feature request: option on double precision position values

General Discussion
1

i don’t know whether it’s impossible, i wish all position values in Panda are stored as double float.
i know double float values require more memory and CPU time, so can it be an option provided to SDK users?
i know graphic cards only accept single float position values, but values can be converted from double to single only at the final step to the graphic cards. the concept is the same as keeping higher res images and downscale them only at the final output.
double float may work slower than single float, but there is something that only double float can do.

2

I think using doubles instead of floats is already an option in panda, you have to use “–override STDFLOAT_DOUBLE=1” in your build command of panda I believe (Didn’t test it, maybe there is an easier way?)

3

Yes, it is possible today, and the instruction tobspr gave is correct: add --override STDFLOAT_DOUBLE=1 to the makepanda command-line when compiling Panda3D from source.