This IS definately an idea that has merit. A number of other people have discussed this issue. Tiled vector datasets work well for points, pretty good for lines, but run into some serious issues with polygons. There is already a python based tool developed by Microsoft Research Labs that breaks up point & line datasets into small javascript files containing object literals (I don't know if GeoJSON is used or some other format). You can find this at: http://research.microsoft.com/MapCruncher/TiledVectors/TiledVectorsDemo-1.1/description.html http://research.microsoft.com/MapCruncher/TiledVectors/TiledVectorsDemo-1.1/description.html
You can also read some ideas on this on the MapWrecker blog (the most interesting part is in the comments section): http://mapwrecker.wordpress.com/2007/07/17/well-know-vs-rational-vector-delivery/ post 1 , http://mapwrecker.wordpress.com/2007/08/24/rational-vector-delivery-pre-game/ post 2 hope that helps validate your idea and give you some perspective on what other's have thought about this problem. Matt Priour Kestrel Computer Consulting Guillaume Sueur-2 wrote: > > Hi, > > Has anyone of you had a thought about caching vector data in a web > format (GML, GeoJSON, GeoRSS...) ? > What I mean is a tilecache system for which one would set a tolerance > depending on the scale, and which would store the little files > corresponding to the generalized original data on the tile extent. > > Do you think the performances of vector display would be increased or is > this idea complete crap ? > > Thanks > -- > Guillaume SUEUR > > > _______________________________________________ > Dev mailing list > [email protected] > http://openlayers.org/mailman/listinfo/dev > > -- View this message in context: http://www.nabble.com/tilecache-for-vector-data-tp17339988p17342298.html Sent from the OpenLayers Dev mailing list archive at Nabble.com. _______________________________________________ Dev mailing list [email protected] http://openlayers.org/mailman/listinfo/dev
