Dos 20 achados do /simplify, 12 entraram na entrega e 3 ja estavam
cobertos (o modo `territory` pertence ao E12 por rabbit hole do plano, a
shoelace duplicada e descarte declarado, e o B5 F2 ja contava os 5 call
sites dos helpers de CLI). Sobraram dois fatos sem registro.
O revisor de perf mediu que `quantize` aplica um unico numero de digitos
aos dois eixos: a grade de ~103 m e exata em longitude e a latitude segue
o aspect ratio do bbox (102,8x109,2 m no estado, 102,9x80,1 m em
Salvador). Corrigir exige `{scale, translate}` explicito e reescrever os
tres artefatos, perdendo a prova de byte-identidade da malha estadual por
um erro de ~9% num eixo num mapa que para em maxZoom 10 — entao vira
gatilho no plano-pai, nao item de roadmap.
E o B8+ respondeu por precedente uma pergunta em aberto do B5 F2: com
`scripts/lib/topology.mjs`, o helper de cache da F2 pousa em
`scripts/lib/` em vez de abrir um segundo endereco. A recomendacao antiga
(`src/lib/cliCachedDownload.ts`) fica rejeitada: `src/lib/` e client-safe
por contrato e nao e lugar de codigo que faz mkdir/fetch/writeFile.
Sem novos IDs de roadmap.
F1 — a tolerancia de simplificacao e os digitos de quantizacao viraram
politica em unidades de solo em scripts/lib/topology.mjs, consumida pelos
dois scripts de geometria. Os quantis 0.35 copiados cortavam as 19 ZE de
Salvador a ~0.6 m2 e os 417 municipios a ~1.78 km2; com tolerancia
absoluta e a grade de ~103 m do estado, a malha das zonas caiu de 480 para
30.2 pontos/feature. bahia-municipalities segue byte-identico e o tripwire
de bytes deu lugar a pontos/feature.
F2 — featureMapKey passou a ser asserido sobre features reais das tres
malhas comprometidas (417 codarea, 27 code, 19 slug), e o comentario de
MAP_KEY_PROPERTIES agora diz o que o codigo faz: as propriedades sao
mutuamente exclusivas, nao ordenadas por especificidade.
F3 — a malha das zonas e um detalhe sobre o mapa, nao o mapa: se o chunk
falha, Salvador cai no poligono municipal que ja esta desenhado embaixo e
o resto da Bahia continua funcionando, com o motivo logado. A ressalva de
aproximacao saiu do gate de comparacao e passou a depender de
bundle.hasZoneMunicipalities, derivado no servidor de municipality.kind.
F4 — orfaos deletados: os dois acessores sem chamador, isZoneMapKey (o
servidor ja sabe a regra), o wrapper simplifyTopology e o helper
polygonCount com um branch inalcancavel. MunicipalityMapNavigation
colapsou em string | undefined e MunicipalitiesByMapKey em
Record<string, string>.
After B8 F2 polygons, the textual zone breakdown under the dashboard map is redundant; register the removal in the roadmap and align B8+/B8 plan notes.
The three /simplify reviewers reported after the merge, so their findings
became a fill-in instead of a fix. The lot is anchored on a measurement that
contradicts "small mesh, small artifact": the 19 Salvador zones carry 9,120 arc
points against 9,139 for ALL 417 Bahia municipalities, because both simplify
constants are relative to the file's own bbox and were copied verbatim from the
sibling script -- the same "1e4 grid" is ~103 m statewide and ~3.9 m in a city
26x narrower, on a map that stops at maxZoom 10.
That measurement inverts one reviewer's recommendation, which is why it is
written down rather than left to be re-derived: extracting the shared constants
as they stand would cement the defect and give it the authority of a policy, so
what gets shared between the two build scripts is the ground scale.
The other three: nothing runs featureMapKey over a real feature outside the
e2e, so the three spellings of the map key can drift and paint grey with no
error anywhere; Promise.all over both meshes plus the memoized rejection turned
a 19-polygon chunk into a single point of failure for the whole map, where the
municipal mesh alone used to suffice; and the "approximate, not the official TSE
boundary" caveat hangs off the zone LIST's render condition, so it vanishes in
comparison mode while the approximate mesh is still painted.
Also records what this delivery orphaned (getTerritoryFeature lost its last
production caller when fitMapToHighlights stopped consulting geometry modules,
and knip cannot see it because specs are entry points), and bumps B5 F2's
call-site counter to five.