Currently we recalculate all positions based on the tvgs position and viewport every frame.
This is obviously "slow", therefore we should implement a "dirty" system, where the tvg keeps track if something changed (position, scale, line width scale, viewport) and updates itself once instead of every frame.
Currently we recalculate all positions based on the tvgs position and viewport every frame.
This is obviously "slow", therefore we should implement a "dirty" system, where the tvg keeps track if something changed (position, scale, line width scale, viewport) and updates itself once instead of every frame.