First, thank you for open sourcing this excellent framework. It is allowing us to do things that would be near impossible in GE.
I see a similar problem has been addressed earlier in the forums, but was hoping to gain some insight on the issue.
Our problem set includes displaying up to 50,000 entities that are updated at intervals ranging from 100ms to several hours. When the camera distance is large we hide the details of the entity and show a single point. When the viewer is zoomed in we show a more detailed view of the entity including a label, billboard icon, path, and a line to to the entity’s children.
Performance on a laptop with a K3100 GPU is sluggish when I increase the number of entities above 10,000.
We send a bounding box to the server requesting updates only in the camera’s current FOV. **Is there an efficient way to remove all of the entities in a collection that are not in the current view? **Or another sane strategy to handle a huge number of dynamic entities.