Plus additional goodies, such as
- equirectangular camera support
- OpenGL 3.1 support
- more accurate voxel pruning for training
- cleaner GUI

Can also pass datasets, network configs, or snapshots as positional arguments now.

Makes f-theta and latlong modes more meaningful and paves the way for a future Orthographic mode

Courtesy of Eric Haines

- Makes only a small performance difference (~20%) and matches original NeRF behavior. Remove the need for --nerf_compatibility when wanting to draw rough parallels to original NeRF.
- Also adds a GUI slider to adjust cone tracing angle live.

mesh export with python api and run.py support; added PLY export as well as OBJ; added vertex color export (sRGB); fixed exception when loading nerf with camera path json in same folder; added UI when failing to load snapshot or camera path; some UI code consolidation; change render aabb ui to min/max rather than center/rad.

run.py still contained an early, subtly incorrect PSNR computation that has to do with how prior art in NeRF papers (incorrectly) interprets the PNG alpha channel as sRGB-space rather than linear space. This is further complicated by the fact that NeRF's raymarching procedure also operated in sRGB space (rather than linear), which is both non-physical and introduces a subtle discrepancy with traditional PNG background blending. The new run.py code works around all that -- as the code that generated the paper results already did.

The following line of code will now produce a PSNR of ~36.35, matching the paper and reflecting the actual quality of our results.
`./run.py --mode nerf --scene lego/transforms_train.json --test_transforms lego/transforms_test.json --nerf_compatibility`