Hoping for some review of these changes. I simplified the centering logic (at least I find it easier to read) and it seems to work in Python 2, and since it's not using any division I presume the same for Python 3.
Coverage decreased (-0.03%) to 91.054% when pulling 4fdb07af8e039992e788a67e1a8bef894b36600f on camera-fix into f1ef0af47fd6f30fabf83433a0a5df6d50f27f2b on master.
Hoping for some review of these changes. I simplified the centering logic (at least I find it easier to read) and it seems to work in Python 2, and since it's not using any division I presume the same for Python 3.