Hi, I'd like to use mris_flatten_cuda to accelerate the flattening process. However, it seems that the code doesn't handle ripped vertices correctly. mris_flatten by itself works correctly, but attempts to run mris_flatten_cuda results in thousands of "PANIC: RIPFLAG IS NOT SUPPORTED !" messages. The result is a completely folded surface.

Peering through the code, it seems that the MRISremoveRipped command is being run (line 236 in mris_flatten.c), but somehow doesn't operate correctly. Is there any way to get the cuda version to run correctly, or is that not supported at this time?

-James Gao