Raster tiles to vector tiles: how to cross the gulf
summary
Migrating raster to vector tiles proves messy not magic: vector tiles enable dynamic styling and smaller packaged geometry but hit a hard 500 KB per tile Mapbox limit and force choosing a client (Mapbox GL JS, OpenLayers, Leaflet, Tangram). Practical fixes include generating a separate label layer with geojson-polygon-labels, using Tippecanoe with per-feature minzoom to trim data, using tile-server-gl or MapTiler to emit raster fallbacks, and prebuilding MBTiles to eliminate a backend — demonstrated on a million-tree site that now runs serverless.
AI-written summary · about
chapters (8)
- 0:00Why use vector tiles
- 1:01Working with constraints
- 3:17Labeling technical challenges
- 5:18Displaying terrain data
- 7:12Handling edge cases
- 8:43Data optimization strategies
- 11:14Architectural shifts
- 14:15Q&A and final thoughts
related talks
15mMaking maps work on mobile
15mHave you checked your vector tiles lately?
06mSwitching to PMTiles
21mPoint-in-polygon lookups with vector tiles: quick, effective, and no server needed!
49mClustering in the Mapbox Tiling Service
37mStreamline data pipelines with incremental tileset updates in the Mapbox Tiling Service (MTS)
46mCreating vector tilesets with the Mapbox Tiling Service
21mMake a map in minutes: How Mapbox is streamlining map development