Global/MVT acceptance tests: the rest of the config-level checks

art commented

2026-08-15 07:22:29 +00:00

Owner

Part of the #115 plan: the config-level (global) acceptance tests landed in #119 with two (every map_* table's geom column has a gist index; the densest z5 tile stays under 8 MB). This issue tracks the rest of the global/MVT checks the review comment asked for ("check that the MVT is sensible"). The general instruction from #115: a test is a sentence citing the R-number where relevant, a SQL query whose first column is the pass/fail boolean, and a timeout from {1, 10, 100, 1000}.

Candidates:

R30: a request below a layer's floor returns an empty tile, and a request above maxz serves the maxz table — callable per layer via the zxy_ functions on a fixed tile below the floor (cheap, and it exercises the MVT functions themselves, which no test does yet).
Per-source tile budgets: extend the z5 Stockholm check to each country source (zxy_is, zxy_se, zxy_pl, zxy_fi) and a couple of mid-zoom tiles, so a runaway layer in ONE country fails that country's run rather than the shared check.
Empty-vs-missing: an all-ocean tile must be genuinely empty (0 bytes) — R30's sibling, catches a layer serving its floor table to every request.
Index coverage for the per-country prefixes (ismap_, se_, pl_, fi_), not just map_* — once #12's orphan cleanup lands, the legacy tables stop being a reason to scope the check.
MVT attribute set: a decoded tile's keys must match the contract's mvt_exclude-adjusted attr_names (R121's structural half) — needs tile decoding in SQL, so likely a 1000s-bucket candidate or a mapsrv-side test.

Part of the #115 plan: the config-level (global) acceptance tests landed in #119 with two (every map_* table's geom column has a gist index; the densest z5 tile stays under 8 MB). This issue tracks the rest of the global/MVT checks the review comment asked for ("check that the MVT is sensible"). The general instruction from #115: a test is a sentence citing the R-number where relevant, a SQL query whose first column is the pass/fail boolean, and a timeout from {1, 10, 100, 1000}. Candidates: - R30: a request below a layer's floor returns an empty tile, and a request above maxz serves the maxz table — callable per layer via the zxy_<layer> functions on a fixed tile below the floor (cheap, and it exercises the MVT functions themselves, which no test does yet). - Per-source tile budgets: extend the z5 Stockholm check to each country source (zxy_is, zxy_se, zxy_pl, zxy_fi) and a couple of mid-zoom tiles, so a runaway layer in ONE country fails that country's run rather than the shared check. - Empty-vs-missing: an all-ocean tile must be genuinely empty (0 bytes) — R30's sibling, catches a layer serving its floor table to every request. - Index coverage for the per-country prefixes (ismap_*, se_*, pl_*, fi_*), not just map_* — once #12's orphan cleanup lands, the legacy tables stop being a reason to scope the check. - MVT attribute set: a decoded tile's keys must match the contract's mvt_exclude-adjusted attr_names (R121's structural half) — needs tile decoding in SQL, so likely a 1000s-bucket candidate or a mapsrv-side test.

art referenced this issue

2026-08-15 07:23:15 +00:00

#115: Formalized acceptance tests. #119

art referenced this issue

2026-08-15 07:23:27 +00:00

Formalized acceptance tests. #115

art-bot was assigned by art

2026-08-15 08:55:04 +00:00

art referenced this issue

2026-08-15 09:47:42 +00:00

#93: places dedupe picks the cluster's highest register tier before font size #126

art referenced this issue

2026-08-16 00:22:19 +00:00

#122: Sweden's remaining R-checks as acceptance tests (R38, R40/R125, R90/R91, R95/R96, R119) #134

art added the

claude-wip

label

2026-08-16 01:05:59 +00:00

art referenced this issue

2026-08-16 01:17:24 +00:00

R121's decoded-tile check needs a mapsrv-side MVT decoder (not a SQL acceptance test) #138

art referenced this issue

2026-08-16 01:17:27 +00:00

mapsrv/importer tile tool: R121 decoded-tile check + standardized benchmarks + size validation #139

art referenced this issue from a commit

2026-08-16 01:19:40 +00:00

#125: the rest of the global/MVT acceptance tests in conf/common.yaml

art referenced this issue from a pull request that will close it,

2026-08-16 01:19:42 +00:00

#125: the rest of the global/MVT acceptance tests in conf/common.yaml #140

art commented

2026-08-16 01:19:54 +00:00

Author

Owner

Result is in #140 — #140; the review note is the PR description. Shipped: R30 floor/ceiling via the zxy_map_* functions, per-country z5/z8 tile budgets, all-ocean empty check, index coverage extended to ismap_*. The MVT attribute-set (R121 decoded-tile) half cannot be a SQL test (MVT is protobuf, no decoder in Postgres) — recorded here and filed as #139 (mapsrv-side).

Result is in #140 — https://git.blahonga.org/art/ismap/pulls/140; the review note is the PR description. Shipped: R30 floor/ceiling via the zxy_map_* functions, per-country z5/z8 tile budgets, all-ocean empty check, index coverage extended to ismap_*. The MVT attribute-set (R121 decoded-tile) half cannot be a SQL test (MVT is protobuf, no decoder in Postgres) — recorded here and filed as #139 (mapsrv-side).

art referenced this issue from a commit

2026-08-16 05:39:43 +00:00

#125 review: true rationale for the all-ocean test, and two robust ceiling probes

art

2026-08-16 05:40:02 +00:00