Potree shouldn't care about annotated-lanes anymore. Separate from potree code. From a data side, mostly on the aptiv bucket, I see we have to take care of three cases:
If both annotated and lanes exist
annotated lanes should be changed to lanes, lanes should be changed to original-lanes
If only lanes exist
Also make a copy of lanes.fb to original-lanes.fb
If only annotated-lanes exist:
Copy that file to both lanes.fb and original-lanes.fb and delete annotated-lanes.fb
So essentially, the important bucket in my mind is Aptiv. Either we change these filenames in the aptiv bucket for: SH 1,2,3 datasets, GJ_F 1,2,3,4 datasets, SATA1 and 2 datasets, and Woodward. Or you have potree handle the three cases above? I think the latter option is a bad option.
The script you mentioned to query S3 directories is a good idea. I imagine it will be helpful also in the future.
Potree shouldn't care about annotated-lanes anymore. Separate from potree code. From a data side, mostly on the aptiv bucket, I see we have to take care of three cases:
lanes,lanesshould be changed tooriginal-lanesSo essentially, the important bucket in my mind is Aptiv. Either we change these filenames in the aptiv bucket for: SH 1,2,3 datasets, GJ_F 1,2,3,4 datasets, SATA1 and 2 datasets, and Woodward. Or you have potree handle the three cases above? I think the latter option is a bad option.
The script you mentioned to query S3 directories is a good idea. I imagine it will be helpful also in the future.