> >(3) Consider deliberating choosing INTEGER PRIMARY KEY values based on >a "Morton code" or "Z-Order curve" of the coordinates. >(https://en.wikipedia.org/wiki/Z-order_curve) That will cause >features that are close together geographically to tend to be close >together within the file.
My primary key is actually the Id of the specific object in the OpenStreetMap-database, and we also need this osm-id in the app. I was trying just to order all rows by the "Z-Order curve"-value first before inserting them. for each (line in lines.ordered.by.zvalue) { insert line into lines insert line into lines_rtree } But this doesn't seem to work. At least, I don't see any improvements in page-usage. So is it just the value of the primary key controlling in which page the row is stored? _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users