Hey, just curious if there's any cost/benefit factors not immediately obvious if DGD were to expand its floats from 36 bits to the full 64 bits. I think it could gain on precision and if converted properly wouldn't cause anything to choke in conformant code that checks float.h