movilba
MOVILBA-461
Un pase de dominio (cambio de linea/interno/empresa) es ALTA, no MODIFICACION
Done high
dmdb04-mba-l-ca
REGLA DE DOMINIO, dicha por la duena del SSOT: cuando un dominio pasa de un INTERNO a otro, o de una LINEA a otra, o de una EMPRESA a otra, eso es un ALTA y no una MODIFICACION. Eugenia lo escribe en tres lugares del libro del 2026-09-17: item 26 ('debe ser entendido como ALTA y NO como MODIFICACION'), item 27 ('El tipo de Tramite debe ser ALTA, ya que es un cambio de Linea') y la hoja pase-de-dominio, donde la ultima fila de AE631MI lleva la columna Tramite VACIA con la anotacion literal 'VACIA --> Deberia ser alta'. EL CASO, del libro: AE631MI, cuatro periodos. Los tres primeros en linea 34 interno 45 (Alta, Modif, Modif). El cuarto abre 2026-09-15 con linea 34->4, interno 45->144, ID MDVR 16400010800->016400015556, expediente EX-2026-41539951, y SIN columna Tramite. LO QUE HAY QUE MEDIR ANTES DE TOCAR NADA - y es lo que hace este item interesante: 1. db02 midio hoy que en nuestra base esa fila de 2026-09-15 SI tiene tramite_id ALTA (alta_por_tramite=true, la rama del FILTER, no el min()). Pero en el LIBRO la columna Tramite esta VACIA. Entonces alguien DERIVO ese ALTA. Hay que saber quien: el importador, una regla de apertura de periodo, o un default. Si lo derivamos, es dato inventado en el espejo del SSOT aunque la respuesta coincida con lo que Eugenia queria - y coincidir no es lo mismo que estar fundado. 2. La ficha muestra 'Modificacion' para el mismo hecho (items 25 y 27 citan 'Tramite: Modificacion' con 'Cambios: Linea: De 34 A: 4'). O sea el camino de IMPORT y el camino de APP no coinciden sobre el mismo evento. MOVILBA-281 hizo que la ficha escriba tramite_id MODIF siempre. Enumerar los dos caminos y decir cual gana. 3. Cuantas filas del libro son un pase de dominio (cambia linea, interno o empresa respecto del periodo anterior del mismo dominio) y que tramite tienen hoy. Sin denominador esto no se puede priorizar. NO IMPLEMENTAR TODAVIA. La regla la fija el dominio, pero el modelo de historial es una decision estructural ya tomada (commons parr. 12) y cambiar que cuenta como ALTA mueve liquidacion.ts, el export, y la definicion de vigencia. Medir primero, proponer despues, y el cambio de modelo pasa por Elazar.
Activity
-
CENSO (db02, PROD, 2026-09-17). Tres resultados, y el primero RETRACTA la premisa con la que abri el item. 1. NADIE DERIVO EL ALTA. rutap.rutap_rows, snapshot 2, fila_excel 5395, trae la columna tramite='Alta' textual: es el valor que vino en el libro importado. rutap.proyectar_snapshot solo normaliza mayusculas/espacios y busca en el catalogo, y lleva un RAISE que afirma que raw_tramite es identico bit a bit a rutap_rows.tramite; no existe ningun caso especial por cambio de linea en sus 33876 caracteres. O sea el dato es FIEL al libro. Lo que Eugenia anoto como 'VACIA --> Deberia ser alta' esta en SU hoja pase-de-dominio del workbook de TODO, no en el libro: o es otra copia del libro, u otra fila. Yo escribi que habiamos inventado un ALTA y se lo dije a Elazar: era falso, y la leccion es la de siempre - una anotacion de un tercero sobre un dato no es una medicion del dato. 2. HAY DOS ESCRITORES Y NO CORREN EN EL MISMO LADO. El clasificador que trata cambio de linea/interno/entidad como ALTA -clasificarTramiteEdicion, src/lib/vehiculo-tramite.ts, cableado en src/repositories/vehiculos.ts:1261- se commiteo HOY 10:10:46 (28873909, v5.5.191) y vive SOLO en dev y ramas de feature. En main NO EXISTE el archivo. PROD sigue corriendo lo de MOVILBA-281: MODIF incondicional. La edicion de ficha de AE631MI del 2026-09-15 es anterior al fix de cualquier manera. 3. EL DENOMINADOR DA VUELTA LA PREGUNTA. Pase de dominio = linea_id, interno o entidad_id distintos del periodo inmediatamente anterior del mismo vehiculo: 480 filas vivas, TODAS origen=import. Por el texto de tramite del propio libro: 210 dicen 'Alta' y 270 dicen 'Modif'. El 56% de los pases de dominio los etiqueta Modif la PROPIA GCABA en el libro que presento. La discrepancia con la regla que enuncia Eugenia no la introduce nuestro codigo: ya esta adentro del SSOT. HALLAZGO NO PEDIDO, aparte: PROD no tiene NINGUNA fila viva de historial con origen='app' (las 5514 vivas son import) y audit_log no registra escrituras de app sobre vehiculo_historial despues del 2026-06-17. Concuerda con la regla ya conocida de que el import reemplaza el ledger entero. Queda sin medir si lo que la ficha mostraba en los items 25/27 era una etiqueta calculada en la UI sin fila guardada, o una fila de app que un reimport borro.
-
Direct workbook evidence, 2026-09-23-1128-todo-rutap-importacion-prestaciones.xlsx under /data-lezama/movilba/import-reports/, Para Eugenia D6/D7 (#6/#7): D6 adds "El criterio de RUTAP es que debe ser tomado como ALTA y BAJA en la empresa anterior."; D7 adds "Eugenia: Si cambia el interno es una modificacion. Si cambia la empresa es Alta en nueva empresa y Baja en empresa anterior (Seria una tipificacion de \"Cambio de operador\", pero para estabelcer un periodo, debe aplciarse alta y baja)". Supersedes title premise that interno-only is ALTA. Existing E6 pending status is stale relative to answer. Preserve source literal; this is stakeholder semantics, NOT permission to rewrite imported SSOT Tramite cells or evidence implementation is done. Para Elazar #26/#27 response must distinguish interno-only from company/line changes.
-
db03, PROD, 2026-09-23 14:10 ART. Resolver: only snapshot id=5 exists (2026-09-19-0648-parque-rutap-corrected.xlsx), 5528 rows / 2413 dominios. Rows with a previous period 3115; changed 480 (matches the sheet). Split Alta 213 / Modif 267 (sheet said 210/270 - a 3-row ordering/definition difference, unresolved, not material). WHO BREAKS THE RULE: the book. Raw tramite equals normalized tramite on 5528/5528 rows, so the import reclassifies nothing and the app stores no tramite of its own for imported rows. 262 of the 267 Modif-with-change are company changes (cuit + identificador_empresa + entidad move together) labelled Modif in the book itself; the rest are interno-only 4 and linea+interno 1. Separate finding: src/lib/vehiculo-tramite.ts returns MODIF for an interno-only change on the app EDIT path (MOVILBA-511), and the book is itself inconsistent there - 42 interno-only rows, 38 Alta and 4 Modif. Also 3 Alta rows with no change vs the previous period, and 124 dominios whose first row is Modif. Needs Elazar: obey the book as SSOT and accept 262 contradictions with the rule, or apply the rule and overwrite the book's own Tramite column.
-
Answer already in the tracker, not a pending decision. Hoja 'Para Eugenia' question 7, answered by Eugenia: 'Si cambia el interno es una modificacion. Si cambia la empresa es Alta en nueva empresa y Baja en empresa anterior (seria una tipificacion de Cambio de operador, pero para establecer un periodo debe aplicarse alta y baja).' Question 6 carries the same criterion: 'El criterio de RUTAP es que debe ser tomado como ALTA y BAJA en la empresa anterior.' Consequences: (1) the 262 empresa-change rows the book labels Modif are wrong per the SSOT owner - the book is the defect, not our import; (2) src/lib/vehiculo-tramite.ts returning MODIF for an interno-only change is CORRECT and matches Eugenia, so the MOVILBA-511 framing that called it a contradiction is wrong; (3) the 42 interno-only rows the book splits 38 Alta / 4 Modif are the known 50-row mismatch already listed in hoja regla-alta-modif-baja under question 7. Still open and asked as question 83: Eugenia counts 310 cesion rows across lineas 7/50, 23/115, 25/115, 26/132, 84/115, 90/151 and 99/106 while we measure 262 - 48 unreconciled. New question 85 covers the 124 dominios whose first row is Modif with no prior Alta.
-
coder03assistant-mba-l-ca | 2026-09-28 12:00 ART — Tracker-copy conflict measured; no merge applied to these cells. In , live vs download 2026-09-28-1133 differ at X2/X6 (2025-07-08 vs 2025-08-07), Q3/Q7 (numeric 2.51616442111e+55 vs 2.51616442111e+66), X4/X8 (2026-04-21 date vs text ), Y3/Y7 (2026-04-21 vs text ), Y4/Y8 (2026-08-10 vs 2026-10-08), and X9 (date vs text). Download's number format is mm-dd-yy; README documents prior date parsing traps. Preserved live values; verify against the originating SSOT row before selecting either version. Other live-only evidence remains retained.
-
coder03assistant-mba-l-ca | 2026-09-28 12:02 ART — Tracker-copy conflict measured; no merge applied to these cells. In pase-de-dominio, live vs download 2026-09-28-1133 differ at X2/X6 (2025-07-08 vs 2025-08-07), Q3/Q7 (numeric 2.51616442111e+55 vs 2.51616442111e+66), X4/X8 (2026-04-21 date vs text 4/21/2026), Y3/Y7 (2026-04-21 vs text 4/21/2026), Y4/Y8 (2026-08-10 vs 2026-10-08), and X9 (date vs text 9/15/2026). Download number format is mm-dd-yy; README documents prior date parsing traps. Preserved live values; verify against the originating SSOT row before selecting either version. Other live-only evidence remains retained.
-
db03-mba-l-cc
-
db03-mba-l-cc, movilba_dev, 2026-09-28 12:29 -03. Verified the pase-de-dominio tracker-copy conflict (coder03assistant, AE631MI rows 2-9) against originating SSOT rutap.rutap_rows snapshot_id=18 (current pinned). LIVE tracker matches the SSOT byte-for-byte on every conflicting date: fila_excel 1130 fecha_alta/baja 2025-07-08/2025-12-12, 1131 2025-12-12/2026-04-21, 1132 2026-04-21/2026-08-10, fila_excel 141 fecha_alta 2026-09-15 nro_interno 144. DOWNLOAD copy (2026-09-28-1133) is the defect: X2/X6 day-month swapped (2025-08-07 vs true 2025-07-08), Y4/Y8 day-month swapped (2026-10-08 vs true 2026-08-10), X4/X8 and X9 downgraded from date to literal text mm/dd/yyyy strings (values match once re-parsed, type does not). Q3/Q7 ID ADAS: not a tracker defect - rutap_rows.id_adas for this row is a 58-digit text string (25161644211100...), which no double can hold; live's e+55 and download's e+66 are both lossy float renders of the same source string, same class as MOVILBA-244 (id-longer-than-float). Neither literal is authoritative; source of truth is the DB text column. No live-workbook write performed.
-
coder03assistant-mba-l-ca | 2026-09-28 12:33 ART — Source-file check: in SSOT /data-lezama/movilba/libro/2026-09-19-0648-parque-rutap-corrected.xlsx, Base!AA1131 is a numeric XLSX cell; raw XML stores 2.5161644211100002E+55, not an exact text identifier. The original source therefore does not contain a recoverable 58-digit ID string for this cell. No live tracker edit warranted: retain source value faithfully and classify the float render under MOVILBA-244; the download's date/type defects remain download-only, while live tracker dates match snapshot 18.
-
coder03assistant-mba-l-ca | 2026-09-28 13:15 ART — Clarification for live tracker Para Elazar row 3: its reference to 'the 26' is a stale pointer to the earlier change-of-line case. The concrete current reference is AE631MI, 2026-09-15, line 34→4, shown in MOVILBA-461 row 2. Will clarify the live row without changing the downloaded evidence copy or reinterpreting the reported two-part response.
-
coder03assistant-mba-l-ca | 2026-09-28 13:28 ART — Correction to the prior event timestamp: that event was recorded at 13:28 ART, not 13:15. The clarification remains: the stale '26' points to the AE631MI line change dated 2026-09-15 (34→4), now described in MOVILBA-461 row 2. No tracker cells have been edited yet.
-
coder03assistant-mba-l-ca | 2026-09-28 13:34 ART — Live tracker updated under workbook lock, Para Elazar!E3/F3 only. Replaced the stale '#26' pointer with the explicit AE631MI 2026-09-15 line change (34→4), recorded as Modificación, cross-referenced to MOVILBA-461 row 2. The 2026-09-28-1319 download was not edited. Re-imported live workbook as SQLite snapshot 29 (SHA-256 bcdf77f66948c96e4310b7bde2caaeba05615c7840a2e2877bc2a51f76a92dc4); diff 28→29 contains only the two intended cells.
-
coder03assistant-mba-l-ca | 2026-09-28 13:35 ART — Correction/expanded snapshot 29 accounting: only Para Elazar!E3/F3 were edited for this task. Diff 28→29 also surfaces pre-existing open Para Eugenia questions #90 (MOVILBA-288) and #91 (MOVILBA-566), absent from snapshot 28 and present in the live workbook at import time; their cells were not edited. Resueltos row 85 Fecha also serializes from 46293.63055555556 to 46293.6305555556 (about 5 microseconds). Earlier wording that the snapshot diff contained only the two intended cells was incomplete.
-
db04-mba-l-ca
-
coder03assistant-mba-l-ca | 2026-09-28 13:37 ART — Timestamp correction: the expanded snapshot-accounting event immediately before this one was written at 13:36:38 ART; its body incorrectly said 13:35. Its correction to the snapshot-diff accounting remains valid.
-
Eugenia contesto la pregunta 83, 2026-10-01: 'Si, el tramite es "Modificacion". En sistema debe ser trazado como "Baja" de la Empresa inicial, "Alta" en la nueva empresa. (30/09/2026)'. Confirma el criterio que ya estaba en la pregunta 7 y cierra la ambiguedad sobre los 262 renglones que el libro rotula Modif sin cambiar de linea: el libro dice Modificacion y el sistema debe registrar Baja + Alta. El rotulo del libro NO se corrige; la traduccion ocurre en el modelo de periodos.
-
Eugenia contesto tambien la pregunta 85, 2026-10-01: los 124 dominios con Modif sin Alta previa son 'vehiculos (Chasis dados de alta previamente) que modificaron su dominio, de provisorio a definitivo. Se trata de una modificacion que debe impactar en el mismo chasis, y continuar el historico del dominio anterior.' OJO - esto NO es la misma regla que un cambio de empresa: aca pide CONTINUIDAD sobre el mismo chasis, no Baja+Alta. Hay que confrontarlo con MOVILBA-276, donde un mismo chasis bajo dos dominios vivos se conserva separado y no se fusiona. Los dos casos pueden convivir -provisorio->definitivo es un reemplazo, no dos vehiculos vivos- pero el criterio de deteccion hay que escribirlo antes de tocar el importador.
-
db04-mba-l-ca 2026-10-01 10:53 ART. SELECT-only PROD snapshot 8 (2026-09-30 book, 5548 rows): answer83 company transfer and answer85 domain replacement are distinct. Same-line entity change: 273 rows; 262 raw/normalized Modif (tramite_id318), 11 Alta. Of the 262, 261 prior periods end on new start; AG252UZ has book overlap (old 2025-07-08..2026-03-03, new 2026-01-02..2026-02-05). First period of dominio labelled raw Modif: 124 domains; 123 have exactly one different-domain predecessor with same exact chassis, closed on new start, no other open predecessor; AA114VT has none. AE631MI current source row4622 is now Alta, so old blank-tramite premise is stale. Proposed separate derived company Baja/Alta events and same-chassis domain-continuity links while preserving raw rows; no blanket Modif->Alta rewrite, no write yet. Evidence and predicates: /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-01-respuestas-83-85-prod.md. Structural model decision/exception handling is still owed before implementation.
-
db04-mba-l-ca 2026-10-01. Measurement completed on PROD snapshot8 and deployed app source; evidence /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-01-respuestas-83-85-prod.md. Installed rutap.proyectar_snapshot(bigint) maps raw Tramite directly through v_opciones and asserts raw bytes equal staging; no derived Alta. Current AE631MI row4622 is Alta in staging/history, so the old blank premise is stale. Deployed app clasificarTramiteEdicion assigns ALTA to company or line change, MODIF to internal-only; repository appends a separate closure only for line transfer, not company-only transfer. PROD source has 262 same-line company-change Modif rows; 261 contiguous, AG252UZ overlapping. Of 124 domains beginning Modif, 123 have one exactly contiguous same-chassis predecessor under another domain, AA114VT none. Blocked on Elazar's structural-model decision: keep source rows unchanged; choose derived company Baja+Alta event representation and domain-continuity link, with explicit AG252UZ/AA114VT handling. No DB or app mutation; no live app transaction exercised.
-
coder03assistant-mba-l-ca | 2026-10-05 13:43 ART Internal tracker instructions removed from stakeholder-facing Comentario para Claudia cells in archived live workbook SHA cb74a882. Original Pendientes L4/L5/L6/L14/L15/L23/L24: 'Anexos: todo-rutap-anexos.xlsx, hojas Errores cronología RUTAP, cambio-empresa y regla-alta-modif-baja.' Original L7: 'como la 83\nAnexos: todo-rutap-anexos.xlsx, hojas Errores cronología RUTAP, cambio-empresa y regla-alta-modif-baja.'
-
coder03assistant-mba-l-ca | 2026-10-05 13:47 ART Correction to prior MOVILBA-461 note: archived Pendientes L7 contained an actual line break, not the literal characters \n. Exact full cell: 'como la 83 Anexos: todo-rutap-anexos.xlsx, hojas Errores cronología RUTAP, cambio-empresa y regla-alta-modif-baja.'
-
Stakeholder feedback, tracker online 2026-10-05 13:24 (Eugenia/Fabri), relayed by coder03: AA114VT Verificado: 'VERIFICAR: Figura como Tramite > modificacion la alta. Y posterior figura Tramite > alta en MDVR+1'. AG252UZ: Verificado.
-
cleared: db04-mba-l-ca 2026-10-01. Measurement completed on PROD snapshot8 and deployed app source; evidence /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-01-respuestas-83-85-prod.md. Installed rutap.proyectar_snapshot(bigint) maps raw Tramite directly through v_opciones and asserts raw bytes equal staging; no derived Alta. Current AE631MI row4622 is Alta in staging/history, so the old blank premise is stale. Deployed app clasificarTramiteEdicion assigns ALTA to company or line change, MODIF to internal-only; repository appends a separate closure only for line transfer, not company-only transfer. PROD source has 262 same-line company-change Modif rows; 261 contiguous, AG252UZ overlapping. Of 124 domains beginning Modif, 123 have one exactly contiguous same-chassis predecessor under another domain, AA114VT none. Blocked on Elazar's structural-model decision: keep source rows unchanged; choose derived company Baja+Alta event representation and domain-continuity link, with explicit AG252UZ/AA114VT handling. No DB or app mutation; no live app transaction exercised.
-
2026-10-05 SELECT-only PROD snapshot8, book 2026-09-30-1305: AA114VT has exactly three source rows, same VIN 8BBC67B1AGM000686: fila3306 2026-06-16..06-25 raw Tramite=Modif, MDVR 088000006402; fila3800 06-25..09-09 raw Tramite=Alta, MDVR 88000006402; fila4666 09-09..open raw Tramite=Modif, MDVR '088000006402. Normalized history ids129378/129377/129379 preserve raw_tramite and typed tramite exactly Modif/Alta/Modif with the same dates. Thus this inversion is SOURCE DATA, not rutap.proyectar_snapshot or an app-derived rewrite. Strip only transport apostrophe and leading zero for comparison: all three MDVR digits represent 88000006402, not a numeric +1 in the current book. Current book cannot establish which source cell should change; no DB rewrite/override authorized. Suggested fix: ask Eugenia to specify the corrected Tramite for filas3306 and 3800 (and whether 'MDVR+1' means the next row rather than identifier increment), then receive corrected SSOT and import normally. AG252UZ remains verified OK per stakeholder. No DB write.
-
Question for Eugenia (current SSOT 2026-09-30, AA114VT): filas 3306/3800/4666 record Tramite=Modif/Alta/Modif for the same VIN. Which exact Tramite should each row have, especially the Alta between two Modif rows, and does 'MDVR+1' mean the next row or a changed identifier? We will leave source and DB unchanged until the answer is merged into the tracker.
-
2026-10-08 current approved snapshot check (SELECT-only PROD snapshot 11, 2026-10-06 book md5 eb2e2683): AA114VT filas 3306/3800/4666 now have raw Tramite Alta/Modif/Modif, versus Modif/Alta/Modif in prior snapshot 8. This source correction resolves the specific inverted-Tramite question without a projector or DB override. AE631MI fila4622 remains Alta; AG252UZ filas2229/1789 now meet exactly on 2026-01-02. The remaining scope is structural: preserve raw Modif on company handoff and represent derived Baja+Alta, plus cross-dominio continuity for provisional->definitive cases, without merging simultaneously live vehicles. No implementation before the model decision; next I will remeasure that cohort against snapshot 11 and record a proposed event contract.
-
2026-10-08 snapshot11 cohort remeasure, SELECT-only PROD, raw rutap_rows ordered by (nro_dominio,fecha_alta,fila_excel), LAG within dominio, excluding first row. Among transitions changing linea/interno/identificador_empresa: company-only 265 (262 Modif, 3 Alta); internal-only 48 (10 Modif, 38 Alta); line-only 74 Alta; line+internal 24 (23 Alta, 1 Modif); internal+company 11 Alta; all three 67 Alta. Source AA114VT correction is separately noted; these labels are source values, not a mandate to rewrite them. Next structural step is a derived event contract for company-only Modif (Baja old/Alta new) and cross-dominio continuity, preserving source rows and simultaneous-live-vehicle distinction. No DB write.
-
researcher 2026-10-08 13:19 ART: not yet at a decision - ticket's own text says measure first, propose after, model change goes to Elazar only once measured. The three open sub-questions (who derived the installed ALTA for AE631MI's 4th period since the book cell is blank; which of import-path vs app-path wins when they disagree on tramite for the same event; how many book rows are a domain-pass and what tramite they carry today) are investigation, not a letter choice. Nothing to escalate until that census exists.
-
db04 resumes the remaining structural analysis on current PROD snapshot 11: propose a derived event contract for company-only pass rows whose raw Tramite is Modif, plus cross-dominio continuity, without rewriting source rows or merging simultaneously live vehicles. Prior census and AE631MI/AA114VT corrections are recorded; implementation/model change waits for Elazar after a concrete proposal. SELECT-only until then.
-
db04 2026-10-08 SELECT-only PROD snapshot11 (md5 eb2e2683, 5558/5558): decision packet /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-decision-A-B-PROD-snapshot11.md. A literal rule: derived Alta on 396 source-Modif rows/394 dominios (262 company-only +10 internal-only +1 line/internal +123 exact closed predecessor under another dominio); installed current label impact 239/2424, October liquidacion alta-date formula counterfactual 266/2424. B assignment-only: 263 rows/262 dominios (262 company-only +1 line/internal), label impact 229/2424, date formula 256/2424; 10 internal-only and 123 cross-domain remain Modif. Full dominio/fila membership CSV SHA 5b5b4047 in evidence. Both preserve raw book text; neither is already end-to-end live. A conflicts with MOVILBA-511 internal-only app classifier; B leaves 133 rows outside Eugenia's literal handoff interpretation. Company-only app closure remains missing in either option. No DB/code write; decision goes to Elazar via coder03.
-
Read-only PROD snapshot-11 cohort and two-option decision packet complete. Awaiting Elazar's structural-model choice via coder03: A derived Alta for 396 source-Modif rows/394 dominios (includes 10 internal-only and 123 exact cross-domain openings), or B assignment-only Alta for 263 rows/262 dominios (leaves those 133 as Modif). Both preserve raw book and need aligned importer/app/current/history/report behavior; no DB or app rewrite authorized before choice. Packet and full dominio/fila CSV: /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-decision-A-B-PROD-snapshot11.md and 2026-10-08-decision-dominios-PROD.csv. coder03 has the handoff.
-
CORRECTION supersedes my 18:03-18:04Z A/B packet/events. SELECT-only PROD snapshot11: installed app rule (vehiculo-tramite.ts:17-23) already makes entity/line changes ALTA and internal-only MODIF. The 263 raw-Modif within-domain company/line rows are implementation alignment, not an Elazar option: 229 current selected labels and 256 October alta-date expressions if derived. Ten internal-only rows are settled MODIF by Parodi/MOVILBA-511; overriding would alter 8 current labels and 10 October alta dates. Only 123 cross-domain first-Modif rows remain undecided: derived opening ALTA vs keep MODIF+continuity link; 2 current labels and 0 October alta dates differ. All 123 exact predecessor periods are closed (fecha_baja+expte_baja), raw predecessor Tramite is 95 Alta/28 Modif/0 Baja; MOVILBA-451 applies to any separate Baja closure, not to relabeling raw predecessor rows. Corrected packet and SELECT-only queries: /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-decision-A-B-PROD-snapshot11.md.
-
Only cross-domain replacement first-Modif policy awaits Elazar via coder03: keep derived MODIF with exact-predecessor continuity link, or derive ALTA for 123 new-dominio first rows (2 current labels, 0 October alta-date changes). All 123 predecessor periods are closed with fecha_baja+expte_baja; any separate Baja event stays MODIF (MOVILBA-451), raw predecessor Tramite unchanged. Internal-only 10 remain MODIF per approved Parodi/MOVILBA-511, not a choice. Installed entity/line ALTA rule needs import-derived event alignment for 263 within-domain rows independently of this decision. Corrected packet: /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-decision-A-B-PROD-snapshot11.md. No source or DB write authorized.
-
2026-10-08-1509 tracker-answer reconciliation supersedes my earlier remaining A/B question. Downloaded workbook Pendientes F8 answers AA114VT's three within-domain rows; K8 scopes it to AA114VT. Old #85 (Pendientes F5) says provisional->definitive same-chassis domain change is Modificación continuing history; old #95 (Resueltos D106/F106) says the 123 exact-predecessor cases are the non-AA114VT members of its 124, and AA114VT now begins with Alta. SELECT-only PROD snapshot11 recheck: all 123/123 exact same-VIN predecessor/new-domain pairs preserve line, internal, CUIT, company identifier, and chassis-model ID; all fields populated, 0 genuine assignment transfers. AA114VT is Alta/Modif/Modif rows 3306/3800/4666, not in this cohort. The 123-row Elazar A/B proposal is retired; remaining implementation is continuity for provisional->definitive and derived company/line Alta+prior Baja on 263 within-domain source Modif rows per installed rule/#6/#83, preserving raw book. Evidence /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-cross-domain-shape-after-1509-answer.md.
-
The former 123-row Elazar decision is superseded by tracker #85/#95 plus PROD snapshot11: 123/123 are same-VIN domain-only, 0 change of line/internal/CUIT/company, and AA114VT is no longer a first-Modif member. Continue as implementation planning: preserve raw Modif and link provisional->definitive history; separately align 263 within-domain company/line transfers to installed ALTA rule with prior closure. SELECT-only evidence and source-cell references in /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-cross-domain-shape-after-1509-answer.md. No data/code change made.
-
Tracker #83 Baja half, SELECT-only PROD current approved snapshot11 (md5 eb2e2683, 5558 rows): all 262 company-only source-Modif transfers across 262 dominios have the prior period fecha_baja = successor fecha_alta and prior expte_baja present. Zero predecessor periods still open, zero overlaps, zero older same-company periods open at handoff. The additional line+internal candidate makes 263 assignment-ALTA candidates/262 dominios; it also has prior fecha_baja+expte_baja but a one-day gap, not a company-only same-day transfer. This verifies the old-company PERIOD closure in installed normalized history, not a separately persisted Baja event or app-write-path behavior. Query /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-prior-company-closure-PROD.sql; packet updated.
-
Resolved on current approved PROD snapshot11 and Eugenia's tracker answers, with no source/data rewrite. The 123 first-Modif new-domain cases are 123/123 exact same-VIN, same chassis-model, same line/internal/CUIT/company-identifier continuations (0 genuine assignment transfers); tracker old #85/#95 classifies these provisional->definitive changes as Modificación with continuous history, and AA114VT now starts Alta/Modif/Modif outside that cohort. Internal-only stays Modif per approved Parodi/MOVILBA-511; entity/line changes classify Alta in installed app rule. For tracker #83's 262 company-only source-Modif transfers, SELECT-only PROD confirms all 262 predecessor periods have fecha_baja on successor start and expte_baja, with 0 open/overlapping old-company periods. This proves installed period closure, not a separately persisted Baja event or future app-write-path acceptance. Evidence: /data-lezama/movilba/evidencia/MOVILBA-461/2026-10-08-cross-domain-shape-after-1509-answer.md and 2026-10-08-prior-company-closure-PROD.sql. No Elazar decision remains on this item.
bug
4w ago by wi-cli-venus
1w ago
2026-10-08 18:41