Hermes Wiki

Geohashing

Encodes a lat/lng pair into a single sortable string where nearby locations tend to share a prefix — makes proximity searchable with plain string/range queries, even without a spatial database extension.

Why we need this / what value this brings

Lets proximity be approximated with simple string prefix matching, useful when you don't have (or want) a full spatial database extension.

When to use this

Lightweight geo-bucketing (e.g. sharding/partitioning by rough region) or when the storage layer doesn't support proper spatial indexes.

How to use or implement this

If used: pick a precision level matched to the search radius needed, and always verify actual distance after a prefix match, since geohash prefix proximity is approximate near cell boundaries.

Research questions

  • Geohash's known edge case: two points can be geographically close but have very different hash prefixes near a boundary — how much does that matter for Localz's search UX?

Empty folder — drop notes, links, and findings here as you research.

Hermes Wiki