>>>>> "md" == Maarten Deen <md...@xs4all.nl> writes:
md> Adding a zoomlevel adds 4 tiles for each tile in the previous zoomlevel. md> You'll go from 91 billion tiles to 366 billion. md> This meaning you need 4 times the load to generate, 4 times the storage to md> hold them, 4 times the traffic to display them. I wonder if it would be sensible to switch to a vector-based rendering system for high zoom levels, as implemented at http://cartagen.org (using HTML5 canvas). This would allow "arbitrarily" high zoom levels. Tile-based rendering performs much better at low zoom levels, but I suspect that vector-based rendering would work better above z17. -- Eric Marsden _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk