Feature/globebrowsing speedup (#735)

* Removal of ChunkRenderer, ChunkedLodGlobe, ChunkCuller, chunklevel evaluator, culling classes, layershadermanager, GpuLayer, GPUData, ChunkNode, Grid, BasicGrid, Chunk files, Angle, AABB classes, PointGlobe, Ellipsoid, TileSelector, tiledatatype, iodescription, simplerawtilerreader, rawtilereader
 * Less dynamic allocation for SkirtedGrid, LayerManager, RenderableGlobe, TextureUnit
 * Clean up memory management in RawTiles
 * Code simplification
 * Optimize shader uniform setting
 * Introduce UniformCache
 * Callback simplification
 * Turn ChunkNode into a struct
 * Use a MemoryPool to organize all ChunkNodes rather than use unique_ptr and the necessary memory allocation
 * Collect draw calls
 * Consolidate fragment shaders between local and global renderer
 * Shader cleanup and optimization
 * Update CMake to not include included shaders
 * Integrate traversal function into the looping
 * Replace std::queue with std::vector
 * Merge TextureContainer into MemoryAwareTileCache
 * Lazy computation of chunk bounding boxes
 * Memory management of LayerGroup
 * Remove class hierarchy from tileproviders (temporaltileprovider not working yet)
 * Remove PBO classes
* Chunk status stored in Chunk
* Don't create a copy of the ChunkTilePile
* Enable culling by projected area on default
* Have raw tile reader return a Tile instead of a shared_ptr to a tile
* Start making GDAL mandatory
* Increase the default lod scale factor to 15

modules/touch/src/touchinteraction.cpp

@@ -41,9 +41,8 @@
 #include <glm/gtx/quaternion.hpp>
 
 #ifdef OPENSPACE_MODULE_GLOBEBROWSING_ENABLED
-#include <modules/globebrowsing/globes/renderableglobe.h>
-#include <modules/globebrowsing/globes/chunkedlodglobe.h>
-#include <modules/globebrowsing/geometry/geodetic2.h>
+#include <modules/globebrowsing/src/basictypes.h>
+#include <modules/globebrowsing/src/renderableglobe.h>
 #endif
 
 #include <cmath>