One leftover: a news-weekly run 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:

  • the 13 superseded drafts have every qualifying item there;
  • News drafts — lectures, website (2026-07-12 → 2026-07-19) #240 has no item that qualifies.

  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:

• Releases are covered through 2026-09-27, by the daily run of 2026-09-28.
• Lectures are covered through 2026-09-27, by the weekly run of that day.
• The replay starts at 2026-09-28.
• A lecture gap. The weekly run due on 2026-09-13 never started, so lecture updates from 2026-09-07 to 2026-09-12 were never reported. Include that week if the backfill reaches back that far.
• A file-name clash to avoid. _data/activity/2026-09-28-software.yml already 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.