deterministic OCaml engine to compute and validate liturgical calendars for multiple rites, template-driven output to year 9999 https://git.labunix.xyz/colitur.git/atom?h=v0.10.1 2026-08-21T13:29:47Z 2026-08-21T13:29:47Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:29:47Z urn:sha1:4f87cbde4ee79ac96aa7ebfed105b3a5ec02be43 2026-08-21T13:28:13Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:28:13Z urn:sha1:43bea0b314c902e148f0ad9154a2ef05fd459d9a 2026-08-21T13:12:08Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:12:08Z urn:sha1:d129b9b9a766c3530d2dd6ac131ebafd1bddf7a5 The Latin Mass Society Ordo 2024-2025 for England and Wales (v2.21, compiled by Peter Day-Milne) gives 9 August 2025 as "OUR LADY on SATURDAY IV Cl W / V Mass of BVM / Commem of St Romanus M at Low Mass" -- no vigil -- and 10 August as the 9th Sunday after Pentecost with St Laurence commemorated. That is colitur's output on both days, field for field. It is discriminating where the Ordo already in docs/research was not. This edition lists six vigils as the day's office across its year, including the Assumption's on 14 August 2025, and its footnotes discuss arranging a diocesan feast around the Vigil of St Laurence in other years. Its absence on 9 August 2025 is an application of the rule, not a data gap. The extraordinaryform.org Ordo omits that vigil on every date it covers, including 2027-08-09 where the clause says to keep it, which is why it witnesses nothing here. Also confirms the Vigil of the Assumption as II Cl violet, the RG 128 colour this project corrected from the bootstrap's White. The clause's other shape -- feast transferred away, 2022/2033/2038/2044 -- remains unwitnessed; no accessible Ordo covers those years. Stated in the note rather than left implied. 2026-08-21T13:07:44Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:07:44Z urn:sha1:35c11c7ccf95acfb35a7ec34edfa86d92da60e4a The last unbuilt clause in the EF module. A II/III-class vigil is omitted entirely when the feast it precedes fails to keep its own day -- 1 744 days across 1583-9999, three verified shapes, zero unexplained. Rule and vigil inventory independently confirmed by O'Connell 1964, outside the Divinum Officium -> missalemeum -> lectio lineage. The 10 affected days in 2005-2050 are allow-listed as C39; no external day-level witness exists, which the note states rather than glosses. 2026-08-21T13:04:52Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:04:52Z urn:sha1:f3ac69d0a679200a48716be065a0fa2e70a2c6da Replaces the withdrawn Ordo claim with a real one of a different kind. O'Connell, The Celebration of Mass (4th ed., 1964) -- a rubrics commentary contemporary with the 1960 Code, outside the Divinum Officium -> missalemeum -> lectio lineage, already cited here for RG 128 -- states all three of RG 33's triggers in English, the third included, verbatim in substance with the Latin. The same passage confirms vigil_feast_table's membership independently: Christmas Eve and Pentecost I class and so outside the rule; Ascension, Assumption, John the Baptist and Peter & Paul II class; St Lawrence the sole III-class vigil. Five entries, the same five, same class split. It confirms the RULE and the ENTITY LIST, not the 10 days -- O'Connell publishes no day-by-day calendar. The distinction is stated in the note rather than blurred, since blurring it is what the withdrawn Ordo paragraph did. 2026-08-21T13:00:03Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T13:00:03Z urn:sha1:afd82f97ff440add51ea3e87c7ab53938443d5a0 C39 and the 2025 golden pin both claimed the published Ordo independently confirms RG 33's third trigger, because it omits St Lawrence's vigil on 2025-08-09 where lectio and missalemeum keep it. It does not. The Ordo omits that vigil on 2027-08-09 as well, where the clause says to keep it (10 August is a Tuesday, the feast holds its day), and 2026-08-09 is a Sunday where the rule's first trigger omits it anyway. The 2024 date has no row. So the Ordo never shows St Lawrence's vigil on any date it covers, and the 2025 agreement is a blanket gap coinciding with the right answer, not evidence. The rule itself is unaffected: it rests on RG 33's text, which admits no latitude, and on lectio not evaluating the clause at all -- "RG 33" appears nowhere in its production code, and the third trigger cannot emerge from per-day rank comparison because it turns on what is observed the following day. The divergence is architectural, not a disagreement about the rubric. Recorded rather than deleted: the withdrawn claim is why this entry was first written up as unusually well corroborated. 2026-08-21T12:46:34Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T12:46:34Z urn:sha1:8072287da9d8e8a9e2de8b7845923f5d576dea55 The note apportioned the entry's ten rows as seven commemoration-reduced and three transferred, counting 2038-06-23 in both halves to make the arithmetic work. It is six and four; the row count itself was always right, and is pinned by the comparator. 2026-08-21T12:42:36Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T12:42:36Z urn:sha1:651eb993240164123067001c053c9e0b8d970459 Agent research reports, SDD ledgers and review packages, same local-only reasoning as docs/. Untracked but not ignored until now, so they showed up in every git status. 2026-08-21T12:42:30Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T12:42:30Z urn:sha1:12b97761019cfa02ca0da8a5fb50ef815d07685c RG 33 omits a II/III-class vigil in three cases: it falls on a Sunday, it falls on a I-class feast, "vel si festum cui praemittitur in alium diem transferri aut ad commemorationem reduci contingat". Only the first two were built; the third was recorded in precedence_ef.ml as unimplemented on the grounds that no witness existed in the shipped data. That reasoning was wrong, and the rule fires on 1 744 days across 1583-9999. Both halves of the clause reduce to one observable question -- is the feast the OBSERVED office on the following day (RG 34 puts it there) -- so the kernel asks it once per candidate, after place_transfers has settled the year. No fixed point is needed: a vigil is a candidate only on its own day, never on its feast's, so suppressing it cannot change what the next day observes. Precedence.rules gains vigil_feast, which returns the slug of the feast a vigil precedes; the kernel cannot infer that itself, because only two of the five vigil/feast pairs share a slug stem. Blast radius, measured pre-change binary vs HEAD over the whole domain and classified: 1 744 days, three shapes, zero unexplained. 1 199 are the feast reduced to a commemoration (10 August on a Sunday, St Lawrence); 478 and 67 are the feast transferred under RG 96 after the Sacred Heart or Corpus Christi takes its day. The Assumption's and the Ascension's vigils never qualify -- their I-class feasts always keep their own day. Independently witnessed, which is unusual here. The published Ordo -- the only witness outside the Divinum Officium -> missalemeum -> lectio lineage -- omits St Lawrence's vigil on 2025-08-09, agreeing with colitur against both engines. That date had been read earlier as an Ordo gap; the Ordo was right, and correcting the misreading is what surfaced this clause. On 2027-08-09 the feast does keep its day and the Ordo omits a vigil colitur correctly keeps, which is a genuine Ordo gap. Allow-lists: C39 (lectio, 10 rows) and a 2038 oracle class citing the register, the 2026-2027 window having no instance. The golden pin asserting St Lawrence's vigil is violet moved 2025 -> 2027; its own comment had reasoned about the vigil's weekday and missed that RG 33 also looks at the feast's. Two new pins cover both shapes of the clause. The vigil/feast table is built with Slug.of_string_exn: mutation testing showed that of_string plus Result.to_option turns a typo into None, which this hook's contract reads as "not a vigil", switching the rule off in silence. Two tests assert the table against the shipped data in both directions. 2026-08-21T10:33:30Z Lukasz Kasprzak lukas@labunix.xyz 2026-08-21T10:33:30Z urn:sha1:0eae32f67145065b0e4960650ae06c293fd9f272 colitur.1 documented the flag and colitur-config(5) documented every [defaults] key except this one, which the previous commit added. Says what a flavour IS rather than only listing the values: the set of characters that must be neutered so a value cannot corrupt its target. A semicolon is ordinary text in LaTeX and a field separator in iCalendar; an unescaped & or _ does not look wrong in LaTeX, it aborts the build. Also why unset means INFER rather than 'no escaping', and why there is no silent fallback.