Clone this wiki locally
The tips and tricks here are oriented to BYOD (bring your down data) mapping and OpenStreetMap obliquely.
Presentation cartography and basic inventory mapping is the focus here. Spatial analytics is touched upon briefly but is not the main focus. This site makes assumptions that OpenSource GIS mostly means: PostGIS as a spatial data store and Mapnik as a rendering engine, a Mappy CSS styling language like Cascadenik or Carto to setup the layers and their appearance. Interactivity is briefly touched on. Actually embedding the resulting tile assets is briefly touched on.
Looking for an overview of OSM instead? Check out the Switch2osm site.
Software - Install critical software and links to tutorials.
Data - Download OSM and Natural Earth data.
Example workflow - Start a project, import data, export data, create tiles.
Web mapping = API + tiles - Slippy maps on the web have two parts: images tiles that make up the map, and an API framework that stitches tiles together and handles interactive panning and zooming.
Why go open source geo - Read what the community is saying.
Map examples - Pretty basemap tiles and compelling interactive maps.
When tiles aren't enough - Use your web map stylesheet to render larger raster images and export to PDF, EPS, and SVG.
Tutorials - From Dymo, PostGIS, to Imposm.
Map Styling - Color your map features using cascadenik or carto since raw Mapnik XML is hellish.
Rendering map tiles - Once you're done designing your map, you'll need to render 10s of thousands of tiny image files, hundreds and thousands of megabytes in size.
Hosting Tiles - Publish your tiles so others can see them. Amazon S3, EC2, CloudFront, your own server, MapBox.com, etc.
Setting up an EC2 Ubuntu map server - Work in the cloud, it's so fluffy!
Upping your game:
Map Labels - Tips for labeling features.
Style Hub - Pre-baked stylesheets and icons for drawing OSM and Natural Earth data.
Adding interactivity - UTF-8 grids + hybrid tile vector maps
Is your map effective? - Take a critical look at your work, evaluate it with a consistent rubric, and ask, "Is my map great?"
Viewing map tiles in desktop GIS apps - For QGIS, ArcGIS, and Manifold.
Map scales/zooms, coordinate systems - Web maps have 20 preset scales, learn these “zoom” levels and their natural scale equivelents. Also more general map projections.
Generalize your data - Seriously. Your maps will look better and load faster.
Tools of the trade:
QGIS - The open source alternative to ArcMap, but not as industrial strength. It's a GUI for composing maps and doing simple analysis.
PostGIS - Store your geography in a spatial database (kinda the open source version of an Esri GeoDB). Faster than raw SHP files when speed matters. Storage default for OSM data.
OGR, GDAL - Reproject, sort, filter and otherwise modify Shapefiles and other vector map data formats. GDAL for raster. These are amazingly powerful.
TileMill - Desktop "web" app with an easy installer, powered by Mapnik and Carto. From Development Seed / MapBox.
Mapnik - Draw web map tiles, industrial strength.
TileStache - Cache web map tiles, from Stamen.
Dymo - Create beautiful map labels, from Stamen. Like Annotation from ArcGIS.
Shapely - Python library for geometric objects, predicates, and operations without requiring PostGIS.
pprepair - Validate and automatically repair planar partitions.
GeoJSON - Make it smaller. Giving your data the precision it's accuracy deserves.
Kartograph - Kartograph is a new framework for building interactive vector map applications without Google Maps or any other mapping service. It was created with the needs of designers and data journalists in mind. Tons of map projections, gracefully degrades to IE7. Great for thematic maps that are smaller scale than 1:250,000.
MaPublisher - Make maps in Adobe Illustrator.
OpenStreetMap - aka OSM. How to import it and explore the tag universe.
Geocoding - To geocode is the hardest part about going open source geo.
Routing - The MapQuest Open API returns routing based on OSM data
Shaded relief - set of utilities for processing DEM (digital elevation models, digital terrain models) and shading that to indicate the shape of mountains and other relief features.
Land cover - set of utilities for processing classified TIF images and colorizing techniques to indicate type (or presence) of vegetation and land use.
Wicket - View WKT geometries on web map.
Unix command line
Makefiles - Key to organizing your workflow and making it repeatable.
Python - Makes almost everything else possible.
General Unix commands - Including screen.
GREP - Data formatting problems? Advanced text find-replace to the rescue.
- Markdown is the devil - Yet another random raw text formatting style without a GUI.