Skip to content

Geographic and mailing fixtures

These fixtures describe test inputs, not the operator's installation. Runtime receiver locations remain explicitly saved per receiver/path; tests do not seed those records or change an existing installation.

Public reference points

test/fixtures/publicGeography.ts uses station points from the pinned NWS nws-geography-v2026-04-16.json dataset. The weather tests check that the constants match the dataset. They are reporting-station coordinates, not claims about surveyed airport reference points or an actual rfSentry receiver.

Reference Latitude Longitude Purpose
KGTU 30.68333 -97.68333 County/zone membership, receiver distance, synthetic UAT ground coordinates
KBOS 42.36056 -71.01056 Distant receiver-history reference and distance/bearing endpoint
KAUS 30.18304 -97.67987 Surface CPR point with a nearby reference offset by 0.05 degrees

The NWS KGTU station page identifies the public airport reporting point. The repository's pinned station and polygon data provide the exact test snapshot; current weather observations are not required to run the tests.

KGTU belongs to county TXC491, zone TXZ173, and office KEWX in that snapshot. Synthetic warning products use those real codes to exercise local coverage, cancellation, expiry and mixed local/distant segments. KGTU to KTPL is approximately 58.5 km, preserving both the 75 km nearby case and the 10 km distant case. These are synthetic products, not recorded warnings.

KGTU is about 68.5 km from the center of grid EM10; KBOS is over 2,000 km away. The receiver-history tests retain those bounds, capture-time selection, and unknown-receiver behavior. The KGTU-to-KBOS great-circle check uses a mean Earth radius of 6,371.0088 km: about 2,697.760 km and an initial bearing of 53.8272°. Its original rounding tolerances remain unchanged.

The UAT fixtures encode and check the same public point using the existing 24-bit angular scale. They do not assert that a real ground uplink exists at that point. Existing wire-format vectors, mathematical grids, and official geographic datasets retain their own purpose and are not deployment locations.

Controlled records and geometry

The FCC relationship and address-presentation records use fictitious Fixture Alpha, Exampleville, test-only callsign strings and non-deliverable postal placeholders. They exercise case/punctuation normalization, current versus prior callsign precedence, exact name/postal overlap, unrelated-person rejection, PO-box deduplication and preservation of individual address fields. They are not anonymized records of a real licensee. TX, IL and their FIPS representations remain valid state-code inputs where the parser needs them.

The geocoder unit dataset is a wholly synthetic street grid, with synthetic locality and ZIP keys. Its unequal-vertex line spans longitude 0 to 0.1 on the equator; address 150 in the 100–200 range must resolve to longitude 0.05. This preserves the distinction between distance-weighted interpolation and vertex-count interpolation. It makes no claim about a real Texas street or post office. The integration test likewise constructs temporary synthetic data.

Evidence and scope

Deployment narratives explicitly omit the original locality. Their retained results and open acceptance gates still refer to the original execution; no qualification at these public fixture locations is implied. Public airports in fixtures are reference points, not deployment locators.

The FCC benchmark uses the neutral lexical term license, with its existing workload bounds and search semantics. Earlier benchmark measurements retain their original workload provenance; this cleanup does not claim a new timing run.