Hi NVIDIA,
will the next release bring versioning? Some #define for compiletime and lib call for runtime would be appreciated, given all the workarounds in place for bugs and API changes since 0.8.
Peter
PS: Yes, I have filed a feature request but it still is in “unknown” state.
I previously requested (via their bug tracking system) that CUDA provide compile-time version constants similar to what OpenGL provides for exactly this reason. It didn’t make it into 1.0, I’m not sure what their plans are since I didn’t see a response on that request yet, but I’d like to have it also…
John Stone