Removed requirements for PoseableBatch-es to have a specified

number of rendered elements.

This was pretty easy since the elements were getting re-allocated
in the graphics card every frame. TODO would be to make the
re-allocation not happen every frame but rather only happen when
a re-allocation is nescessary and rather just not fully re-populate
the buffer if not every element is used or have some other sort of
automated way of handling the VRAM.

Example/MyBuiltinCubeScene.cpp

@@ -10,7 +10,7 @@ MyBuiltinCubeScene::MyBuiltinCubeScene(Application& application)
 	: Scene(application),
 	m_shape(meshgenerator::gen_cube_p<basic::Vertex, basic::Index>(DRAW_TRIANGLES, 2.0f, 2.0f, 2.0f), DrawMode::DRAW_TRIANGLES),
 	m_camera((float)TAU_1_4, (float)m_screen_size.x / m_screen_size.y, 1.0f, 10.0f, vec3(0.0f, 0.0f, 5.0f)),
-	m_batch(&m_shape, 2)
+	m_batch(&m_shape)
 {
 	m_pipeline.add_batch(&m_batch);
 	m_pipeline.set_camera(&m_camera);

Example/MyBuiltinLitScene.cpp

@@ -18,7 +18,7 @@ MyBuiltinLitScene::MyBuiltinLitScene(Application& application)
 		), DrawMode::DRAW_TRIANGLES
 	),
 	m_camera((float)TAU_1_4, (float)m_screen_size.x / m_screen_size.y, 1.0f, 10.0f, vec3(0.0f, 0.0f, 5.0f)),
-	m_batch(&m_shape, 2),
+	m_batch(&m_shape),
 	m_light(
 		Position(0.0f, 2.0f, 2.0f),
 		PhongLight::Power(0.2f, 1.0f, 1.0f),

Example/MyBuiltinTexturedScene.cpp

@@ -23,8 +23,8 @@ MyBuiltinTexturedScene::MyBuiltinTexturedScene(Application& application)
 		texturegenerator::gen_quick_sampler()
 	),
 	m_camera((float)TAU_1_4, (float)m_screen_size.x / m_screen_size.y, 1.0f, 10.0f, vec3(0.0f, 0.0f, 5.0f), vec3(0.0f, 0.0f, -1.0f)),
-	m_cube_batch(&m_cube, 2),
-	m_sprite_batch(&m_sprite, 1)
+	m_cube_batch(&m_cube),
+	m_sprite_batch(&m_sprite)
 {
 	m_pipeline.add_batch(&m_cube_batch);
 	m_pipeline.add_batch(&m_sprite_batch);

Example/MyPongScene.cpp

@@ -49,11 +49,11 @@ MyPongScene::MyPongScene(Application& application)
 	m_life(meshgenerator::gen_rect_p<basic::Vertex, basic::Index>(DRAW_TRIANGLES, LIFE_THICKNESS, LIFE_THICKNESS), DRAW_TRIANGLES),
 	m_paddle(meshgenerator::gen_rect_p<basic::Vertex, basic::Index>(DRAW_TRIANGLES, PADDLE_THICKNESS, PADDLE_HEIGHT), DRAW_TRIANGLES),
 
-	m_outline_column_batch(&m_outline_column, 2),
-	m_outline_row_batch(&m_outline_row, 2),
-	m_ball_batch(&m_ball, 1),
-	m_life_batch(&m_life, STARTING_LIVES * 2),
-	m_paddle_batch(&m_paddle, 2),
+	m_outline_column_batch(&m_outline_column),
+	m_outline_row_batch(&m_outline_row),
+	m_ball_batch(&m_ball),
+	m_life_batch(&m_life),
+	m_paddle_batch(&m_paddle),
 
 	m_outline_top_pose(glm::vec3(0.0f, (m_screen_size.y - OUTLINE_THICKNESS) / 2.0f - OUTLINE_OFFSET, 0.0f)),
 	m_outline_bottom_pose(glm::vec3(0.0f, -((m_screen_size.y - OUTLINE_THICKNESS) / 2.0f - OUTLINE_OFFSET), 0.0f)),

Example/MyTestTextScene.cpp

@@ -31,7 +31,7 @@ MyTextTestScene::MyTextTestScene(Application& application)
 		texturegenerator::gen_quick_sampler()
 	),
 	m_camera(m_screen_size),
-	m_text_batch(&m_text, 1)
+	m_text_batch(&m_text)
 {
 	m_pipeline.set_camera(&m_camera);
 	m_pipeline.add_batch(&m_text_batch);

charcoal-builtin/PoseableBatch.h

@@ -14,39 +14,35 @@ namespace charcoal
 		class PoseableBatch : public builtin::Batch<VertexType, IndexType, 1, RenderableType>
 		{
 		public:
-			PoseableBatch(
-				RenderableType* renderable,
-				int element_count
-			) : PoseableBatch(renderable, element_count, element_count)
-			{}
-
-			PoseableBatch(
-				RenderableType* renderable,
-				int element_count,
-				int element_render_count
-			) : builtin::Batch<VertexType, IndexType, 1, RenderableType>(renderable, element_render_count), m_orientation_elements(element_count)
+			PoseableBatch(RenderableType* renderable)
+				: builtin::Batch<VertexType, IndexType, 1, RenderableType>(renderable, 0)
 			{}
 
 			virtual ~PoseableBatch() {}
 
 			void reset_rendered()
 			{
-				// Also clear the matrix array?
+				m_orientation_elements.clear();
 				charcoal::InstancedBatch<VertexType, IndexType, RenderableType>::set_render_count(0);
+				// Just thinking, maybe want to not make want set_render_count to be public... The end-developer
+				// could screw up things by using that function. Then again, it could allow for preventing of
+				// needing to resize the element buffers once that optimization is made.
 			}
 
 			void add_rendered(const Poseable& poseable)
 			{
-				// Maybe just push back and then increment the render count?
-				m_orientation_elements[charcoal::InstancedBatch<VertexType, IndexType, RenderableType>::m_render_count++] = poseable.get_orientation_matrix();
+				m_orientation_elements.emplace_back(poseable.get_orientation_matrix());
+				charcoal::InstancedBatch<VertexType, IndexType, RenderableType>::m_render_count++;
 			}
 
 		protected:
 			void setup_element_buffers() override
 			{
-				// This feels worthless because it is also done in update_element_buffers.
-				glBindBuffer(GL_ARRAY_BUFFER, charcoal::ElementBatch<1>::m_element_buffers[0]);
-				glBufferData(GL_ARRAY_BUFFER, m_orientation_elements.size() * sizeof(mat4), NULL, GL_STREAM_DRAW);
+				// This is (was) worthless because it is also done in update_element_buffers.
+				//glBindBuffer(GL_ARRAY_BUFFER, charcoal::ElementBatch<1>::m_element_buffers[0]);
+				//glBufferData(GL_ARRAY_BUFFER, m_orientation_elements.size() * sizeof(mat4), NULL, GL_STREAM_DRAW);
+				// TODO: Find a way to not have to do a re-allocation every frame.
+				// While we're like this, I guess the adding and resetting of rendered can be changed though!
 			}
 
 			void resize_element_buffers() override