Skip to content
Sniffari

Sniffari

Our OpenStreetMap data

Last updated 7 August 2026 · All legal & data pages

Sniffari's seeded place layer is built from OpenStreetMap — © OpenStreetMap contributors, available under the Open Database License (ODbL) 1.0. This page is our license notice, and — because ODbL §4.6 asks publishers to share their derived database or the method of making it — the complete method is documented below.

What is OSM data, and what isn't

  • The seeded place layer is OSM-derived — names, locations, and categories of imported places. We never edit OSM-origin values in place; community corrections are stored as separate override records, which keeps the OSM layer intact and re-syncable.
  • Community contributions are not OSM data — reviews, photos, check-ins, dog-friendliness verifications, and places added by members are Sniffari content, held in a separate layer alongside the OpenStreetMap-derived database (a "Collective Database" in ODbL terms).
  • The basemap is neither — the satellite imagery is © Esri, Vantor, Earthstar Geographics, and the GIS User Community, and the place-name labels drawn over it are © Esri, HERE, Garmin, © OpenStreetMap contributors, and the GIS User Community. Administrative boundaries are presented as supplied by our map data vendors; their alignment does not imply endorsement by Sniffari, Esri, or any governing authority.

How we derive our layer (the method)

Sources. Two pipelines, one method. Country-scale imports use Geofabrik daily country extracts (*-latest.osm.pbf, one national load per country in our live coverage) processed with osm2pgsql (flex output) applying the selection and mapping below; other regions are loaded from the live Overpass API with the equivalent query, per administrative area. Re-syncs run monthly on both paths.

Selection. A feature is a candidate if any of these hold: it matches one of the always-selected tags (leisure=dog_park/park/garden/nature_reserve, natural=beach, craft=winery, the pet-service tags below); it is a café/restaurant/pub/bar/fast-food or one of the six accommodation tourism values and carries a dog key; or its dog tag is one of yes / leashed / unleashed / designated / permissive.

Category mapping. First matching rule wins:

Sniffari categoryOSM tags
off_leash_arealeisure=dog_park
parkleisure=park · leisure=garden · leisure=nature_reserve
beachnatural=beach
wine_estatecraft=winery · shop=wine
cafeamenity=cafe
restaurantamenity=restaurant · pub · bar · fast_food
accommodationtourism=hotel · guest_house · chalet · apartment · camp_site · caravan_site
trailroute=hiking · highway=path · highway=footway
Pet services (a separate map class — not dog-friendliness destinations)
animal_hospitalamenity=veterinary · healthcare=veterinary, where the name matches /hospita+l/i
vetamenity=veterinary · healthcare=veterinary (all remaining)
groomingshop=pet_grooming · craft=pet_grooming
pet_shopshop=pet
boardingamenity=animal_boarding

Filters and flags. Destinations tagged dog=no are dropped (services keep no dog tag — it isn't meaningful for them). A destination whose dog value is in the explicit set above is marked as explicitly dog-tagged; everything else imports as a "help verify" candidate until a member confirms it on-site.

Retained fields. Per feature: OSM type + id (for re-sync), name, derived category, location (point), the dog tag, the feature's complete OSM tag set (from which display fields — address, contact, website, and amenity flags such as off-leash and fenced — are derived), and the sync timestamp. Country codes come from the extract's own country, or by point-in-polygon against Natural Earth boundaries (public domain) where an extract spans multiple territories.

Download

Machine-readable extracts of our unmodified OSM-derived layer are being published as monthly builds and will be listed here with checksums. Until then, the method above reproduces the layer from public sources, and we will provide the current extract on request: support@sniffari.app.

Questions

License questions or attribution concerns: support@sniffari.app.