A map looks like a picture, but an open map database is not a picture. It is a structured store of facts about the world, coordinates, shapes, and descriptions, from which pictures, routers, and analyses are all derived. Understanding that one distinction explains everything else in this corner of the map world: why the same data can render as a road map, a cycling map, or a hiking map; why you can ask it questions a picture cannot answer; and why "the open map database" is actually several databases with different anatomies and different rules. This guide opens the anatomy first, using a single cafe, then compares the three databases an ordinary project will ever meet.
Home / GIS and Open Map Data for Beginners
Open Map Database: What Is Actually Stored, and How the Big Ones Differ

Photo: Ömer Tosun / Pexels
Coordinate Explorer: Touch One Record in the Database
Every click reads one point from the open OpenStreetMap database: a node with coordinates, the smallest record this article takes apart. Ways and relations are built from records like this one.
Map data © OpenStreetMap contributors. Coordinates are read from the map itself, in the WGS 84 system your phone uses.
The Anatomy: Nodes, Ways, Relations, Tags
OpenStreetMap, the database most people mean, stores the world in three element types and one labeling system.
A node is a single point: a latitude, a longitude, an ID, and tags. A bench, a tree, a cafe entrance, a mountain summit: if it can be a point, it is a node.
A way is an ordered list of nodes, and it means either a line, a road, a river, a fence, or, when the list closes back on itself, an area, a building, a park, a lake. The same structure does both jobs; closure is the difference.
A relation is a group of nodes, ways, and other relations with roles: a bus route threading forty road ways in order, a boundary assembled from a dozen border segments, a lake with an island, where the outer ring and inner ring need each other to mean anything.
Tags are the descriptions: key=value pairs hung on any element. Tags are where the meaning lives. The geometry says where; the tags say what.
One Cafe, Stored End to End
Make it concrete. A cafe on a corner exists in the database as one node at its location, carrying tags like these:
```
amenity=cafe
name=Corner Cup
cuisine=coffee_shop
opening_hours=Mo-Fr 07:00-18:00
website=https://example.com
```
The node gives the where. The tags give the what, in a vocabulary any application can read: a router can ignore it, a restaurant guide can feature it, an analysis can count it, which is exactly what project three of our GIS projects for beginners does with an Overpass query for amenity=cafe. The street outside is a way, a node list tagged highway=residential and name=Elm Street. The building the cafe sits in is a closed way tagged building=yes. Three elements, a handful of tags, and a renderer can already draw a recognizable corner, while a database query can answer questions no drawing could: every cafe open before 8 on weekdays, within 500 meters, on a street named Elm. That is the difference between a database and a picture, and it is why the OpenStreetMap API serves data, not images, to programs.
The Three Databases You Will Actually Meet
| OpenStreetMap | Overture Maps | Natural Earth | |
|---|---|---|---|
| What it is | A community built database of everything mappable, edited daily by volunteers | A foundation project combining company and open sources into standardized themes, places, buildings, transport, addresses | A curated cartographic dataset built for making world and regional maps |
| Structure | Nodes, ways, relations, free-form tags | Themed datasets in fixed schemas, released as data files | Tidy vector layers at set scales (large, medium, small) |
| License | ODbL: free to use with attribution and share-alike on derived databases | Per source, mostly permissive licenses; check the theme you use | Public domain |
| Update rhythm | Continuous; the living database changes every minute | Monthly releases | Occasional, deliberate revisions |
| Best use | Local detail, questions, analysis, anything needing current, rich data | Application ready places and buildings data at scale | Beautiful base maps of countries, continents, and the world |
The choice rule is simple once the anatomy is clear. Need the cafe's opening hours, the trail that changed last month, the answer to a question: OpenStreetMap. Building an application that needs millions of place records in a clean, uniform shape: Overture exists for exactly that. Drawing a wall map of a continent where the data must be cartographically tidy rather than exhaustive: Natural Earth, public domain and proud of it. Confusing them produces the classic disappointments: Natural Earth has no cafes to query; OpenStreetMap's raw tags are gloriously inconsistent compared to Overture's schemas; Overture's monthly rhythm cannot show you yesterday's new road.

A Database Is Not a Map, and That Is the Point
The map picture on openstreetmap.org is one rendering of the database, made by one style sheet's choices. The same database renders as a cycling map that elevates bike paths, a humanitarian map that elevates water points, and a transport map that elevates railways, none of which changes a single node. When someone says a place "is not on the map", the database question is whether the feature and its tags exist; the picture question is whether this renderer chose to draw it. Features can exist undrawn, which is why data tools find things the default picture hides, and why editing the database, through an editor, is a different act from reporting a picture problem. Our open map data explainer covers the licensing that makes all of this legally shareable, and the OpenStreetMap download guide covers taking the database home, from a neighborhood extract to the full planet file.
Living Data: How Fresh Is Fresh?
OpenStreetMap's database is alive in a way the others are not: edits land continuously, and derived products refresh on their own schedules, minutely for some extracts, daily or weekly for most downloads and applications. That rhythm has a practical consequence for your projects: record the date of any extract you analyze, because a count of cafes is a count as of a day, and the honest caption says which. Overture's monthly releases make dating automatic. Natural Earth's deliberate pace makes dating nearly irrelevant, which suits its purpose; continents change slowly, cafes do not. Choose your database partly by the clock your question keeps. A question about the world as it is this morning belongs to the living database, with all its tag level messiness forgiven, because the alternative databases simply have not been told yet.
Who Writes It, and How a Record Earns Trust
A database anyone can edit raises an obvious question: why trust a record? The honest answer is that you trust it the way you trust a path worn across a field: because it is used, watched, and corrected in public. Every element in OpenStreetMap carries a history: which account created it, when, in which changeset, and every later edit, all readable through the element's history view and through the API itself. A cafe node placed six years ago by a local mapper, refined by three others, and confirmed by a recent survey note is a different grade of fact from a node dropped yesterday by a brand new account, and the database lets you see the difference. Vandalism and mistakes happen, and they are handled the way open projects handle them: other mappers revert, discuss, and repair, usually fast in populated areas, slower in remote ones, which is why experienced users weigh local detail in busy places heavily and treat a lone record in an empty region as a claim awaiting a second witness.
That transparency is a structural difference from the other two databases, not a footnote. Overture's records arrive through its member organizations' pipelines, with provenance tracked per source but no public edit history for an individual feature. Natural Earth's features are curated by its editors on a release cycle, authoritative in the way an atlas is authoritative, and equally uninspectable at the level of one river bend. None of these models is simply best. For a classroom map of the world, curation wins. For an application ingesting places at scale, pipeline uniformity wins. For knowing whether the footbridge at the end of your street reopened last week, and being able to check who says so, the living, inspectable database wins, and it is not close. Choose by the question, again.
There is a second trust habit worth building: cross-checking between the databases. Where OpenStreetMap and Overture agree on a place, confidence rises. Where Natural Earth's coastline and OpenStreetMap's disagree, the difference is usually scale, one dataset drawing for a continent, the other for a street, and knowing both purposes dissolves the argument. The mature way to use open map databases is not loyalty to one but fluency in all three, with the question choosing the source and the date stamped on whatever you take. That fluency is the real subject of this pillar of the site, and the OpenStreetMap download guide is the practical next step: it puts an actual extract of the living database on your own machine, where you can open it, query it, and watch a cafe become a record with your own eyes.
Related Guides
Readers working through this topic usually need its closest companions too. These GIS and open map data guides cover the neighboring ground:
- QGIS Print Layout: Compose a Map Someone Can Actually Use: The map on your QGIS canvas is a working view; the map in a print layout is a finished argument, with a fixed scale, a legend that explains it,...
- The QGIS Georeferencer: Pin a Scanned Map to the Real World: A scanned historical map is a picture until you georeference it: teach QGIS a set of ground control points, places whose real coordinates you...
- Overpass Turbo: Query OpenStreetMap Like a Database, From Wizard to Export: Overpass Turbo turns OpenStreetMap into an answerable question: every bench in a town, every drinking water point along a trail, every bakery in a...
Frequently Asked Questions
What is an open map database?
A freely licensed, structured store of geographic facts, coordinates, shapes, and descriptions, from which maps, routers, and analyses are built. OpenStreetMap is the best known, storing the world as nodes, ways, relations, and tags. "Open" refers to the license: you may use, copy, and build on the data under its terms, with attribution.
What are nodes, ways, and relations in OpenStreetMap?
A node is a single point with coordinates. A way is an ordered list of nodes forming a line, or a closed area when it loops. A relation groups nodes, ways, and other relations into a larger structure, like a bus route or a boundary. Tags, key=value pairs, attach meaning to all three.
What is the difference between OpenStreetMap and Overture Maps?
OpenStreetMap is a single community edited database with free-form tagging, updated continuously, under the ODbL license. Overture Maps is a foundation project that combines company and open sources into standardized, themed datasets, places, buildings, transport, addresses, released monthly under mostly permissive per-source licenses. OSM is richer and fresher; Overture is more uniform and application ready.
Is Natural Earth really public domain?
Yes. Natural Earth is public domain data, built by cartographers for making world and regional maps, with tidy layers at set scales. Its trade is detail: it is curated for cartography at small scales and carries none of the local, queryable richness, cafes, benches, opening hours, that OpenStreetMap holds.
Can I download an entire open map database?
Yes. OpenStreetMap offers the full planet file and regional extracts, updated regularly; Overture distributes its themes as data files through cloud storage; Natural Earth downloads as layer packages. Our OpenStreetMap download guide walks the sizes and formats. Expect the planet file to be large, and an extract of your region to be the sensible first download.
Why does a feature exist in the database but not show on the map?
Because the map picture is one rendering with its own choices about what to draw at each zoom. The database can hold a feature, a footpath, a bench, a small business, that the default style omits or shows only when zoomed close. Data tools and other renderings, cycling, hiking, transport styles, draw different subsets of the same database.
Photo credits: Ömer Tosun (Pexels); ThisIsEngineering (Pexels). Photos are served from this site; follow the credit to the library page for the original.