Skip to content

sync/bib.yml: re-seal intermediate's six take rulings - #90

Draft
mmcky wants to merge 1 commit into
mainfrom
bib/reseal-intermediate
Draft

mmcky wants to merge 1 commit into
mainfrom
bib/reseal-intermediate

Conversation

@mmcky

@mmcky mmcky commented Nov 22, 2026

Copy link
Copy Markdown
Contributor

What

QuantEcon/lecture-python.myst#1140 sets intermediate's copy of every entry a ruling covers to the pool's result (the owner's "All ruled entries (Recommended)", 2026-11-22). Six rulings take intermediate's copy and change it, so their seals must name the new copy. Otherwise tools/bib-merge reopens all six once intermediate's pin moves, since the copy a ruling takes must be its sealed version, and the bibliography check goes red.

Ruling intermediate's seal before after
AKR1990 a229f98efbe3e257 0517d1216224e38a
hansen2020twisted 3521470dba8bdbc0 12c2453dc0203a9f
HansenRichard1987 3bbc65ca3cdefc33 dff11e9d6d462585
Helpman1985 cf417eed8a3dacef 26992dc65b708457
MarcetMarshall1992 5bc8e38b782615e6 1340b06a614cbba3
pareto1896cours 3af6b01cc3c7cc6c 04762a3685a82856

The digests are tools/bib-merge's own, of intermediate's copies at QuantEcon/lecture-python.myst#1140's head, 148a901. Each ruling's because notes the re-seal. The merged entries do not change. Every other ruling that takes intermediate's copy still matches its seal, so no other one reopens.

Checked

A simulation with tools/bib-merge's own code:

Order (draft until it can pass)

This cannot pass until intermediate's pin names QuantEcon/lecture-python.myst#1140's merge, because its CI reconstructs intermediate's bibliography at the pin in sync/state.yml. It also waits for #88, which changes the same header. So:

  1. sync/bib.yml: re-seal intro's two take rulings for its converged copies #88 merges (after bibliography: set the 42 ruled entries to the lecture monorepo's lecture-python-intro#875);
  2. bibliography: set the 23 ruled entries to the lecture monorepo's lecture-python.myst#1140 merges;
  3. this branch is rebased on main; tools/sync --update --pin-only intermediate moves intermediate's pin to #1140's merge (it writes only sync/state.yml), and tools/bib-merge --reconstruct --write regenerates lectures/_static/quant-econ.bib, whose header must then read intermediate=06466b7d241875b4;
  4. its CI goes green, including the four-site build that a change under lectures/ runs, and it merges before the next nightly sync (03:47 UTC), best straight after #1140. A nightly sync between the two would move intermediate's pin on main by itself, and the bibliography check would go red with these six refusals until this merges.

The pin covers every change merged to intermediate's main by then, not only #1140. Moving it changes no lecture in the pool here: drift-check reports any promoted lecture those changes touch as a refresh, and its refresh pull request brings them, reviewed as usual. At step 3 this body lists the upstream changes the new pin covers. The dates in the because notes are those of this draft; they are corrected at step 3 if #1140 merges on a later day (UTC).

This and #89 can go in either order after #88; whichever goes second is rebased on the first.

🤖 Generated with Claude Code

https://claude.ai/code/session_01WT8bMacV9Lmy1VLUc6XJoL

QuantEcon/lecture-python.myst#1140 sets intermediate's copy of every entry
a ruling covers to the pool's result. Six rulings take intermediate's copy
and change it, so their seals now name the new copy (digests from
tools/bib-merge's own code, of intermediate's copies at 148a901). The
merged entries do not change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WT8bMacV9Lmy1VLUc6XJoL
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant