diff options
| -rw-r--r-- | CLAUDE.md | 25 |
1 files changed, 25 insertions, 0 deletions
@@ -135,6 +135,31 @@ against missalemeum; layer 5 pins ~30 dates. Litanies, the RG 110 Peter companion, St Barbara). Not wired in as a layer — its PDF columns shift between pages and extraction is unreliable; a structured Ordo would be the only way to add a fourth *lineage*. +- **THE CHAIN IS NOW BROKEN AT ONE LINK, AND LAYER 4 IS THE OUTLIER** + (2026-08-18). lectio has ADOPTED eight corrections colitur argued from the + Missal — `ubaldus`/`didacus` promoted to III-class feasts, both August + vigils recoloured violet under RG 128, and both Ember Saturdays' readings. + So on those days the lineage no longer runs Divinum Officium → lectio: + lectio now follows the Missal, and **missalemeum is the only party still + carrying the inherited error.** Layer 3's fixture was regenerated, and + **C27, C28 and C37 went to zero and are closed** (C38 narrowed to a + readings-only divergence, which cannot close without lectio gaining a + Commons concept). 227 of 16 801 rows changed; agreement 86.6% → 87.5%. + Read the percentage as the *least* interesting part: those 227 days now + agree BECAUSE BOTH ENGINES FOLLOW THE MISSAL, where before they agreed + because both inherited the same errors. This is the first time the project + has RESOLVED the lineage problem on a set of days rather than documenting + it. + **What this changes for layer 4, practically.** On those days the oracle + now disagrees with BOTH engines, and its allow-list entries (M29, M30, and + the Ember-reading classes) should be read as "missalemeum is the outlier", + not as "colitur is unusual". Do not treat a fresh colitur-vs-missalemeum + divergence as presumptively colitur's fault: the base rate has changed. + **What it does NOT change**: everywhere else, layers 3 and 4 remain one + lineage, and lectio's own data is still generated from missalemeum by + `scripts/gen-sanctoral-ef.go`. That generator would silently revert all + four calendar corrections if re-run, which is why lectio's ini headers and + the generator's own doc comment now each carry the list. - Layer 3 **never compares commemorations** (lectio has no RG 111 admission logic, so its "others" are losing candidates, not the admitted set) and never compares the week column. It also **shares colitur's own lineage**: colitur's data was |