2 de 130 pares de chasis cambian tambien interno y entidad: no es un re-patentado, parece transferencia entre operadores (o chasis compartido)
db-mba-l-cc + coder02-mba-l-cc, 2026-08-20, snapshot 4, SELECT-only. De los 130 pares chasis viejo/nuevo, 128 son re-patentados limpios: mismo interno, misma entidad, misma linea, patente vieja en Baja y patente nueva en Alta. 2 NO. Ejemplo medido, chasis LA6C3K1GXTB102089: AH846RR Alta interno 709 entidad TRANSPORTES NUEVE DE JULIO SAC contra AI058QQ Baja interno 478 entidad LA NUEVA METROPOL. Cambian interno Y entidad, o sea o es una transferencia entre operadores (que es un tramite distinto de un cambio de patente y deberia quedar registrado como tal) o son dos vehiculos distintos que comparten el string de chasis - y en ese segundo caso cualquier join por chasis los abanica. El segundo par falta identificar. Clasificar los 2 antes de proponer nada; no drafteear SQL. Contexto: la cadena por chasis ya esta en prod (v_vehiculo_cadena, MOBAAPI-689) y asume que un chasis = un vehiculo.
Activity
-
Second pair identified (db-mba-l-cc + audit-mba-l-ca agree, SELECT-only): chasis LA6C3K1GXTB101976 - AH846RG Alta interno 708 TRANSPORTES NUEVE DE JULIO linea 109, vs AI058QN Baja interno 61 LA NUEVA METROPOL linea 65. Same shape as LA6C3K1GXTB102089. In BOTH pairs the Alta side is Nueve de Julio/linea 109 and the Baja side is Nueva Metropol on a different linea - a consistent pattern across both, not noise, and it reads as operator+linea change rather than a plate-only replate. Note this is the one case the chasis-group gate does NOT cover: audit measured 2247 groups in snapshot 4, 153 with exactly 2 dominios, max 2 - so the chasis join never fans out, but 'no group over 2' does not prove every group of 2 is one vehicle. These 2 are exactly that doubt.
-
Verified correct, no defect, no write. db-mba-l-cc pulled the full RUTAP chain: there are THREE tramites per chasis, not two - prod only ever held two rows. LA6C3K1GXTB101976: AI058QN (Nueva Metropol, linea 65, interno 61) -> AH846RG (same operator, plate renewed, same interno/linea) -> AH846RG (Nueve de Julio, linea 109, interno 708, current/open). LA6C3K1GXTB102089 identical in shape. One physical unit, two plate changes and one operator transfer. Prod's Alta row already carries the FINAL tramite's interno/entidad/linea; it collapsed the middle same-plate/same-entidad renewal into no extra row, which is correct since the dominio did not change on that hop. The Baja row correctly holds the closed first-plate period. Kept as a documented edge case for the chasis-chain view: apparent interno+entidad divergence across a chasis pair can be a legitimate 3-tramite chain seen through prod's 2-row projection, NOT two vehicles sharing a chasis string.