I just saw on the Miro whiteboard the suggestion of an experiment for putting 3D Tiles into GeoPackage. Personally, I think this is a bad idea. Why? First even members of the OGC GeoPackage Standards Working Group (SWG) are worried about GeoPackage becoming a grab bag for any and all possible geo content. They also worry about extensions ad nauseam. Putting on my OGC Architecture Board member hat, I would strongly question this as a valid approach. FYI, I would say the same thing if I3S were proposed in this way. I think a better way would be to provide a logical model of how and where structured 3D content/models could be placed in a CDB X data store. This is much as OpenFlight files are handled now but in a more open and general approach so that glTF, CityGML, InDoorGML, COLLADA and other formats could be easily integrated into the data store. I do not think overloading poor GeoPackage is the best answer :-) Then the experiment demonstrates how 3D Tiles files can be encoded into the structure defined by the logical model.
Remember that the original design goal for GeoPackage was as an easy to implement, light weight container for vector geometry and tile matrices for raster data. A number of key extensions were defined, such as for R-Trees and Tiled Gridded Coverage Data.
I just saw on the Miro whiteboard the suggestion of an experiment for putting 3D Tiles into GeoPackage. Personally, I think this is a bad idea. Why? First even members of the OGC GeoPackage Standards Working Group (SWG) are worried about GeoPackage becoming a grab bag for any and all possible geo content. They also worry about extensions ad nauseam. Putting on my OGC Architecture Board member hat, I would strongly question this as a valid approach. FYI, I would say the same thing if I3S were proposed in this way. I think a better way would be to provide a logical model of how and where structured 3D content/models could be placed in a CDB X data store. This is much as OpenFlight files are handled now but in a more open and general approach so that glTF, CityGML, InDoorGML, COLLADA and other formats could be easily integrated into the data store. I do not think overloading poor GeoPackage is the best answer :-) Then the experiment demonstrates how 3D Tiles files can be encoded into the structure defined by the logical model.
Remember that the original design goal for GeoPackage was as an easy to implement, light weight container for vector geometry and tile matrices for raster data. A number of key extensions were defined, such as for R-Trees and Tiled Gridded Coverage Data.