Development Ideas

This is a list of ideas for future improvements of TopOSM. Feel free to contact me and/or edit this page if you have additional ideas, comments on existing topics, or want to implement one.

For bugs or minor things, you can also add an issue on GitHub.

Large projects

• Use Carto, Cascadenik, MapCSS or a similar language for styling. A more compact and expressive language would make styling tweaks much more manageable.
• Support on-demand rendering. The (experimental) tilestache branch already supports this, but significant performance improvements and cleanup would be required for this to be practical.
• Worldwide rendering. This would require additional data sources for elevation (SRTM, ASTER, ...?), and hydrography (fall back on OSM?) outside of the US.
• Localized highway shields, probably on a state level. See existing work on this topic, including Richard Weait's writeup (down?), Phil Gold (prototype here), and the OSM Wiki on US Road Signs.
• Remove dependency on mapnik patches. This might be achieved by using the compositing features in future mapnik versions (see e.g. Artem Pavlenko's smart halos experiments) or through a third party compositor like tilestache. Note that the tilestache branch already works correctly with a stock mapnik. Each adds some complexity and performance penalty, however.
• Try Mike Migurski's DEM-tools for hillshading? Other layers?
• Add an accurate map legend. Preferably, it would adjust per zoom level so that only currently relevant features are included. TopOSM-Massachusetts has an example of this (click show legend at bottom right.)

Smaller Ideas

• TODO: Fix nodata transparency for .vrt layers

Styling and Cartography

• Switch to using the NPS map symbols. (kkenny)
• Render additional point features to the map? (kkenny)
• Better trail visibility.
• Improved typography.
• Try point-in-polygon solution for labels as suggested by kkenny (while waiting for native mapnik support...)